[ Interview ] ProfitWell’s Patrick Campbell on the Pros & Cons of a Multi-Product Strategy in SaaS

A few weeks back, I spoke to Patrick Campbell for The Lean B2B Podcast. We talked about metrics, customer research, pricing, segmentation, and the pros and cons of having a multi-product strategy.

You can watch the full interview below, or access it on iTunes, Google Play or Spotify.

Interview Transcript

Patrick Campbell – Multi-Product Strategy

Etienne Garbugli: My guest today is Patrick Campbell. Patrick is the founder of ProfitWell. ProfitWell is a business intelligence platform helping more than 8,000 subscription companies reduce churn, optimize their pricing and grow their subscription business end to end.

Patrick is an economist by training. Before starting his company in 2012 he worked in data and strategy at Google and Gemvara in Boston.

Patrick, welcome to the podcast.

Patrick Campbell: Yeah. Awesome to be here. Talk about research, all kinds of fun stuff today. Excited.

Etienne Garbugli: So, ProfitWell has a really interesting story. It started out bootstraped providing services, helping businesses with their pricing strategies, but the vision was always to be a product company from what I understood. How were you going about figuring out what the product should be initially?

Patrick Campbell: Yeah. So this is one of the biggest misconceptions about our company, and it’s kind of our fault if it’s a misconception, right?

So we, we actually ended up, starting off as a pure product company. Um, so we had this piece of software that we started off with that it did, essentially pricing research. So you build these surveys and you send it out. Um, and then what we discovered is, basically people, they liked the data. And they liked the output. They didn’t want to do the work to get the output, and they didn’t want to,, how do I put it? They didn’t want to make decisions without talking to a human. So they wanted like, services and things like that on top of it. And we were like, Oh, that’s dumb. We don’t want to do that. And then it was like, well, no, we’ll pay you money. And we’re like, Oh, okay. So it was one of those things that kind of worked out in a good way.

And then, you know, when we were doing our research, we wanted a better way to get data than just surveys. And so we, we started building other products and things like that. And that’s what came about with ProfitWell Metrics. And now we have a whole suite of products out there.

Etienne Garbugli: So, at what point did you guys start thinking about adding multiple products? Like was it the, what made you think that this was the right strategy?

Patrick Campbell: So, um, yeah, that’s a good question. I think it really just came down to the math and, and just the velocity. So what I mean by that is, we realized very quickly that our TAM, our total addressable market from a logo perspective was actually quite small. There’s only, you know, depending on how you measure it, about 150,000, maybe only a hundred thousand subscription companies.

And when you cut that by SaaS, you know, mainly the B2B SaaS market, then all of a sudden you’re talking about 20 to 30,000 companies, which is not a lot from a market perspective. And so. You know, the advantage of that market is the, the revenue on all of these businesses is growing exponentially, right? It’s compounding as the world of subscriptions and recurrent revenue does, which is our market. And so we looked at it and we said, okay, well, to be a hundred million dollar company, a billion dollar company, we can’t sell $50 a month products to a lot of people, right? We have to sell like higher priced.

We needed high LTV, and the best way to get a high LTV is through good pricing and good monetization as well as, good retention, obviously. But then, you know, the way that you increase that ARPU [ Average Revenue per User ] or that ACV [ Average Contract Value ] further, the average revenue per user, or the average contract value is basically by adding more products, right?

And so it kind of created this thesis of being multi-product. And we’ve taken this on really early in our life cycle. Most companies don’t take this on until they’re $100 million company and we’re, we’re a $10 million company. So, it’s kind of fascinating.

Etienne Garbugli: But so how did the opportunities for ProfitWell Metrics, Retain and Recognize come about? How much effort did you guys put into finding the next product opportunity?

Patrick Campbell: Yeah. So, the metrics product actually. So, like I kinda said before, we were looking for a product that we could, so survey data is amazing. A lot of people misunderstand survey data. They don’t know how to clean data properly and, we’re terrible as human beings or as operators at sending surveys, but the people who are really good at it and they realize how much power is there. The issue that we were having is it’s a higher cost way to get data.

And so we were thinking, well, there were two things, like one, we were thinking, well, how do we get the data we need at a lower cost? And what that became was, well, what if we had access to their financial data? Or what if we created personas? And it was like, well, then we would need financial data probably top of the funnel data, engagement data. Ooh, all of a sudden it started to become, you know, really, really, really complicated, right? And then it’s like, um, okay, well, uh, how do we get that data in a better way? Well, we’re always going to have to start with financial metrics, right? So that was kind of a little bit of a thesis that we’re having. And then all of a sudden, we were helping a company that was about to IPO with their pricing, and we discovered that they were basically calculating MRR churn completely incorrectly. It was a public company that was about to be public. And so, we kind of started of putting the two together.

And we just were like, oh, well we have multiple things that are showing us like that this is something that we should go into. And that’s when we jumped into ProfitWell Metrics. And then everything else was like, very much like the flow of information and the flow of resistance, right? Which was like, oh, like, this is not good, let’s solve it. Or, hey, we’re right next to being able to solve this. Let’s do that. Or this is what our customers are screaming, you know, all types of things.

Etienne Garbugli: So, one opportunity came out of the previous and the previous, and then the previous, and then… It was a sequence of different opportunities that came out from the previous.

Patrick Campbell: Yeah, exactly.

Etienne Garbugli: Okay. And how much effort did you guys put into finding what the next opportunity or the next product opportunity should be?

Patrick Campbell: Um, so yeah, that’s a really great question. It’s really hard to quantify effort, right? If it was physics, it’d be a lot easier, right? It’s hard. Yeah, it’s hard. It’s hard to quantify just because you’re always like rationalizing why you’re doing things right, and so you’re like, well, of course. Of course, we put a lot effort into it because we just rationalize why we’re doing it, and it’s like, well, hold on a second. No, maybe we didn’t put that much work into it.

So I think for us it’s like. The effort, the flow starts with a thesis, right? And it’s like, Hey, I have this thesis. And that thesis is probably based on something that someone said or like some feedback that we got or someone we really trust, like being like, Oh, you guys should totally do this, or something like that.

And then what ends up happening is, is our CPO, our head of product and I, we then just start debating. Oh, we should do this, or we shouldn’t do this, or we shouldn’t do that. Right? And so there’s a lot of effort that actually goes there and it’s very much a qualitative, and yes, there’s some quantitative aspects that, you know, we go through, like with this of, Hey, well why would we go into that space? Well, let’s look at the size of the space. Well, this, these three articles, say the size of this space is this. Like, that makes sense. Or that handles that objection. So, there’s a lot of debate that goes on. And I think that a lot of companies, this is where they should start, but they don’t because it’s uncomfortable to debate. And we’ve created a pretty big, you know. A culture of debate and feedback.

Then what we ended up doing is let’s say we’re like, cool, we’re going to go after it. Then we write a narrative, which is basically like a memo of sorts, or we just go through like X, Y, and Z of this is the logic behind it.

And then what we end up doing is if the logic makes sense and we send it to like another group of people internally. We then go and, we then start doing market research. Now, this market research, will probably start with just qualitative interviews. Hey, how are you doing this right now? What does this look like?

Kind of like being a prosecuting attorney. Like you’re not asking them directly, do you want this, right? And I think that’s another problem is people don’t do the narrative, they don’t do the debating. And then they just go out and be like, Hey, do you want this? And then what you’re doing is you’re just, you’re just doing the path of least resistance, which is, I’m just going to build the thing that someone says.

And it’s like, yeah, but there might be one person, they might not be articulating things properly. Like you have to be that researcher to basically dig into it. And then after those conversations we’ll start to have a clear thesis on some of the details of what we’re doing. And then we’ll do some quantitative research.

Sending out some surveys to our customer base and then like, disregarding certain people’s answers if they’re, you know, Hey, we know this isn’t going to be a startup product. This is for big dogs. So, we have to like, get big dog data and things like that. And then, you know, what ends up happening is, and then there’s this iteration cycle, right?

So, we’ve decided on this big thing that we’re going to go after, and now there’s just constant iterations. And those iterations have their own research cycles. And their own debates and things like that. So like right now for our Retain product, which is a churn reduction product, we did those big debates and the big narrative like years ago at this point.

Now it’s like, hey, should we go into this part of the market or should we add this feature? And so those cycles are shorter because the overarching thesis has been thought through and now it’s just more detailed work. Even if it’s a big feature or something like that. So hopefully that’s helpful.

I mean, that’s kind of the process that we follow and it probably isn’t as linear as I’m saying. It’s probably all over the place and in some cases as is the case when you’re building something.

Etienne Garbugli: No, no. It’s great. So, is there an overarching persona or like a profile that you’re specifically targeting across these opportunities?

I there like a unifying tread behind these opportunities in terms of the targeting or the market that you guys are going after? Like you mentioned SaaS, but within SaaS?

Patrick Campbell: Yeah, so we’re subscriptions targeting. So that’s not just SaaS. So, it’s DTC or direct to consumer, which includes subscription e-commerce, subscription media, subscription nonprofits, all kinds of stuff.

And so I think that the big thing to kind of think through for us has been. Multiple products means you typically will have multiple personas, but maybe the same company you’re targeting. So, this has made it a complicated, and we admittedly have not figured out all of this from a go to market perspective or even a product perspective because you can’t just spin up four teams for four products, right?

