Highlights:
- [00:06:03] The benefits of ideal customer profiles (ICPs) and who uses them
- [00:08:15] Andrew discusses the details of how they developed ICPs at Hotjar
- [00:11:34] Anti-ICP: Ruling out a large portion of your audience
- [00:24:08] Andrew details how to discover criteria that reveal insight
- [00:25:40] Tip: Don’t overcomplicate ICP, three or four defining characteristics are enough
- [00:29:19] Anti-ICP: How to communicate with customers that you don’t want
- [00:32:16] An effective ICP contributes to retention, alignment, and customer satisfaction
- [00:35:44] Andrew's final piece of advice: ICPs “needs to be a company-wide initiative”
Sources mentioned in the episode:
About our guest
Andrew Michael is the CEO at Avrio – a software development research platform – host of Churn.FM, and Founding Member of Startup Cyprus. Andrew is also a teacher and speaker featured at the How to Web Conference 2022 and Advancing Research 2022. Before joining Avrio, Andrew worked at Hotjar, where he held leadership roles in marketing, management, and experience design.
Transcript
Andrew Michael - 00:00:00: It's always important to be keeping like a pulse on who the ideal customer profile is because it can really shift quite rapidly as well when different market conditions maybe present themselves or new technology emerges. Like there's many, many different reasons that can evolve, be targeting at any given time.
Erin May - 00:00:21: This is Erin May.
John-Henry Forster 00:00:22: I'm John-Henry Forster. And this is Awkward Silences.
Erin - 00:00:35: Silences. Hello everybody and welcome back to Awkward Silences. Today we're here with Andrew Michael who is the CEO of Avrio. Today we're going to be talking about how you can use research to identify your best customers, your ideal customers, which is something I'm very excited to talk about because we are doing just that right now as we like to redo from time to time. So selfishly. It's going to be a fun one for me too. So thanks for joining us Andrew.
Andrew - 00:01:04: Thanks for having me. It's great to be on.
Erin - 00:01:06: We got JH here too.
JH - 00:01:08: Yeah. When we talk ideal Customer Profile ICP, I can't help but think Insane Clown Posse so just focus back on the topic at hand.
Andrew - 00:01:17: Yeah, it's one of those acronyms as well. I have a slide at some point as well. It's like just this like hearty character peeing on the wall. It's like ICP as well. In marketing we use way too many acronyms.
JH - 00:01:31: An important one too though.
Erin - 00:01:33: Yeah. Just to reiterate, you were talking about ideal customer profiles and none of those other ICPs. Yeah. Awesome. So we just discussed what is an ideal customer profile? We know what the acronym spells but like really what are we talking about when we talk about an ICP?
Andrew - 00:01:49: Yeah, so I mean, I think from my perspective it really depends on the stage of growth and where you're at. And I think like this is maybe often looked in the beginning, but ultimately I think in the early stage startup, it's really like your ideal customer profile. Who's ever going to give you money at that point in time? You literally just need to survive and you need to make things work. But then I think as the company grows and as the product evolves, your understanding and your definitions should be evolving constantly. And like you say, you revisiting the work now and you visit the research. I think this is just something that you should be doing all the time. And also another thing I think on this is that when we think about our ideal customer profile, if we start looking at our customers who we have today, they're a direct reflection of the product we've built and the marketing we've done up until now. But they don't always necessarily represent the best ideal customer for the business or for where the market is moving. So always thinking about these things and thinking there's a constant evolution that you build on top of with your understanding over time and it's an evolving research process.
Erin - 00:02:47: I think in the beginning your best customers are the people who will pay you. I imagine that doesn't ever really change. Right, but it changes how you want to think about those opportunities. Yeah?