Like, this is just not how it should work, especially when you’re, we’re a bootstrapped or a customer-funded company of about 80 folks. And we have four different products, right? So like, that’s a lot to like. Yeah, a lot of surface area. And so, you have to, you’re probably going deprioritize a product on some level and then you have to come up with some framework of targeting.

Now, from a go to market perspective, they’re all subscription companies and we tend to target companies that are doing tens of millions, if not more of revenue. But we get a lot because we have a free product, we get a ton of startups and we love them and we like them. They like us, they’re getting a ton of value.

But because of that LTV problem that I was describing, we need to go after like more upmarket companies, and then each product has different personas. So, on the Price Intelligently product that typically is targeting marketing and product folks, and normally the head of product or the head of marketing unless it’s like $100 million plus company then it might be a director of some sort.

Retain product is customer success or product. And then our Recognized product is very, very much, the CFO or the. There’s not really Chief Accounting Officers, but the lead accountant at a company. And so, it’s kind of interesting too, because we use the free product as basically a signal.

So, depending on who comes in, you know, cause it might be the founder, well, if it’s a founder at a smaller company, we’ll go at them with a certain product, right? Probably Retain, if it’s a founder at a larger company, we’ll probably be do an enterprise type deal where we try to sell them the suite.

But if it’s customer success or CFO or something like that, we’ll lead with like those different products mainly because those are the ones that they’re, they’re most interested in.

Etienne Garbugli: So, in a way, the free product helps qualify a certain set of prospects and then the behavior within the free product or the people that sign up will help you qualify what the next step should be.

Patrick Campbell: Yeah, 100%. So, basically the idea and that sounds really elegant and smooth. It’s not that elegant and smooth. So, it’s just one of those things that there needs to be a lot of, a lot of adjustments depending on how things, how things go forward. But it’s one of those things where, you know, it’s, it’s at least like, that’s, those are the, not only the constraints, but those are the, the playing field or the framework that we’re working with, which allows our team to kind of work within it.

Etienne Garbugli: So, you mentioned before, like that a lot of times businesses will start thinking about a multi-product strategy later on. You guys did it a bit earlier, so when you think is the right time to start thinking about that specifically? You guys did it earlier. There’s a specific reason for that. But when do you think for other businesses, when should they start thinking about expanding their product line?

Patrick Campbell: As late as humanly possible? No. Just because of our pain. No. I think that our, yeah, it’s a good question.

I think the model there is more around, like if you were seeing. If you were seeing like crazy traction on one product, like I think you need predictable growth for one product before you move on to the next product. And that’s the thing where you have infinite optionality. And so, what’s really, and this is the classic product advice is.

Hey, you need to say no more often than you say yes, right? And so much more like 99 “Nos” to one “Yes”, right? And I think what ends up happening is, we, we can even get a little bit too like, oh yeah, we could build that, we could build that because we have 20% of the entire subscription market using ProfitWell.

And so it’s one of those things where it’s like, well, you have 20% of any market. There’s just infinite possibilities of what you could build to monetize that market, right? And so I think for most folks, you gotta have predictable growth in one area. And then the market’s going to pull you. And I know that feels like fluffy, but the market’s going to pull you into a direction of like, hey, this is what makes sense, or this is not what makes sense.

And I think that that’s kind of the structure of how to think through things. And your measurement of the market pulling you is going to be, it’s going to be different depending on the vertical and the company.

Etienne Garbugli: So, so when you guys start looking into a new opportunity, how do you start thinking about pricing when it comes to the new products? From what I understood, the different products you have, have different value metrics and different, not business models, but different ways of charging specifically. Like how do you guys start thinking about it? How do you guys start thinking about the pricing?

Patrick Campbell: Yeah. So, our directive is, so we basically have this spectrum and this spectrum came about in just thinking through like the analytics space.

So, there’s a spectrum of data, analytics, insights, outcomes, right? So, the only reason you want data is to solve an outcome. The only reason you want analytics is to help you solve an outcome. The only reason you want insights is to help solve an outcome, right? So, everything leads to this outcome. And it could be reducing churn, improving ARPU, it could be a whole host of things, right?

And so, our thesis was, okay, anything that’s just data or analytics, we’re probably going to not charge for, right? Like that’s where we’re going to have freemium. Anything that’s insights or outcomes, we’re going to be driving towards an outcome. The only reason we have those insights is to get towards the outcome, but we’re going to try to charge based on the outcome as close as humanly possible.

And so, what’s kind of held us back from building certain products has been. Oh, there’s no effective way to charge based on an outcome. So, for example, like, we want to come up with this, there are certain retention aspects for retention products that we wanted to build and we’re probably going to build them, but we’re going to include them basically in the Retain products rather than spinning them into something else because it’s like, hey, we can’t charge in a clean way that isn’t tied towards that outcome. But if we put it into Retain, we can still charge based on that outcome. And yeah, maybe we’re not monetizing that feature as effectively as we could, but we’re monetizing the entire product.

And so, our Price Intelligently product, that’s morphing into a pure software product, or a lighter service product, you know, at some point in the next couple of years, because we now have a clean thesis on hey, this is how we could charge based on an outcome, you know, for a pricing product, right? And so that’s what really drives us. And I think that when you have some initial traction and you’re moving, this is how you should think about your pricing, or your products like, Hey, is there a way that I can charge effectively to increase my overall ACV or ARPU.

I think this is the reason why we gave away ProfitWell for free. The metrics product was one, because, metrics products are terribly difficult to monetize. They all go up market. They all start out saying, hey, we’re going to be like the democratization of this type of metric. And then it’s like, nope, we’re going to Fortune 500 because people just don’t appreciate metrics and analytics products as much as they should. And you know, we’ve tried to change that, but it just doesn’t, it doesn’t work. That’s just how people are hard-coded. And so it’s a gateway for us. And the only reason it’s a gateway is because from a unit economic perspective, as well as from a network effect perspective, not only for a network effect of the data feeding our algorithms, but also the network effect of people like referring other people to use the product. That just kind of worked out, right? And so, I think that that’s, that’s been the big thing that we’ve kind of focused on is how do we charge for products that, basically make people more money. And that’s like the central thesis.

Etienne Garbugli: So, each of your products, the products that you’re charging for have a specific outcome that, that people are buying it for.

So, like Retain for example, is churn reduction. And all of the other products. So, the way you kind of segment these products as well, is based on the actual outcomes that’s coming out of the products.

Patrick Campbell: Yeah, that’s the goal. Now, we are not in a perfect place because we have some, some pricing debt, we’ll call it.

So, ironically, our Price Intelligently product, the way that it’s set up. We don’t, we can’t charge purely based on outcome. Now we have a plan, like we have this ROI plan that people can pay for, but what ends up happening is, it gets really complicated when you’re selling to an enterprise company, right?

We’re selling to like giant, Autodesk or something like that. They need a little bit more predictability and they can’t handle the ambiguity of, the type of pricing that we would come up with. Now we’re working on basically. A new product coming down the pipe where we would be able to do that.

Now, that’s going to be some time though. So, but we have this deck where it’s like not a perfect thesis yet, but it’s something that we are, we are basically guiding ourselves towards.

Etienne Garbugli: Okay. Okay. So, once you have these products, how do you figure out what the value metric should be for these? It’s tied to the outcome I assume?

Patrick Campbell: For the new products?

Etienne Garbugli: Yeah. Well for the new products or the products that you add or you created in the past?

Patrick Campbell: Yeah. So that, that, that’s what we’re trying to do. So, all new products, we’re trying to get as close to outcome as possible. Now, some of them might still be in the insights like world, which isn’t necessarily, so that’s where Price Intelligently is, Hey, we’ve done this research. We have these insights, we have this data or these insights that tell you that you should do X, and that’s what’s going to help you improve that, that ARPU or that ACV, right? So, it’s basically like one degree away from that outcome or two I guess one degree to do your, I don’t know how the degree, but it’s like basically right next to that outcome.

So, we’re okay with that for now. But basically, what we’re trying to do is we’re trying to get all of our products into a world of outcomes, which, you know. Might not be possible right now, but it will be possible as we kind of advance in the future.

Etienne Garbugli: Okay. Okay. So, when you’re working on say the pricing model for Retain or another product, how do you go about creating the segmentation for your pricing strategy?

Patrick Campbell: Uh, so for us, like which axis of segmentation are you referring to? I just want to make sure I understand.

Etienne Garbugli: I think that’s part of the question. So, how do you figure out which axes of segmentation are important for a specific product versus another one?

Patrick Campbell: Got it. So, when you’re referring to segmentation, are you referring to like verticals, these types of things or like types of personas?

I just want to make sure I understand cause there’s segmentation. I can take that way. Or packaging.

Etienne Garbugli: That’s also the question. Which dimensions are you’re looking at specifically, and like, how do you. How do you figure out which dimensions matter for that specific product?

Patrick Campbell: Yeah. So, okay, I’m going to answer based on my interpretation of segmentation.

Yeah. Cause segmentation gets tough because most people what they refer to is, which aspect of the customer or the segment of the base that you have. So, for us it’s. When we’re doing research, and we’re putting together our pricing specifically for some, most of our products.

We just go back to those, those ideal customer profiles that we’ve set out and we really focus on the data and the information that’s coming from those target folks. So, for example, on the pricing side, we will weight data or we will wait information for customers that come from those companies that are doing $75 million or more in revenue per year versus those who are doing less than that, right?