Andrew - 00:03:00: Absolutely, I think I mean, in the beginning, you accept pretty much any customer. And I think from my perspective, on CHURN.FM we talk about churn in the context a lot. And I think in the early days, when you just accept anybody and everybody as a customer and you just try and acquire a new one and everyone, what ultimately ends up happening is you end up seeing that impact on show and retention, ultimately, which ends up like draining your company's resources, like support. Ends up spending time serving customers who aren't the ideal fit. But at the early days, you need that for growth, and you need that to sustain the business. But then over time, you can really start to realize, okay, the time and effort that we're spending serving these customers doesn't make sense for us, and we're better off getting a better grasp of who we should be serving better and then focusing on that audience. So, there are times when it makes sense to fire customers.
JH - 00:03:45: Yeah, that phrase always is a good one. How do you think about it as an early team and your kind of going through this transition of, we just got to find customers and then we have some critical mass, and now we can start to be a little bit more intentional about which of these customers are the best fit. Where does that happen? Is there like, a certain stage or scale where that tends to come in or it varies? Or how do you know when you should start thinking about this? I guess this may be the question I'm trying to get at.
Andrew - 00:04:07: I think you should always be thinking about it. So, I'd use that as an example. I think before we started the business, we already had an idea of who the ideal customer profile was, and that started as a result of early customer research that we were doing. So really just trying to dig into the pain point into the problem we're solving. We started to formulate which group and which audience resonated best, understood the problem most, had the biggest pain. So, we started building the product with this ideal customer profile in mind. But then, as you said, as you start to grow and as you start to acquire new customers, you start to notice new segments and new audiences that you perhaps didn't consider. And we're busy going through this evolution now as well, is starting to look at some of these other adjacent use cases, adjacent audiences, and trying to see, okay, like, does this make sense for us to be serving this audience now? And it's something that I think you're always doing as part of your business, but really it starts before you even get started building a product. t
JH - 00:05:02: That makes sense, yeah. So, it's not that you don't do it early on it's that you don't have enough signal and data and probably until later on to be more confident in it. But it's something you're always thinking about refining and having hypotheses around.
Andrew - 00:05:11: Yeah. And I think this is something that changed my perspective from our time at Hotjar, actually, is that when I joined Hotjar came from a previous startup that I founded. We were VC backed, so we had this heavy reliance on, like, we needed to share data and numbers and statistics when it came to how the business was operating, running. And when I joined Hotjar actually, like I found that for the business that at scale it was heavily, heavily reliant on qualitative research and just speaking to customers and almost nothing on data alone. And I think the reason for this is that at early stage the data that you have is more like noise than signal. And only when you start to hit a certain size and scale can data become really reliable and your best source of information, it always is your customers. But even more so at the early stage you really need to be speaking to your customers day in and day out and never lose that pulse.
Erin - 00:06:03: So, I think we'll go back and forth between sort of the what and the how and the why here. Right, so what is an ICP and why does it matter and how do you use it? So maybe we can jump into that. How do you use it for a minute because one of the things that we talk about a lot is how a lot of artifacts you get out of research such as an ICP ideally can be used by a lot of people in the company. A lot of different departments may be in different ways, but curious how you've seen an ICP actually be used by different departments and to the point about sort of different stages, your ICP is going to look a different way that might play out different ways in different stages as well.
Andrew - 00:06:41: Yeah, so I think what's interesting is like most companies go through this evolution where a lot of teams have what they think is the understanding of the ideal customer profile and they'll go to marketing and they'll have one sort of a definition, they'll go to customer success or sales and product. But at some point, in time, like someone in the team says hey, we need to get a grasp to this. We need to have a common understanding of who we are going after so that when we are using what we deem to be ideal customer profile we have like consistency across the board. And by that the way you can use it is pretty much every team in the company can use the ideal customer profile in some form or manner. If you take marketing for example, like understanding who you're selling to really then starts to think about the communication, the ad network channels that you're going to go through, how you're going to distribute content, thinking about the product and positioning. That's how marketing can really use the ideal customer profile with those insights in mind, and then it just flows down the chain. So, sales, instead of spending the time trying to sell anything and everything, you give them really a good fit profile. So depending on if you have the luxury with the high inbound amount of leads, like you can filter them using your ICP or when you're doing your outbound. So it's pretty much like going throughout the organization, you can find different use cases for the team to be able to use it effectively. And then what it ends up happening is that at the end result, you get a higher ration, higher proportion of the ICP that you acquire versus non-ICP. And overall, what you have to see over time is the significance that it has on the bottom line and the revenue for the business.
JH - 00:08:15: You mentioned a little bit like on inbound stuff. If you can identify who is an ICP, you can route them differently and do important to kind of filter them. Does that mean that when you're defining an ICP, you need to do it off like characteristics that are easy to identify or publicly available? Or should you also be thinking about more proprietary information or data you're going to have on users as being a factor too?
Andrew - 00:08:35: So, I think it's always going to be a combination of both. And so, I can tell you a little bit about how we did it at Hotjar in the early days and then how that understanding evolved over time. So, when we first started looking into identifying ideal customer profiles, the very first thing that we just did was we took a look at our own internal data and we said, let's take a cohort of users that had been with us or customers that had been with us for longer than twelve months paying and had a certain level of activity. So, the activity component was their own proprietary data. We were looking at how active were they within the product and how regularly were they using us. And then we just use a service like Clearbit or any one of these data enrichment services that took a look at, okay, who is our overall demographic, and then started breaking that down by various demographic and demographic properties. So, company size, age, Alexa rank, role of the individual. And from there you could really start to see some patterns emerging from different segments and different properties, individual properties themselves. And what we tried to do in the beginning, the first step of this was we just said, okay, like just random things now, but they say like mid markets, product managers, product designers, and one or two different use cases. This was the profile. And what we slowly started to realize over time was it served us well for the beginning, like, that's the data we had at hand, we needed some formal definition that we could go to and we could draw to. But as we grew, our understanding and our learnings as well grew with that. And we slowly started to realize it was just a really narrow view of looking at only our existing customers. Because again, like as I mentioned in the beginning, there was a direct result of the early product that we had built and the market yet. And it also has survivorship bias in it of the existing user base. So, we then slowly started to layer on different levels of complexity. We also at some point did a lot of pricing and packaging research, which a big part of that is understanding who the buyer personas are. And through that, we actually started doing a panel study. So, this was like the next layer of evolution was like not only looking at our own audience but introducing like Hotjar to a new panel who didn't know our products or service. And then tried to understand, okay, what were those thermographicor demographic properties that showed interest that we had the highest likelihood to buy or the most willingness to pay. And we then could say, okay, did this audience align with our previous understanding? Or there may be some audiences that we haven't attracted well enough yet through our product or through our marketing that we could be going after because there's a big opportunity in the willingness to pay or the likelihood to buy. And then it just like every stage we started layering on a new understanding of our customers and new learnings. Not to go on a long run along. But I think what it ended up getting to in the end, what we realized was that having a fixated definition of like, this is who they are, they're like product managers. This role, that- is not necessarily the best way to go about using the ICP because you end up like negating and ruling out a large portion of the audience that could be a good fit. So, what we ended up doing was looking at building together a scoring model to understand to what degree do they fit the ideal customer profile. And through that research, what we ended up doing was, let's say we found that product managers perhaps were much more engaged. They stuck around for longer, they would get a three points in the scoring model, whereas maybe, let's say digital marketer, some in the medium range, they would get a two. And then other roles like customer success would get a zero or one. And we did this for each one of the thermographic properties that we had and each one of the things. And we based it not only off of base of the panel, we based it off retention, we based it off recently acquired customers. So what ratio made up the recently acquired and we also based it off of conversion rates. So, we took a cohort that was six months old and we looked okay out of everyone that signed up there who ended up converting and were there any further thermographic or demographic properties that stood out that converted better than most? And again, that could have been fed back to marketing and say, hey, there was only a small amount of these properties that we got through in this cohort, but they had really strong signals and they were really strong conversion rates at the end. And then we took all of that data and ultimately what we ended up having was the scoring model. And it was quite beautiful in them because you could see, like, as the score increased, so did net MRR retention and so did overall revenue that we're generating from these, which is ultimately the end goal that we're trying to achieve.