Because we’re pushing ourselves to continue to go up market. Now in terms of like, the segmentation of when a deal comes in. We typically price, also along that axis on some level. So, we have a couple of different levers within our Price Intelligently pricing. One is like the size of the company, and that’s very much like pure segmentation-based where it’s like, Hey, if they’re above this threshold, the price is different than if it’s below this threshold.

And that’s basically just because the value that they’re getting, even if it’s the same level of work is very different, right? If I improve ARPU for $100 million company by 30%, and I improve a startup by 30%, that $100 million company is going to see those gains in a much, much quicker way, but also a much more dramatic way. So, that’s kinda how we weigh things there.

What’s kind of beautiful with the Retain product is the segmentation is just built into the value metric. We charge based on recovered revenue. So, you know, if you’re a large company, you have a lot of failed payments and the price is then going to be higher because it’s recovered revenue versus a smaller company where you might have a lot of failed payments, and so the effective percentage we take based on revenue, we don’t charge based on percentage. We charge based on, we have some tiers that basically backs in you based on how much we recover. But either way, like that allows us to, basically get as close to pure value as humanly possible. And yeah, we’re pretty proud of the elegance of that pricing model because it’s, it’s pure ROI.

If they don’t make any money, it’s $0. It doesn’t cost you anything. And if we make you a ton of money, we take, a good amount of money, but it’s nowhere near the amount of money we made you.

Etienne Garbugli: Hmm. So, in that sense, the segmentation that you guys have is based on… So, you have an overarching hierarchy of the market that you guys are looking at and the products, matter to different types of audiences and different types of segments that you guys have or different ideal customer profiles within, in relation to that specifically.

Patrick Campbell: Yeah. So, from that perspective, yeah, exactly. So, we have, depending on the product, we have different segments that we target, and then our pricing model shifts depending on who actually comes in, right?

So, on our inbound side, on, on the lead to demo, to our ops team, you know, they’re not BDRs [ Business Development Representatives ], we don’t have BDRs, but they’re like BDR equivalent. They are only taking certain, or they’re only going after folks outbound who are of a certain size or of a certain caliber, but they’ll take inbound from different types of groups as well.

So, it’s structured based on those segments coming in, but that’s, that’s the nature of a value metric is that it can kind of save you a little bit from having to overthink segmentation.

Etienne Garbugli: Okay. Okay. So, in that sense, that also helps you guys minimize a little bit the diffuseness of having a multi-product strategy where you guys are targeting the same profiles, but approaching them with a set of products that cater to their realities?

Patrick Campbell: Yeah, it gets tough. It just gets tough because like again, there’s a lot of surface area, but yeah, the long and short of it is we are, we are trying to make somewhat of a path of least resistance for folks coming in. And we don’t want to, obviously get it wrong when they come in just because they might not. They may not, you know, we don’t want to confuse the sales process, if that makes sense.

Etienne Garbugli: Yeah. Yeah. I think that’s actually a very good strategy. Like if you, I don’t know if you’ve read Lost and Founder by Rand Fishkin.

Patrick Campbell: Yeah.

Etienne Garbugli: Yeah. So, he talks about how the multi-product strategy that they add actually confused there because they had different personas, different targets, and the products didn’t overlap that much.

I think you guys have a structure that kind of makes it. More consistent, which is really interesting.

Patrick Campbell: Yeah. Yeah. Which works out well. Yeah. It’s interesting. I think it’s super hard, with whenever you add multiple, multiple pieces to a situation, and that’s kind of what we’ve done.

Etienne Garbugli: Well, what do you feel are the, the pros and the cons of having a multi-product strategy?

Patrick Campbell: Cons we’ve talked a lot about, so I’ll summarize them. I mean, it’s, you have a lot more surface area for limited resources, right? And there’s, there’s definitely an argument not to do it because you have, again, if you were targeting, if we were targeting different types of companies AND different types of buyers, I think it would be a very dumb decision and we’ve, maybe it’s still a dumb decision, but we’ve tried to convince ourselves it’s a smart decision because we’re going after the same companies and we have this hub and spoke model where they can come in through the hub, which is ProfitWell Metrics, or they can come in from one of the spokes, which is our different paid products.

Now, what’s super interesting is on the pro side. We get multiple bites at the apple, as they say, or maybe I’m using the metaphor incorrectly where let’s say someone, someone stalls on the Retain side of the house, right? You know, the customer success person, they’re really busy or he or she just isn’t interested for some reason.

We can go at them for the Price Intelligently product, right? And that customer success person may or may not even be involved in that process, right? And so, there’s a little bit of a swarming aspect where when one path fails, another path can be successful over time. I think that the other, the other pro is, is that.

We, and this might be just rationalization of a bad idea, but we don’t have to necessarily be as targeted with some of our marketing strategies because there’s a little bit of a pick your own adventure because people can kind of come in and see what’s going on. But the con there is that our homepage is a little generic, right?

It’s a little bit hard to understand exactly what we do from the hero, right? It’s not just like. We do this, we help with this, right? It’s like we have a suite of products that help with this nebulous thing, right? This outcome-centered products, right? And it’s like, what does that mean? Well, you have to read a little bit more to understand, right?

And so, the goal is we have to be so good at our marketing, so good at our content that people are willing to read that whole page, right? If they can read the whole page, we can get them. If they don’t read the whole page, then it’s really complicated, right? So, I think that there’s, those are some of the pros, some of the cons.

I think it’s, it’s something that we’re constantly, literally every quarter we’re like on some level debating, should we focus all on this thing right now? Should we, should we do this? And we always kind of talk our way out of it. But yeah, it’s something that’s, you know, something that’s super interesting.

Etienne Garbugli: So, how would you assess a product that would be too far off the platform that you guys have created? Like would you say, for example, a product that is for acquisition or something else like that? Is there like something that would tell you that this is, this is too far off, what we currently have.

Patrick Campbell: Yeah. I think that it really comes down to, so it’s, it’s less about the space. So, we believe that we will, we’ll have an acquisition product at some point. And it’s more about. What is it that we can do that no one else can do better, right? Or what makes us unique in doing it? So, one thing that we thought about is like, Oh, let’s build an NPS [ Net Promoter Score ] product.

Cause it’d be really useful to have NPS inside ProfitWell, so you can segment your data by NPS, but then we’re like, eh, like what are we going to do that’s anything different than every other NPS product that’s out there. And we have some ideas of how we could do this in a unique way, because you know, the Price Intelligently software is survey-based. And so there’s some things we’ve learned there, but we’re probably not going to make like a four to five X, let alone a 10 X product in that space, right? So, it’s like, eh, let’s just integrate with people and people can use whatever they’re using right now. And then, yeah, maybe we offer up a version of it because our really basic product there, but yeah, it doesn’t.

It probably doesn’t make sense for us to build that, right? So, it really kinda comes down to that uniqueness factor. Like, what can we do that someone else can’t? Well, right now with Retain, our data set allows us to basically be the best in the world at recovering failed payments because we’re sitting on more failed payment data than anyone else, right?

And data as you know, is the fuel for those algorithms. Now, that is kind of expanding now into active churn as well. We’re not going to be a Gainsight, we’re not gonna be anything like that, but we might be a product compliment to the Gainsights of the world and those types of things.

So, yeah, I think long story short, it really comes down to that thesis of how are we better than, how would we, how can we be better than anyone else who would be building this? And that’s, it’s not always obvious. And so, there’s, there’s definitely a debate that goes into that.

Etienne Garbugli: So, in a way, it’s kind of looking at what your current platform enables. Find a way where you could expand in the most, strategically sound way to be able to address, in a unique way, other outcomes?

Patrick Campbell: Yeah, 100%. I mean, we’re going, so for give you an example. So, we’re going deeper and deeper into revenue operations, which is this new newfangled term, which is basically the combination of sales, marketing, and customer success operations under one roof.

And. What we’re finding is that, you know, we’re, we thought the cleaning of data and getting accurate financial data was going to be, something that, yeah, it’s hard, but everyone was going to be able to do. Well, it turns out, we’ve been in this game now for like six years and we’re still miles ahead of our competition and other people in terms of accuracy.

And so, it’s like, okay, well if that’s the case, maybe this is a lot harder than we thought. And so, as we go deeper into revenue operations, okay, well we can create products that help with revenue operations that are, we’re uniquely qualified to solve, around accurate data and things like that.

And so, that’s kind of how we think about it. I think that there’s a world where you still have to say no to some of those ideas, but right now that’s the guiding light for a lot of the ideas that we have, because a lot of the ideas we have, it’s all of a sudden, we ask that question and then poof, that idea goes away because we’re, eh, there’s nothing that we can do that’s going to solve for this.

Etienne Garbugli: No, it’s definitely a super interesting approach. I like in a way how contrarian it is to a lot of the thinking in startups right now. How would you… Maybe last question, how would you, if you were starting over today, how would you approach figuring out what your opportunity should be?

Patrick Campbell: Say that one more time.

Etienne Garbugli: Like if you were starting all over today. You were able to start anything, any kind of business in the current, in the current space. How would you go about figuring out what your opportunity is?

Patrick Campbell: Yeah. So, from scratch.

No, it’s hard. You asked really good questions cause they’re not obvious answers. I think that. Here. Here’s what I think on approach. I feel as if most companies, they do this a little bit backwards. You know, product-market fit was a really good concept when product was really difficult to build.

Now, creating a great product is still and always will be really difficult because it’s like creating a great piece of art or a great, math theorem or like a great chemistry innovation or something like that. It’s always going to be hard to do great. But I think the barrier to entry is so low right now where all of us can just basically build a product, very quickly, right?

We can spin up a server and all those things that we take for granted of. Being able to just grab a website off of Squarespace or doing X, Y, Z. Like those things were so hard 20 years ago. It took so much effort and so much money to do those things. And I think so now product-market fit made sense because Hey, if you could build this innovation, there weren’t a lot of features and functionality out there, so you could just go find this market, right?

You have a product now I’ve got to go find the market. And I think that really, we’re in a world of market, right? And so like, Balfour, Brian Balfour, good friend of mine who’s, runs Reforge right now, and used to be former growth or leader of growth over at HubSpot. He basically started talking about like, it’s really market-product fit.

What is the market you’re going to go into. And then, gleaning from that market, what makes the most sense to build for that market, right? And I think that what you need to do in terms of approach, if we were starting over, I would, I would think about the market and then there’s a bunch of questions, right?

Well, if the market’s so astronomically big, there’s probably people in it already. So, what makes you different? If the market’s small, but growing, what’s the thing that can make you unique now, but also helps you ride the wave of the market? And just really going deep on what that market is and then kind of thinking through like what you build.

And so. I really like what DC, David Cancel did at Drift where you know, the first year and a half, two years, they built like four or five different products, to chase different markets and it was all in the name of research and then they said, Oh, this is the space we’re in. We’re going into this space. And then they went all in, right?

And so it’s less of, I think right now, I guess to put it in more practical terms, there’s too many people going, Oh cool, I can create this cool thing. And then they go all in on this cool thing they just created. Not enough people who are like, Hey, let me research this cool market, and then based on that, I’m going to go all in on it.

And it’s just because research isn’t as sexy as pushing code. The problem is, is that research helps you understand what code you should push, and enough, not enough people realize that.

Etienne Garbugli: But in that case key, would you define that market?

Patrick Campbell: Which one?

Etienne Garbugli: Well, yeah, maybe Drift is an example. Cause a market could be like, for example, pharmaceuticals or it could be a segment within, within that market. Like how, would you, would you start with a market…

Patrick Campbell: That’s a whole podcast, right? So, I think that. Now, how do you define that market?

I mean, that’s, that’s where you have to go into research mode, right? And again, it’s going to start with an inkling of an idea, right? Or it might start with like, I mean, I know some people who build their companies based on, you know, what are the top 10 largest markets? Or what are the top 10 growing markets, right?

You know, so the, the classic Amazon. Hey, like retail’s interesting, but this internet thing, like it wasn’t an eCommerce play, it was an internet play. He’s like, Oh, the internet is going nuts, there’s going to be a, an eCommerce person or a commerce retail person who wins in that market, right? So, you’re looking at trends, but you’re also looking at, those inklings that you might have and then kind of like sanding down.

And then it comes down to a combination of, all kinds of stuff, right? Like, you. You know, there’s some folks, and I think I’m one of them. I don’t, I can fall in love with any market, right? I can fall in love with every single market that’s out there. I could be making… shower curtains, right? Know all, everything about shower curtains. I could tell you everything and everything about it.

But then I think that there’s, there’s a ton of people who are more, they’re more, I need to love it. I need to like it, right? And that’s a constraint. And that doesn’t mean you can’t find a big market, you just might have to find a, find a little bit of a more unique play on it, if that makes sense.

Etienne Garbugli: That’s really interesting. Thanks for taking the time, Patrick. Where can people go to learn more about your work and your company?

Patrick Campbell: Uh, yeah. So, I’m just profitwell.com is, you know, we write a ton of content and do a bunch of other fun stuff.

And then, I’m just Patrick Campbell on LinkedIn, so find me there. That’s where I publish a lot of fun stuff. But if you ever have any other questions, feel free to email me at patrick@profitwell.com. Sometimes it takes a little bit for me to get back to you, but I always will get back to you.

Etienne Garbugli: Amazing. Thank you very much.

Patrick Campbell: Yeah, absolutely, man.

More on Multi-Product Strategy

[ Interview ] Lean UX Co-Author Jeff Gothelf on How Product Teams Should Do Product Discovery in 2020

A few weeks back, I spoke to Lean UX co-author Jeff Gothelf for The Lean B2B Podcast. We talked about UX, product management, innovation, experiment design, and the importance of doing effective product discovery.

You can watch the full interview below, or access it on iTunes, Google Play or Spotify.

Interview Transcript

Jeff Gothelf – Product Discovery

Etienne Garbugli: My guest today is Jeff Gothelf. Jeff is the co-author of the books Lean UX: Design Great Products with Agile Teams and Sense & Respond. He’s also the author of Lean vs Agile vs Design Thinking. Along with Josh Seiden, Jeff co-founded Sense & Respond Press, a publishing house focused on bringing innovation, digital transformation, product management, and design books to market.

Jeff is also a coach, speaker, and consultant helping organizations build better products. Jeff, welcome to the podcast.

Jeff Gothelf: Thanks so much for having me on, Etienne.

Etienne Garbugli: Maybe as a first question, so when you’re working with product teams, and when you speak with product teams, do you feel that, teams are generally sufficiently user or customer-driven?

Jeff Gothelf: Um, I think that most organizations pay a lot of lip service to being customer centric and, and customer driven. I think a significantly smaller percentage of those organizations actually work that way. That’s not to say that the product managers, the designers, the engineers, the QA folks on those teams are not customer centric or don’t want to be.

I just have seen that many organizations say they are or say they aspire to be, but don’t actually. The majority of them don’t put the effort into actually be that way, despite, despite having really smart customer-centric designers, product managers and engineers on staff.

Etienne Garbugli: Well. So, in that case, what do you feel are the biggest challenges, or the biggest hurdles that these teams or these people within organizations face in attempting to be more user or customer-driven?

Jeff Gothelf: I mean there, there are lots, not the least of which is there’s, especially in successful companies, even high-growth companies, they don’t have to be old companies, but companies that have found product-market fit and that are scaling quickly, there’s a belief that we know what’s best for the customer, right?

Look, we’re big. We’re successful. We’ve been around a hundred years or 20 years, or you know, where we, we’re a hundred-million-dollar company. We’re a $500 million company. We know what customers want. I don’t need to talk to them. I don’t need to ask them. Um, because look, look at what we’ve done on our own without that so far.

So, there’s, there’s a lot of that. Um, the other thing is that being customer centric means that your measure of success changes significantly. It changes from outputs to outcomes. Now that’s an easy thing to say. So, I’m going to unpack it for just a second. Outputs are the things that we make, the features, the products, the services, the apps, the devices, those types of things.

Um, most organizations manage to output for a variety of legacy reasons, but really, most of all, because it is a binary measure. You’ve shipped the product or you did not ship the product. And because it’s binary, it’s easy to measure and because it’s easy to measure, it’s easy to manage and easy to incentivize and reward.

If you’re truly being customer centric, you should be managing to outcomes. Outcomes are the customer behaviors that we see once we give our customers the product, the service, the app, the system, whatever it is. And if their behavior doesn’t change in a way that makes them more successful and then ultimately us more successful, then we have to update that system.

And so, the measure of success changes dramatically from a feature-centric one. We built a thing. To a customer-centric one when you manage to outcomes, which is we positively impacted the lives of our customers. And we know that because now they’re doing things differently that benefit them more and that benefit us more.

Uh, and so that’s, that’s, that’s really the biggest, the biggest obstacle to being customer centric.

Etienne Garbugli: And so, working with a lot of organizations or interacting with a lot of organizations, like how do you see that trend evolving?

Jeff Gothelf: Again, I see a lot of really smart people in the trenches, in the individual contributor roles, in the team manager roles, in the discipline leadership roles, the design leader, the product leader, that type of thing.

But I rarely see that conversation elevated to a leadership or an executive level outside of customer satisfaction or Net Promoter Score. Which again is, uh, not, uh, we should absolutely be worried about the satisfaction of our customers, but they’re using that one metric as a sole measure of that is incredibly risky and not terribly valuable.

Um, and so that’s the trend is. I think you’d be hard pressed, so I think you’d be hard pressed to find a designer who doesn’t believe they are customer centric. I mean, sure, there’s going to be some, some genius designers out there who believe they know best, but for the most part, I think the overwhelming majority of interaction designers, UX designers, those kinds of folks are customer centric by default as part of the profession.

I think product managers, modern product managers, people who subscribe to kind of the modern process of software development and delivery and understanding the market (product discovery). I think many of those folks believe they are and are customer centric. So, you’ll be hard pressed to, to find somebody who says, no, I’m not, I’m not customer centric.

There are, they exist. There are some sort of inward-facing product owners or product managers who don’t really believe in customer centricity. I think there’s probably more of those than there are UX designers. Um, and so I think the trends there are positive at the individual contributor level.

I think even engineers really get it. I mean, there’s still, again, I think, you know, there’s probably more engineers than product managers, than designers who don’t really care about the customer that much. But nevertheless, I think modern software developers really do understand that customer centricity is key.