Erin - 00:13:03: There's a lot to dig into there, which is great. I have so many questions for you, but one of them was if you think of the ICP as a binary or not, and it sounds like no, that the scoring allows you to think of it on a spectrum of, let's say zero to much more than zero. So, I'm curious, how do you use that score? Right, so presumably you would want to market to and find people with higher scores, give them more of a sort of velvet rope experience once you've identified them. But how do you think about breaking up how you address these sorts of different tiers of score or different ranges of score?
Andrew - 00:13:42: So even though you still look at it as a range, you still do form groups within those ranges. But it's not to say there's somebody doesn't fit all five criteria, that they're not an ideal customer profile. So, I think that's sort of one of the mistakes we made in the beginning was just like, okay, they had to fit these five criteria. If they didn't like, if they weren't more than five years old, they're not a good business fit for us, but they might have been like extremely fast growing start up and they would have made an extra customer, but because they didn't have one of the five or one of the ten or whatever it was. So one is that you still definitely want to have this understanding of who the team thinks is the ideal customer profile, like who they are as a general as a whole, but then having them as a range in the spectrum then allows you to sort of qualify leads a little bit better. So, one of the first ways we really started using it was in our lead qualification scores for sales team. So, the scoring model then allowed them to say, okay, if the score was below five, we're not going to try book demos with these companies. If it was between five to ten, we would. And then if it was like ten plus. We really want to be pushing to try and book demos with these companies. So, this really allowed them then to have parity on to what level of degree do we want to be going into driving sales? The same thing again on marketing as well is the score itself. You could end up seeing, okay, like which campaigns were driving, higher acquisition of certain levels of scores and then you could then dive into and see, okay, like, what was driving that score at the end of it as well. So, I think there's many, many ways that you can filter them that down as well into support and to other areas of the business.
JH - 00:15:17: Cool. And on the scoring, itself, you kind of give the example of maybe this job is how to get to three. This job is it gets a two and zero one and stuff. How much of coming up with the waiting and the score scale and stuff for all the different criteria is like an art versus science thing. Is it something where you just kind of have to get started and be like, we kind of have a gut feel that job title is pretty important, or you're doing something more like a stricter analysis to come up with the scales and the weighting of the different factors.
Andrew - 00:15:41: It's a combination of both. So it's strict analysis in terms of what we would do is let's say we're looking at retention now and we're going to say, like, who has the highest retention? We would just take a property and see, OK, how does company size impact retention? And we do like the solo people, like one man companies, like two to 10, 50, 50+, and then we see what is retention look like for those individual segments. Then we take the same thing and we would look at conversion rates. So like, we take the cohort and we see out of that cohort the size of one person company or two to ten things, how many of those converted? Look at recently acquired customers. So we would take these different and then the panel study as well. And then from there you get to start to see a pattern and a trend like which ones end up emerging better than others across all four vectors. And then from there it's like giving a scoring and the scoring, then it's based off of the data. But you need to somehow use your intuition in the gut feel sometimes because there may be a little bit of conflict between what you see maybe on retention and what you see on the acquisition front and then those dials and those levers you play with. But you can ultimately see, like, if you've done things right at the end is when you map out your entire user base against the scoring model and then you map it against verticals, like revenue and retention. And you ideally want to see this mass looking slope that increases as score increases.
JH - 00:17:03: All right, a quick awkward interruption here. It's fun to talk about user research, but you know what's really fun is doing user research, and we want to help you with that.
Erin - 00:17:11: We want to help you so much that we have created a special place, it's called userinterviews.com/awkward for you to get your first three participants free.
JH - 00:17:23: We all know we should be talking to users more, so we went ahead and removed as many barriers as possible. It's going to be easy; it's going to be quick. You're going to love it. So get out there and check it out.
Erin - 00:17:31: And then when you're done with that, go on over to your favorite podcasting app and leave us a review, please.