So, I think those trends are headed in the right direction. I think at the C-suite and the leadership level, there is a belief in maintaining competitiveness, maintaining competitive advantage. And they see organizations like Apple, Netflix, Facebook, Google, Amazon, really sort of dominating not only their own spaces, but expanding fairly broadly into other domains. And they want to be like them. And unfortunately, the only real takeaways they get from those organizations is faster time to market, right?

How can we get more stuff to market more quickly? And I think where they fail to see, and this is not true across the board for those big five tech organizations, but what they fail to take away is how customer centric Amazon is. Netflix is.

How focused on the user experience they are. Um, sadly, Facebook is too to their own ends. They’re not really focused on optimizing the customer’s experience with more of their bottom line. But, um, and I think that that is, uh, that’s something that’s missing.

So, they’ll focus on things like agile, for example. Right? Agile is what makes these companies successful. We need agile. And they see it as a, as a recipe for delivering more software more quickly, which again, is managing the outputs, not managing the outcomes.

Etienne Garbugli: All right. So, going in that direction, so say I just got hired by an organization, how would you recommend I go about diagnosing the product organization?

So, how we work and what’s the current situation in terms of how things get delivered? Are we output, are we outcome? Is it, is it working in terms of agile? Is it working as an organization? And then maybe how would you recommend someone within an organization to shift that mindset towards more of a learning, more of an, uh, an output discussion?

Jeff Gothelf: I think how you have that conversation will vary by where in the organization you got hired, right? So what level in the organization. But the conversation you’re trying to have, generally speaking, is the same. It’s a conversation that asks the question, why? Why are we working on this? Right?

What customer is it for? Um, what benefit does it bring to the customer? If the customer is successful, how does that translate into business success for us? How does that fit into the broader product portfolio that we’re building into the strategic plans for the organization for the next year or two years, or whatever it is?

That’s the conversation you want to have. Now, again, asking why is increasingly risky to your career, the lower in the organization you are typically speaking in larger organizations, right? And so, we want to build the conversation to ask why in a way that doesn’t limit your career in that organization.

But that’s the conversation you’re trying to have. So, I think that, you know, at the individual contributor level, it’s about asking, what are the assumptions that went into making the decision to build X, right? Or to design it in this particular way? Who’s the target audience? Right?

When’s the last time we talked to anybody from that target audience? Can I go talk to somebody from that target audience? Right? And then to, as you start to collect more customer feedback, more quantitative data, and as you bring that back into the team. Really seeing how the team, the product manager, the business lead, the business line leader, the executives, the stakeholders, react to that data because that data is never going to be 100% reflective in a positive way of everything that you’re doing, right?

There’s always going to be some contradiction. Hey, we’re making it blue. Everyone says they like red and, whatever. Right? So, how people react to that becomes really interesting, right? It’s a good indicator of the culture of an organization if they’re like, oh yeah, yeah, that’s what the customers say, but we know better. Right?

Or that kind of builds up a better understanding of our, of our customer. Let’s explore and see why that is, and maybe through the next iterations, we can make some improvements, right? It’s two very different conversations.

Etienne Garbugli: So, I saw, I think last week or a couple of weeks back, you released the hypothesis canvas, which I really like.

So, when you’re working on a project, when you’re working as part of a team, so how do you pick which questions to address and how do you figure out what the most important things to learn are?

Jeff Gothelf: Yeah. So, um. The Hypothesis Prioritization Canvas is a tool that I use with every team that I work with to help them determine where to focus their customer discovery work.

There is sadly, not enough time in every sprint to do all the discovery work that we want, and frankly, we don’t need to be doing product discovery on everything. Right there, there. And that’s the purpose of the canvas. The canvas is to say, here are all the things that we’re looking at for the next iteration, two iterations, quarter, whatever it is, whatever the timeframe is.

Where should we focus our learning activities? Our discovery work, our research, our, uh, customer conversations? And, the matrix in that canvas is based on risk of the idea, risk of the hypothesis, low to high, and the perceived value or the perceived impact that we believe that this hypothesis will have on the customer, on the business, etc., again, from low to high.

It’s perceived value because we’re assuming that this will have big impact, right? We, uh, our educated guest tells us that, but we don’t really know. And then risk is contextual to the hypothesis or the idea itself. So, it might be technical risk, it might be design risk, might be brand risk or market risk. The hypotheses that end up being high risk, but also high potential value or high perceived value.

Those are the ones where we should spend our product discovery work on. That’s where we should, if we, if we’ve got, you know, those precious few hours, every sprint that we can build experiments, talk to customers, do research, you know, prototyping, whatever, whatever the activity is. Um, focus on the hypotheses that are high perceived value and high risk.

Because if you get those wrong, you stand to do a lot of damage. But if you get those right. You stand to have a good impact on the customer and on the organization. So that’s, that’s what the canvas is for, is to really help you kind of think through your ideas and where to focus your product discovery efforts.

As far as determining, uh, risky assumptions and that type of thing. The really interesting thing is there’s two key questions that I teach every team that I work with. Um, and the questions are boxes seven and eight in the Lean UX canvas, the first canvas that we did a couple of years ago.

What’s the most important thing that we need to learn first and what’s the least amount of work we need to do to learn that? Now, the focus of your question was box seven, which is what’s the most important thing we need to learn first? And that’s the question that I ask every team at the beginning of a new iteration, right?

What is, you know, uh, the, what is, it’s a conversation about risk and it’s a conversation about the risks that will derail the thing that we’re currently working on. Now, the interesting thing is, is that in a truly cross-functional setting, product managers, designers, engineers, sitting together, you will get at least as many answers to that question as there are disciplines in the room, right?

The designers will talk about some designer challenges, some customer challenges. The product managers will talk about, you know, product-market fit or scalability or business model or business rules or whatever it is, right? The engineers will talk about feasibility, scalability, performance, security, and all of those are valid concerns.
They’re all valid risks to the success of the project. The question is, what do you need to learn right now? What’s the most important thing you need to learn right now? The thing that if we get it wrong today, tomorrow, in this sprint breaks the whole initiative or breaks the whole hypothesis. And that is an interesting conversation to have with your team.

At the beginning of an initiative that’s going to be related to value, the value of the idea. Will people look for it? Will they find it? Will they try it? Will they sign up for it? That type of thing. Do they understand what it is? Right? Um, if you’re in kind of a more mature version of the product, we’re going to move away from value into feasibility, scalability, security, performance. Business viability, those types of things.

Um, and so really thinking about the, where in the lifecycle of your initiative, uh, where in the lifecycle of your idea is helps you to understand the biggest risks, but always at the earlier stages, focus on value. Because if nobody wants it, nobody will find it. Nobody will look for it. Nobody will try it. Uh, it doesn’t matter if you can build it.

Etienne Garbugli: But so, if we talk about that specifically, like one of the things I love about the Sense & Response framework is that agility, but not focused on how you ship, it’s more of that corporate agility where you know when to adapt and change course and, and change your direction.

So, in that case, when do you know, like what are the. So, for example, the trigger points that tell you that it’s no longer about the value. Now we should be learning about that next other thing afterwards. Maybe feasibility or whatever it is. Like how do you know when to transition between say one, a learning goal versus another learning goal?

Jeff Gothelf: It’s a great question. Look, I mean, the interesting thing about all this stuff is that we are inspired by science. We’re inspired by scientific method. You hear us talk about assumptions and hypotheses and experiments, right? Collecting data and all that stuff, and we do all of that stuff and we are inspired by scientific method, but we’re not doing science. Right.

In science, there are more or less absolute truths, right? We fed the bacteria X and they mutated into a monster. That happened or it didn’t happen, right? Uh, in our world, the answers are rarely black or white. They’re usually in, in some shade of gray in between. And so. How you decide when to kind of move on to the next set of assumptions, the next set of risks to say, Hey, you know what that was good data, so we’re going to continue, we’re going to persevere with this hypothesis versus pivoting or killing the idea.

There is, there’s a level of evidence that you want to use, but at some point, you’re going to need to augment that with some level of gut feeling or really confidence, right?

It’s confidence that that is, is the true test. So, my, my friend and colleague Jeff Patton talks about this concept of confidence in terms of bets, right? So, you, you ran an experiment, you learn something and you say, okay, great. Um, what are you willing to bet me. That that this is a good idea that we should keep working on, right?

Are you willing to bet me your lunch? Right? Most teams will say, yeah, are you willing to bet me a week’s vacation? Right? Maybe are you willing to bet me your retirement savings, right? Uh, your house? Are you willing to bet me your left arm? Like, you know, and so and so if you’re, when you, you know, when you collect the data, you know, qualitative and quantitative and you still can’t get to a, you know, it’s going to be in that gray area and you still can’t get to that really kind of definitive, yes, we should move forward. No, we should not. Um, that’s a good conversation to have.

What are we willing to bet that this is still a good idea? And that really gives you a sense, if people are saying, look, I’m not willing to bet you lunch that we should keep moving. That’s a good indication that. We got to change course, right?

If, and if people are saying, I’m willing to bet you my left arm, that this is a good idea. And that’s kind of the collective sentiment, that’s a good indication that the team’s got enough data to move forward. And that’s, that’s, that’s the non-scientific part of this, right? That’s the shades of gray part of this.