How often you talked about updating your ICP and your scoring sort of regularly or constantly. What is your recommendation in terms of how to do that? You pick up the pace as you scale or?
Andrew - 00:17:52: Yeah, so I think this exercise, we probably revisit it every year at Hotjar, and I think it depends on the stage and growth. I think for our stage and early stage now, we're constantly evolving our understanding of the market. We’re religiously reviewing it and we're updating the understanding of it. I think as you start to hit scale, you have the luxury of data. Things get a little bit more predictable in the system. That's when you can slowly like it doesn't need to be as frequent as we are now, but at least, like, I would say yearly to twice a year. It needs to be something that needs to be revisited, that needs to be edited, but it should rather be seen as more of a living organism that evolves as your understanding evolves over time and having a central way for teams to be able to contribute new learnings to it. So, sales is on customer conversations day in and day out, like they're learning new things. The market is evolving, the climate is changing. So, it's always important to be keeping like, a pulse on who ideal customer profiles are, because it can really shift quite rapidly as well when different market conditions maybe present themselves or new technology emerges. Like, there's many, many different reasons that can evolve who you’re targeting at any given time.
JH - 00:19:07: You had called out the survivorship bias of, like, being something to be mindful of, and that was kind of where my mind went from an early part of this conversation. I'm curious how you think about that. So, you have the situation of these customers are doing great, retaining them really highly, and stuff these people are churning out. So, you start to fixate on the idea of people who are maybe the highest retention and stuff. But there are probably things you could do or consider prioritizing investing in to change that churn behavior of this different cohort and like, who's owning that part of it? Is it like product needs to identify those opportunities and make the case? And then if you see the behavior change, it comes into the ICP, or what's the order of events, and how do you make sure that you don't get too narrow-minded?
Andrew - 00:19:45: I think that's actually where the panel comes in really helpful is that you're presenting your product or service to a totally new audience, and then you're trying to understand, okay, is there an audience that we're perhaps not serving today that we could be doing a better job of? And then, like, if you do see the opportunity, then, yes, it's definitely the role of user research, the role of product to really dig into those different use cases, to dig into those audiences and try to understand, is there a bigger or better opportunity for us out there? And another area I think as well is also good to look at. Even just taking a look at things like Google Trends, I think we did this as well as Hotjar at some point and really try to understand, okay, where was the market heading? Like, how did we see things evolving over time? And were we serving the right segment for the way that the market was evolving at that point? So I think there's different factors that you can look to when it comes to trying to understand, okay, are we serving the right audience? And is there perhaps maybe a big opportunity that we've just neglected or maybe our marketing hasn't spoken to, our product hasn't been built for yet?
Erin - 00:20:47: Yeah, this obviously it gets into business strategy, right, because anyone at any stage, no matter how big or small, is dealing with finite resources. Right? And part of what an ICP allows you to do is to focus. Right? We can't serve everyone. We can't serve the entire market of every human being on Earth. So, we need to focus either a lot or a little bit within that. And, yeah, I'm curious how you use some of this datathe ICP to make a call between, say, you know, we're not at 100% saturation in our identified market or ICP. We could get more there, right, but we see this emerging other audience that we think we could serve better in making those sorts of decisions of going deeper versus expanding. Is there research here that you do that can help you make those calls, or is that just a business decision?
Andrew - 00:21:39: I think those are incredibly hard decisions to make, and they always feel like very big, risky gambles when it comes to it like, do you double down on something that's working, or do you try and go after something new? And I think you can build conviction through research and through speaking to customers, like, continuously, like, getting a pulse and feel what's in the market. But I think ultimately that's just, like, it comes down to business decision that needs to be made. It's like, do we want to double down, or do we want to continue playing in the space that we're in today?
Erin - 00:22:11: Yeah, and then once you've made that decision, you can use this sort of work to then really identify those audiences and the hard work of understanding what they care about and how they buy and all those sorts of things.
Andrew - 00:22:21: Exactly.
JH - 00:22:24: I guess actually it may I ask you a different way is how much do you think about the total adjustable market as part of ICP? So, we have these customers, they're great. We retain them at a higher rate. They pay us a ton, but there are not that many of them. Like, should they be our ICP? Is that part of it or is that a separate exercise?
Andrew - 00:22:39: No, I think that's definitely a part of it as well. You want to understand how big the opportunity is, because if there aren't many of your ideal customer roles, then it's not ideal. So, you ultimately want to be ensuring that as well. That's when you define your ideal customer profile, that the segment and the audience are big enough and there is a big enough market for you to go after for it to make sense. Because you really need to pick a focus and you say, this is whom we are going to be serving. This is whom we are building. And if you're going to invest all that time and company's resources and energy, you need to be sure that there is room to grow within that space. And it's not going to be something where like six months or a year or two years down the line, you're scratching your heads and say, okay, what now? And on that as well. I think at some point, like Hotjar we did next. We had defined our ICP, but we also did for Focus to go in further. And something that we got from Reforge, they've got some really great courses, and it's the idea of this anti-persona. So, like, who are you not serving and who is a bad fit? So just as much as you want to understand who the ideal is, it also really helps the team gravitate and set an understanding of whom they're going after, when they know who they shouldn't be going after as well, and who they shouldn't be spending time in the churn. Because ultimately, if you have both sides, you understand whom we are serving and who we shouldn't be serving. Like everything else you do, like the way you handle support or the way you produce content or the different channels, everything just becomes so much more clearer and you're able to then double down on what's really going to work.
Erin - 00:24:08:
Awesome. We talked a little bit about this before, but you're saying your kind of grouping together the different scores and the different levels. So, I'm curious what you've seen this look like in companies you've worked with in terms of, do you have a tier one? Tier two, tier three? And are they sort of groups of, you know, person-level personas combined with former graphic companies? Or like, what's the artifact you might end up with here in terms of this is our ICP or constellation of ICPs and anti-ICPs, and then we're going to build out some playbooks some ways that we're going to work with these different groups once we've identified them.
Andrew - 00:24:47: Yeah, so I think the first thing is you would end up getting to like a generic ICP. So, you're going to say it's mid-market, sourced businesses between 200 to 500 employees, and maybe a certain specific vertical industry that you're going after. So, I think you still get to that end result, and then you can say, this is our main ideal customer profile, and you might have maybe like one or two others that run in general to that that fits in with it. So, you still do the exercise of, like, the grouping and you say, this is whom we are going after, but you're mindful that you're not like, fixated, that you need to hit every one of those criteria in order to be qualified and to meet the needs of the thing. So, it just starts like that to begin with, and then you use the scoring as a way to see to what degree do they fit and do they meet the needs on that end.
JH - 00:25:40:
One thing we like to poke on usually with this stuff is what do you see when teams do this poorly? Like, they go out and they try to do ICP and they end up with something that's not very impactful or useful. Is it that they make the wrong inferences off the data, they overcomplicate the scoring, they don't come up with an artifact or a way of referring to it internally. What are the things where this can go wrong and not be useful despite good intentions?
Andrew - 00:26:01: I definitely think overcomplicating is one of those ways that you can, like, analysis paralysis, where you spend way too many,v much time, you have way too much data at hand that you don't know what to do with it afterwards. And I think the best one’s projects I've been involved in as well, working with other companies without buying ideal customer, is really like when you can get to like, three or four characteristics that clearly define, like, what's going to make a good fit. And then from there, those three or four tend to be really actionable as well, so you can get to the end of it. One of the big indicators we found early on at Hotjar was that the single biggest indicator if a customer is going to become successful with us or not was their Alexa rank. And it seems very obvious. It's like if a company has a really low Alexa rank, they have a lot of traffic and they made a good fit. But it was such a simple metric that allowed for optimization to begin with. And in the early days, we used different services, and we tried different tools out there that analyzed all the different parameters and then tried to put together like an overly complicated scoring model that ultimately didn't produce much results. But that was one of the very first instrumentations of the scoring model was we just looked at what was the alexa rank, and that's something that's very powerful because it's very easy for the team to understand. It's very easy for everybody to actually action and to use when it comes to analyzing who they are, where they're coming from, with something that you could easily go in and take a look at it. What are the top 1 million Alexa sites look like? What are the industries, what are they serving? So, it really adds a lot of value from that perspective. I think the thing I've seen is when we try to overcomplicate things, that's when it's hard for the team to understand, what does all this mean and how do we use this? But the simpler you can get these things; I think the better because they need to be adopted by everyone in the company. And the worst scenario you can be in is when you've overcomplicated things and then people just go on with their own routes anyway and they end up having their own individual definitions and classification systems, which ultimately gets people fighting against each other instead of working together.