Etienne Garbugli: Well, so in that direction, so in one of your talk, you’re talking about, uh, say for example, the team is focused on improving the conversion rate. So, you hit 15%. Maybe you can get to 20%, but the effort between 15% and 20%, will that’d be worth your time?

Obviously, if you were able to always make the right decisions and always be, be changing, uh, to the next assumption at the right time, there would be a lot of value there cause you would probably be growing faster. Are there telltale signs that tell you, so we’ve, we’ve probably milked that objective enough for now, let’s try and, and, and address the next thing?

Jeff Gothelf: Yeah. It’s interesting, right? So, there’s theory of the local maxima is always an interesting one, right? We keep putting, we keep putting effort into this, and, um, we we’re not seeing the kind of returns on it. Um, again, this is, uh. There’s no, I mean, as far as I know, there’s no formula that says that you have now tried squeezing more out of this five times, and that’s enough, right?

There might be a formula. I just, I’m not aware of one. Um. And so, it’s really a thing of like, Hey, we ran this experiment and we couldn’t get this to move forward. Hey, we tried this week. We really got up, you know, we’ve got a 1% lift out of this. We did this, we got a 0.2% lift out of this.

Okay, we’ve been at this for a month, we can’t seem to budge off of a 15% conversion rate, 16% conversion rate. That’s a good conversation to come back to your stakeholders, your clients, whoever it is, and say, look, we spent a month trying to move this. We got to 15 pretty quickly. Getting beyond 15 has proven a challenge.

There are some big things we could try and kind of pick up and move over, but we don’t believe that we, you know, the, the, the return is worth the investment here at this point. So, I think that that, um, again, is one of those sort of gut feelings. Um, but it’s tough. It’s tough because remember, we all love our ideas and letting them go is difficult.

Etienne Garbugli: All right. So, within this like, so what’s the role of experimentation and what’s the role of maybe more in depth research on product when you’re doing product discovery? So how do you see them best working together?

Jeff Gothelf: Well, I mean this, this is how we learn, right? Experimentation and product discovery is how we learn.

Now, experiments, for me, it’s an umbrella term for anything that we do to build learning into our product. Discovering delivery process. Right? So, it can be anything from a customer interview to a paper prototype, to a live data prototype, to beta testing, price testing, A/B tests. All these things are, are experiments.

They’re learning activities to help us determine if we should continue investing in a particular direction. Um, and so to me, they are a crucial part of “the work”. In other words, these are not disposable elements of product development, modern product development that we can not do when we’re tight on time. Right?

The, the work, sadly, it’s primarily perceived the software engineering in most organizations, but in reality, “the work”, capital T capital W, right, is, uh, is software engineering, design, right? Product discovery, research, product management. Uh, uh, you know, all of these things are part of “the work”.

And I think that that’s, to me, if, if I feel successful when an organization understands that. They understand that the work can’t be done without, uh, uh, without all of those components. And so that’s to me is it’s a crucial role. It’s as crucial of a role as software engineering because, and people say, well, that’s not true because the software engineering is the actual making of the thing.

And while I agree with that, right, if you make the wrong thing or an unusable thing, then who cares? Right? And so, informing what work needs to get done is equally as important as, as expertly crafting the experience, both in a design and a development way.

Etienne Garbugli: So, in that case, if you, uh, if you use experiments as an umbrella term like that, so do you always recommend setting exit criteria, for example, you’re doing interviews or you’re doing other product discovery activities that are more, um, customer development or user research. Would you always say that there needs to be an exit criteria to these activities as well?

Jeff Gothelf: Define what you mean exit criteria.

Etienne Garbugli: More like you have an experiment, you’re running a test, so like you would have an evaluation criteria. If this passes, if we reach this certain threshold, would you do the same thing with interviews as well or with other activities like that?

Jeff Gothelf: Yeah. I mean, yes, you need to go in with a sense of what success looks like and what failure looks like. Right? So, um, there needs to be some kind of a threshold with the team that says, we’re going to talk to 10 people today. If less than three of them tell us that there’s any value in this thing, then we’ve got to go refigure, we’ve got to figure this out.

Now look, you may end up with like two and a half, right? Which is again, where this kind of like two people said they hated it, and one person said, you know, I don’t love it. I’m not sure I’d use like, you know, you don’t get that, that perfectly clear answer.

Um, and that’s where the challenges begin with all of this is, is because if you don’t get the, those very clear responses, you don’t get that very clear, you know, black or white, yes, it’s a pass or it’s a fail, but you absolutely need to go in with some kind of success threshold. That, as a team, kind of back to our conversation about bets and confidence gives you enough confidence to move forward. Right?

So, for example, let’s say you’re doing a feature fake. Right where there’s a link, kind of a button to nowhere, right? So, it looks like a feature on your site, but when you click it, you get nothing, or a 404 page or a coming soon message, right? Um, and 10,000 people come through that workflow on a daily basis, right?

As a team, how many people are going to have to click on that to give you the confidence to move forward? Right? And there’s, you know, you start off with a ridiculous. You can get teams to, to conclude, you can say, look, two people out of 10,000 and they’d be like, no way. Right? 9,000 out of 10,000 no way.

Okay, so it’s between two and 9,000 right? You kind of start to narrow that down. Is it a thousand is it 500 is it 2,500 right? That becomes the kind of conversation that you want to have and be very clear that if we hit this, we feel confident enough to move forward. That’s it.

Etienne Garbugli: So, it’s about increasing the level of confidence you have to move forward to the next step?

Jeff Gothelf: Yeah. Because I told you like, this is not, it’s not science. I mean, you will, every now and again you’ll get lucky, right. And nine out of 10 people will find the thing, tell you they love it, pay you for it. Right. Awesome. Right? But, uh. Well, I’ll give you a real life example. I was working with a team that was building a new subscription-based business for an organization that never had subscription-based businesses before.

So, they didn’t have, they didn’t even have a sense of what success looks like for subscription, but they knew how much money they needed to make and how much they’re going to charge for this service, what they thought they were going to charge for the service, and so they knew how much money they needed to make in order to build this into a viable business.

And so, they did some quick, you know, scaling or some spreadsheets for scaling this up and scaling this down. And they got a sense that about 75% retention rate month over month was what they needed, right? And so they ran this experiment with this new subscription model. A a. Idea, and as long as people converted at or around 75% continuously on a month to month to month basis for three months, that was a good indication that they were delivering value.

So, for them, that number was 75% and it may not be the number for your organization, but that’s what they needed to hit to be able to make a business case for further investment in this idea.

Etienne Garbugli: That’s super interesting. I don’t want to take too much of your time, but thanks for taking the time, Jeff.

It’s really appreciated. Where can people go to learn more about your work?

Jeff Gothelf: Absolutely. It’s my pleasure to be here. Um, so, uh, so I’m super easy to find. That’s by design. If you go to jeffgothelf.com, you’ll find everything there. My blog, What I do, how I do it, links to upcoming events.

I’ve got lots of in-person and online classes available as well as links to Sense & Respond Press. And so that’s a great place to go. Jeffgothelf.com, start there.

Etienne Garbugli: I’ll make sure to share the hypothesis canvas, and as well the revised Lean UX Canvas that you guys have.

Jeff Gothelf: Perfect.

Etienne Garbugli: Thank you very much. That’s really appreciated.

Jeff Gothelf: Thank you, Etienne. That was great.

More on Product Discovery

[ Interview ] Bestselling Author Nir Eyal on Using the Hook Model to Improve Product Engagement

A few weeks back, I spoke to bestselling author Nir Eyal for The Lean B2B Podcast. We talked about product design, retention, growth, and how the Hook Model can help improve product engagement.

You can watch the full interview below, or access it on iTunes, Google Play or Spotify.

Interview Transcript

Nir Eyal – Hook Model

Etienne Garbugli: So, my guest today is Nir Eyal, Nir is the bestselling author of Hooked: How to Build Habit-Forming Products, and Indistractable: How to Control your Attention and Choose Your Life, which was published earlier this fall.

Nir’s work focuses on the intersection between psychology, technology, and business. Nir has taught at the Stanford Graduate School of Business, and prior to teaching and writing bestselling books, Nir was the CEO and cofounder of two businesses Ad Nectar and Sunshine Business Development, both of which were acquired here.

Nir, welcome to the podcast.

Nir Eyal: Thank you. Great to be here.

Etienne Garbugli: Great. So maybe a first question, so maybe at a meta level, why did you decide to explore or move forward the thinking around behavioral design specifically?

Nir Eyal: Yeah. So the idea behind behavioral design is that we can use consumer psychology and product design to help people live better lives and help them form good habits.

And so that’s really what Hooked was about. Hooked, what I wanted to do with Hooked was to steal the secrets of companies like Facebook and Instagram and WhatsApp and Slack, and all these companies that are so good at changing user behavior so that we can take these tactics and not just leave them to the big social media companies and the gaming companies, but that we can use the same exact techniques to help people use all of our products.

You know, anyone out there who’s building a product or service. The idea here is that we can use these same exact tactics to help people form good habits with our products and services. So, you know, whether that’s enterprise software, whether that’s a financial services software, whether it’s, consumer web or fitness software.

The idea is that we can use the very same tactics that these companies use to make their products so engaging and habit-forming. We can use those same tactics for good.

Etienne Garbugli: And as you were doing your research, when you were writing Hooked, did you realize that some of these organizations were already using Hooked-like frameworks within their, their organizations?

Nir Eyal: Sure. I mean, that’s where I learned this stuff. You know, these companies were around much before my book came out, and my book took those tactics that I learned. You know, I looked for the commonalities and that’s where the Hook Model came from. This four-step framework that I talk about in my book is directly pulled from the examples of what we see is common to all of these very habit-forming products.

So, it’s the same four steps, whether you’re looking at YouTube or Facebook or, Slack, uh, you know, all sorts of products and services that are designed this way, have these four basic elements of the Hook Model embedded in them.

Etienne Garbugli: Okay, great. So, within a company’s lifecycle, let’s say from company formation to growth, and maturity, when do you think is the right time to start really thinking about habit-forming designs and functionalities?

Nir Eyal: As soon as possible. So, there’s, there’s two places that my work tends to get used. The first is in the very early ideation paper napkin sketch timeframe when the idea is just germinating and we haven’t committed any code, we haven’t even done any wireframes.

The best time to use the Hook Model is to ask yourself, first and foremost, does your business model even need a habit? Not every product has to be habit-forming, but every product that needs to be habit-forming has to have a hook. And so, if your business model requires people to come back on their own, requires unprompted user engagement, well then you have to understand and design those four steps of the Hook Model.

So, that’s the best place to use this stuff is in the early stages. The other place that I find that a lot of people utilize the Hook Model is, in the later stages when something’s not working. Uh, so many times I’m called the plumber, you know, they call the plumber to stop the leaks. And so many times a company will call me or a venture capitalist will call me and say, we’ve just sunk millions of dollars in this company.

They had amazing growth rates. It looked like everything was going in the right direction, and they acquired all these customers, but then they leaked out, right? We call this a leaky-bucket business; businesses that have high growth, and low retention. And of course, the problem here is that you’re wasting tons of money buying growth, but you can’t buy engagement.

You can’t buy consumer habits. You have to design that into your product. Growth you can always buy, you can just buy a bunch of ads on Google or Facebook or television commercials. Doesn’t really matter. You can buy growth, you cannot buy consumer engagement that has to be designed into the product.

Etienne Garbugli: Okay. So, in that case, the trigger would that they’re realizing that there’s a lack of retention based on their expectations and they’re pouring money into the organization without getting the, the engagement that they’re looking for.

Nir Eyal: That’s right.

Etienne Garbugli: So, how would recommend that a product team or an entrepreneur evaluates how habit-forming their product currently is? So, in that situation, how can they get a clear picture of how habit-forming the product currently is?

Nir Eyal: So, that’s where the Hook Model can be a very good diagnostic tool that if you find that consumers are not sticking around, that they’re not being retained in the way that you’d hoped.

Then that’s when you can take out the Hook Model and figure out what part of your user experience is broken. Does it conform to the four basic steps of the Hook Model that we see in all sorts of habit-forming products? That becomes a central question. Now in terms of a metric that you’d want to look at.

You, you, you want to look at the, what the metric I like best. You know, there’s been a lot of debate whether we want to look at MAU over DAU or, you know, there’s all kinds of metrics out there. My favorite metric is to first define what a habituated user looks like, and that’s going to be contextually-specific to your product.

So, if you’re building a social network. A habituated user is probably someone who checks in every day. If you’re building a enterprise software, same thing, right? It’s probably something that needs to be used every day.

Certainly, needs to be used at least once a week. That’s kind of the critical cutoff for a habit-forming product.

And the product isn’t used within a week’s time or less. Very difficult to change a consumer habit. So, we really want to make sure that we define what would a habituated user look like, if you’re building a fitness app, okay, well that’s gotta be, you know, two or three times a week. If you’re building, you know, a marketing automation software or something that maybe it’s a little less frequent.

It depends on what you think should be the regular cadence of someone using that software. So, you define that number and then you want to ask yourself what percentage of your user base is using that product with that level of frequency. And typically, what we see is that 5% or less is a problem.

If 5% or less of your, of your user base is not habituated and your business model requires them to be habituated for your product to succeed, then you probably need to go back to the drawing board in the early stages. Once we start seeing about 5, 10% of the user base, people who come in through the top of the funnel actually stick around and become habituated. Now we can start innovating. We can start, uh, you know, uh, playing with the different parts of the Hook Model to make them better.

But it’s a good chance that we have some fundamental elements in place already.

Etienne Garbugli: So, if you’re, you’re looking at the data and you see that it’s either 5 or less, does that tell you something more about whether the expectations are maybe wrong, or that the business just does not have product-market fit, for example?

Nir Eyal: Probably doesn’t have product-market fit. So probably, you know, if there’s either, either your Hook Model is horrible, you know, in most cases, when I, when I work with companies. You know, even though the book Hooked sold 250,000 copies, a lot of people still haven’t heard about it. And so, you know, they just think, Oh, well, if I just design a product that is awesome, that’s good enough, right?

If I just give people what they want, then they’ll love it. And of course, that almost never happens. We have to be a bit more deliberate around how we design these products for habit and make sure we have all four elements. So, when I work with a company that is trying to build a product, and yet users aren’t sticking around, right?

They’re leaking out, what I will do is come in with the Hook Model and ask them, okay, what part of the Hooked model is deficient here? Is there not a clear trigger? Have we identified the wrong internal trigger? Is the action easy enough to do? Is there a clear reward, and does it scratch the user’s itch?

And then finally, and probably the most overlooked of the four steps of the Hook Model is are we asking for a user investment that improves the product with use? And so, we can use it as a diagnostic tool. And of course, every single time we do, we always find at least one, if not more, of those four steps of the Hook Model that’s deficient in some way.

Etienne Garbugli: And do you have like an idea of like based on the engagement that you do, of which elements tend to be more problematic when you come in?

Nir Eyal: It tends to be, it tends to be either the internal trigger, meaning they haven’t figured out what the customer pain point really is that occurs with sufficient frequency to form a habit.

Or typically it’s the investment phase is, is, you know, you can get problems in any of the four steps. I mean, I’ve, you know, it’s, it’s, it’s. A pretty even distribution, but I think if, if there’s, if there’s two areas that I tend to see more problems, it’s that either the company hasn’t identified the right internal trigger, the reward isn’t rewarding enough, or they’re not asking for a proper investment.

And many many companies don’t, you know, they forget about this investment phase. They don’t figure out a way to make the product better and better with use. But of course, if you look at every of those, everyone of those products I talked about earlier.

They all have some kind of investment that you put into the product in a form of data, content, followers, reputation, something that improves the product with use. Most companies out there today don’t do that. And it’s a huge, huge missed opportunity.

Etienne Garbugli: So, say in that direction, like how would you recommend an entrepreneur or a product team figures out what the right triggers should be for the product?

Nir Eyal: Yeah. So, what you want to do is to look for nascent behaviors. So, you know, many times I’ll talk to entrepreneurs, uh, and they’ll, there’ll be bright-eyed and bushy-tailed, and they’ll tell me, you know, ask them the question, how does the user currently solve this problem? And they’ll tell me, no, no, no Nir, you don’t understand.

This problem has never been solved before. There’s nothing like what we have for the market. It’s never been tackled. And they think that’s a good thing. And of course, I’m very skeptical because if you can’t tell me how the customer currently solves the problem, it probably means it isn’t a problem.

And so that’s the, that’s the first step is to ask yourself, how is the customer or the user currently solving this problem? And so why does it need a new solution? And I, don’t mind if the solution is terrible. That’s actually a very good sign. So, if they’re, you know, using Google Docs and stitching together data sources from a bunch of different places and they had to, you know, use Scotch Tape and Bubblegum to figure out their own solution to this problem.

That’s great. That’s a wonderful sign. But many times, when I ask companies, you know, how is the person currently solving this problem? They say, well, we don’t know, but they should, and that’s usually a bad sign. So, that’s how we start looking for these internal triggers. Another technique that came out of the Toyota production system is to ask ourselves the five why’s, which is we keep asking why five times until we get to a feeling. Until we get to some kind of emotion that the user is looking to escape, not just the functional needs, but the deep-down emotional needs that they are trying to satiate. Because only when we ask why five times do we open the aperture of a potential solution. So, if you just stop with the functional needs, you know, why does our company, a customer need this?

Oh, because they need to send messages from places, but whatever. It might be like a very, you know, typically teams will, will come up with very functional type needs. You know, this data needs to go over here or whatever it might be. But the aperture is very narrow in terms of what the possible solutions might be when we so narrowly define the problem of why the user needs this.

But when we really back it up and start with an emotion, with an internal trigger, fear, loneliness, uncertainty, fatigue, anxiety. When we back it up all the way back to the emotion. Then the option set for how we can scratch that itch, how we can design for that discomfort opens up and we have much more flexibility and much more opportunity to fix the problem in a creative new way.

Etienne Garbugli: But that does mean that there might be multiple answers to the same questions. How do you go about iterating on the effectiveness of the habit-forming loop, like how do you figure out that this is the one we’re trying, then this one might work better?

Nir Eyal: Yeah. So, then it becomes an issue of segmentation. So, let’s take the internal trigger of boredom. Okay. If we really get down to it, one of the reasons why we use many products and services is because we are trying to alleviate boredom, we do not like that sensation. That is an uncomfortable internal trigger that drives us to check the news, look at stock prices, sport scores, YouTube, Reddit, Pinterest, all of these things.