JH - 00:28:12: Yeah, it's sort of the Gall’s law phenomenon, right? I don't know if you are familiar with that one, but a complex system that works is invariably found to evolve from a simple system that worked. And then a complex system designed from scratch never works and cannot be patched up to work is like the maxim. I think that's very true for a lot of things. That feels like kind of what you're getting at is you need to have something that's working and gets traction and then you're going to revisit regularly so it will get more complex, but don't jump in the deep end.
Andrew - 00:28:36:
I've been extremely guilty of like overcomplicating things to start and then like learning the hard way, just like the simplest solution always, and then iterates from there. What did you say was the law again?
JH - 00:28:48: Gall's law, John Gall. It's like a system, but it's a good one. It's true for a lot of product work, but I think this thing also lends itself too.
Erin - 00:28:55: Well because I think the problems always seem very complicated, right? It's like, how are we going to double our revenue this year? So, it's like, clearly, we need a complicated solution, but that's not the case.
Andrew - 00:29:06: Not the case.
JH - 00:29:07: I think sometimes that you want, making a complex makes it feel more like valid almost, right? It's almost like there's like some theater to it. Look at how much work and effort went into this. It's got to be good. Right. And so there's that trap too.
Andrew - 00:29:18: Exactly.
Erin - 00:29:19: So you talked about this a little bit, but I think the idea of the anti-ICP is really interesting. So what do you do with those people? Just put a big X on your website when they try to create an account or be really mean to them, ignore them? Like, what do you do with folks that kind of fall outside of the folks you want to focus on?
Andrew - 00:29:41: Yes. I don't think you ever want to be mean.
Erin - 00:29:43: Not good business practice.
Andrew - 00:29:47: Yeah. Alienate people coming to you. But I think it just helps give clarity on how you serve these customers. And so, like, an example that came, for example, let's say at Hotjar, we had one of the cases in the early days, the anti-persona was engineers. This evolved over time and changed, but it wasn't going to be a tool that they could use for debugging, like if their app had broken. So what that meant was that when we thought about features and we weren't going to prioritize any feature requests we got, that meant how do we debug or when we produced support documentation, this was not going to be a use case that we wanted to let people know how they could do it. So it just means that further deprioritizes those use cases and it gives clarity to the team. Like, okay, well, I get this in supports requests a lot. Like, what should I do with this? And then the easy answer is this is not a use case that we currently support at Hotjar, and if it does change in the future, we can let you know. So there's ways that you can still come to the table with trying to be supportive, but ultimately just being transparent with the end user and say, okay, this is not what the product has been built for. If you figure out a way, let us know. We'd love to hear any suggestions, but it's not something we have immediately on the roadmap or something that we're going to be supporting.
Erin - 00:31:07: And that's a great point because people might very well stumble upon, especially with product lead growth or just easy-to-use products. And they stumble upon and like, great, this product is going to be great for me. And they start having a really bad experience and for someone to come and just tell them it's like, look, sorry we didn't build this product for you. It's a really good service to that user. Maybe they can go find something better for their use case or like you said, maybe eventually they will be the use case, but not doing anyone any favors by keeping bad-fit customers around, trying to bend it into their will.