Fundamentally, let’s be honest, cater to this uncomfortable, emotional itch, maybe even email for many people. So, even though the internal trigger is the same, how the user scratch that itch might change. So, when you feel bored, maybe you go watch sports on ESPN, but when I feel bored, I read the news and that scratches my itch differently then your itch, even though the itch is the same, it’s boredom.

And so that’s where it comes down to customer segmentation, right? But fundamentally, we have to understand the user’s itch so that we can design the solution and the reward phase, the third step of the hook, the variable reward phase, to make sure that it satiates that particular consumer’s problem.

Etienne Garbugli: So, in, in that end, the, the trigger would need to be more precise for the segment that we’re thinking about specifically?

Nir Eyal: Well, the the, the internal trigger is the same, but the reward changes. So, the purpose of the variable reward phase is to scratch the user’s itch, but leave them wanting more.

So, if you said, Hey, we have a product that’s going to satiate your boredom. I mean, you wouldn’t say that out loud, but that’s the subtext of, of how you design the product, right? You wouldn’t want to put that on a on a banner ad. But if we have a product that’s all about, you know, come watch sports on our website.

You know, for one person that’s really not going to scratch their itch, but to another person that’s wonderful, that’s incredibly entertaining. So, even though the internal trigger is the same, the reward is different. The way it is satiated might change based on a particular person’s demographic, life experience, interests, all those other things.

Etienne Garbugli: So, based on that, how would you make sure that you’re actually making progress as you work with that organization or work on your product?

Nir Eyal: Yeah. So, remember we talked about that, the percentage of users of habituated users? So, once we have that North Star metric. Then once we have that North Star metric of what percentage of our user base is habituated, now we change the product and observe what happens per cohort.

So, we make some changes to the product. We improve the Hook Model in some ways. We, you know, work on the external triggers, the internal triggers, the action phase, the variable reward phase, and the investment phase. We tweak the product based on the Hook Model, and then we observe what happens to the cohort of users who are presented with the new, the improved version of our product based on the changes we made.

And then we’re tracking just one number. What happened to that percentage of habituated users? Used to be, we were at 5%. Oh, my goodness. Look with this cohort now we’re at 10% of our user base is habituated. They’re using the product as much as we would expect a habituated user to use it.

Etienne Garbugli: And gradually you do have a benchmark where you feel like this would represent a product that is really habit-forming?

Nir Eyal: Well, the gold standard, what they saw at Facebook was about 50%. Yeah. But that, we haven’t seen that in a long, long time. It’s very, very rare that you’ll get a product that will be that habit-forming. And of course, it depends on how many users start using your product. And of course, the quality of the users.

You know, if you just spend a lot of money on advertising and people come to your product, but they’re not the right type of users, you might get them to sign up, but they weren’t really great to begin with. But you know, assuming that you’re actually fishing where your consumer, lives right, if you’re actually finding the right type of user, then, then, you know, anything above 5% is a good sign.

Etienne Garbugli: Okay. So, you work with and invested in a variety of organizations. How do you typically see the Hooked framework working out in complex B2B or enterprise settings?

Nir Eyal: Yeah. So, as long as the product is used habitually, you know, and again, not every product needs to be used habitually. So, for example, let’s say you have some, you know, a software that is only used if something terrible happens.

Let’s say you have some server-side software that, brings all kinds of bells and alarms when something terrible happens.

Yeah. So, that’s not going to be a habit-forming product. That’s something that you don’t need to use. It just runs in the background. So, you need to focus on, on selling it once and then you’re kind of done right until something terrible.

So, you don’t have to create a habit with that kind of product. But you know, increasingly that is the minority of products these days that if you look at the revolution in bring your own device (BYOD) and you know, enterprise SaaS software, it is critical that the product is actually used because if people don’t use your product, they, they, they cancel. Right? They churn out.

Yeah. And so, it becomes absolutely critical that we get people to use the product in order for them to keep paying us. Now, that’s a very different business model than what used to be the case, right? Pre-Salesforce days when Salesforce said “No software”, it used to be, you know, you buy per seat and then you can use the software as much as you want, but we got paid no matter whether you used it or not.

That business model is becoming increasingly rare in the enterprise these days because today, you know, people want to pay per month, and so if the product isn’t used, eventually the procurement department says, what the hell are we paying all this money for, nobody’s actually using this?

And then of course they, they churn out. So, it’s very important, increasingly important today that whether it’s enterprise and of course in the consumer web space, that your product does create a habit or you’re very likely to find that your customers will stop using your product and stop paying for it.

Etienne Garbugli: But if your software, for example, Salesforce is used by different roles, different types of users within the organization, how do you, what do you think are the best ways to kind of figure out, would there be multiple, would you use multiple versions of the Hooked framework or multiple instances of the different hooks that you’d be creating in place?

Nir Eyal: Sure. Yeah. If the product does different things for different people, then absolutely. You can modify the different types of hooks per that particular use case. I would argue in the case of Salesforce, Slack, Github, Stack Overflow, all of these things are enterprise products, and, and they all utilize the Hooked model.

Etienne Garbugli: That’s interesting. Hmm. So, from your experience with, with teams and businesses, like what are the criteria that drive the successful implementation of the Hooked framework? Like in what circumstances is it most successful?

Nir Eyal: It’s, it’s really about frequency. That’s the number one criteria for a product that can succeed or will not succeed with the product, is how often should people use it.

You, you can’t, you can’t make people use a product that doesn’t benefit them. And so, if the, you know, it’s a nice aspiration if you say, Oh, I wish people would use the product more. But if it doesn’t make sense, right? And if it doesn’t, if there isn’t a context with which people should use the product and that it benefits them to use it more often, then, then you, you know, you can’t change that, you fundamentally have to create value for the user.
And so, the, the best type of products are the ones that are used with sufficient frequency to form a habit. Those are the right candidates for the Hook Model. Again, if it’s, if it’s a product that’s very rarely used, you don’t need a Hook Model.

You just need to sell it once and then you can walk away. But with a product that needs to be used habitually for the business model to succeed, that’s the kind of product that, that does require habits.

Etienne Garbugli: So, there needs to be a certain, a certain potential for usage frequency, but as well we mentioned before, it’s better if you’re beyond product-market fit

Nir Eyal: You don’t have to necessarily, I mean, if you use it in the very early days, you’re just using it differently. So, the Hook Model can be used in the very, very early days to help you design the experience itself, right? So, if you’re still in a pencil sketch stage, then it behooves you to, before you spend a lot of money on, on the user experience, or of course committing code, spend some time asking yourself what the Hook Model should be so that you could design that into the product from day one. Because again, you can’t buy engagement. It has to be built in.

But if the product is already out there, and maybe it’s even reached some level of a product-market fit, but you, but you find that the consumer engagement rates are too low and the percentage of habituated users isn’t where you want it to be.

Well now you can use the Hook Model as a diagnostic tool to figure out what’s wrong with your user experience to hopefully improve it.

Etienne Garbugli: It’s really interesting having those, those threshold as well. You’re talking about 5% if you’re under that, it gives you a sense of where you should be focusing as opposed to when you’re beyond that. That’s really interesting.

Nir Eyal: Yeah, yeah. And that’s, by the way that I have to say as a disclaimer, because I, I want, I want to make sure this is clear. This is just in my personal experience. I haven’t seen any studies that show 5% is the magic number. This is just in my, you know, decade or so of experience in the industry. What I tend to see, but again, there’s never been a state that says, Oh my gosh, you know, 5% is the number.

Etienne Garbugli: Have you seen a relationship between the Jobs-to-be-Done framework and how people use the Hooked framework?

Nir Eyal: Um, in terms of, I think what I like about the Jobs-to-be-Done framework is that it does focus on the base needs, right?

There’s the, there’s the always that a story of, you know, people don’t want a hammer. Uh, they, they want to hang their picture on the wall. And so how can you help them hang their picture on the wall? I would even take it a step deeper of, well, why do they even need that picture on the wall? Right?

What emotional need are they satisfying? That would even open the aperture even more. Right? Um, so I think there is, there is some similarities between the Jobs-to-be-Done framework in terms of getting down to base needs. Um, but I think what I add to the conversation is, okay, now that we know the internal trigger, now that we know the base needs, we know the Job-to-be-Done.

How do we put this into practice? How do we actually make sure that our product has these four fundamental steps of the trigger, the action, the reward, and finally the investment?

Etienne Garbugli: That’s really interesting. Yeah, that’s what I was thinking about before when you were mentioning how the same level of clarity can be used in both different instances, so thanks for taking the time Nir, that’s all the questions I had for you today.

Where can people go to learn more about your work?

Nir Eyal: Yeah. Thank you. My website is at nirandfar.com Nir spelt like my first name N-I-R, so that’s NIR and far.com and the book is called Hooked: How to Build Habit-Forming Products. And my second book is called Indistractable: How to Control your Attention and Choose Your Life.

And that book is more of a consumer-facing book, not so much a product design book. And it’s about how to make sure that we live with personal integrity and do what it is we say we are going to do.

Etienne Garbugli: And congrats again on the new launch, and thanks for taking the time to chat today. That’s really, really appreciated.

Nir Eyal: My pleasure. Thank you so much.

Etienne Garbugli: Thank you.

More on Product Engagement + The Hook Model