Andrew - 00:31:40: Right, exactly. And ultimately what ends up happening is it does it backfires, is that people in the word of mouth goes. And this product of mouth, like I used it and it doesn't do anything, but it wasn't built for you to begin with, but we just didn't explain that to you clearly or we didn't communicate that effectively. And as you said in product lead businesses where anybody can just get in, they can try the switching costs or the ability to kick the tires is very easy. Being upfront with that and the way you communicate really helps avoid that negative word of mouth that you really don't want to be getting.
Erin - 00:32:16: What are some of the best things you've seen happen in companies who have done a good job with building, and innovating, and iterating on their ICPs.
Andrew - 00:32:25: Retention goes up. I think this is like a big thing as well. Through CHURN.FM, I've interviewed over 180 guests now from some of the fastest growing SaaS businesses and definitely a key thing that stands out in all of them is they have a really good understanding of who the ideal customer profile is. There are strong alignments in the organization, everybody's building for their use cases, working towards it. So, I think the ultimate end goal is that you have happy customers that end up sticking around with you for a long time and don't leave because you get to understand the problems deeply. Like you can empathize with them entirely and then the whole organization really has this good in-depth knowledge that they can serve them better.
JH - 00:33:07: You called out a little bit. One method that can be useful is having a panel of like, prospective users, and non-users getting that signal from them through conversations. Are there other research methodologies that prove useful when you're trying to identify your ICP?
Andrew - 00:33:21: Yeah, I think the panel is typically done like with surveys from that perspective, so it's more like on the quantitative side, but you're still trying to collect qualitative insights. There are different other mechanisms, obviously always use interviews, like number one, I think on my list as well for gathering like really unique insights. And then there are different channels I think that you can explore as well. So, there's really great select channels and slack groups out there where you can gather insights and you can gather knowledge. I think you can also like when you're looking to try and understand which segments aren't being served well, or you could be doing a better job of even sometimes like leaning on looking at competitors and seeing, okay, what are some of the requests that are coming in through their product roadmaps? And the thing, what are some of the complaints that you're perhaps seeing on sites like G2 Crowd, and is there perhaps an opportunity where your product or service is doing better or an audience there that's being neglected? So, there are quite a lot of different channels that you can take a look to. But I would say, like, nothing beats like speaking to customers and really trying to speak to as many people even outside of your customer base as well, but that match the profile.
Erin - 00:34:32: We know how you can find people like that.
Andrew - 00:34:34: Yeah, I have an idea as well.
Erin - 00:34:39: CHURN.FM. It's your podcast.
Andrew - 00:34:42: Yes, that's it.
Erin - 00:34:43: I've definitely listened to it. It's good. Why churn and not retention?
Andrew - 00:34:47: Yeah, I think just playing with the pain a bit more. So, when I decided to start the podcast, the idea was I wanted to start a business, actually, but I was still figuring out what I wanted to do. So, I thought, let me pick a problem that I know if you're asking. You see what they would pay to solve the problem. They would say pretty much everything in the bank because they know it's going to pay back in compounding interest. So double down on the pain. I think that was the reason that makes sense.
Erin - 00:35:12: I find it very interesting. We hear that internally too. You hear a lot more talk of churn than retention and obviously, they're opposite sides of the same thing.
Andrew - 00:35:20: Exactly. And the best way to solve churn is actually to focus on retention because it's like the first-place people say we need to solve churn. They're like, okay, let's go interview people that are exiting and asking why they churn. That's not really where you want to be focusing. What is making those that are successful with your product? Like, that's where you need to focus your time and energy.
Erin - 00:35:43: Totally. Awesome. Well, any parting advice for folks listening on their ICP work?
Andrew - 00:35:50: I would just say that it's a never-ending process and it's something that the earlier you can get your teams aligned and they're working together to define it and understand that it's not just like a qualitative piece of work that needs to be done or quantitative. Like you need to have the what with the why. It's something we talked a lot about at Hotjar. So, like, that is going to tell you what's happening in your business, and then speaking to customers is going to tell you why. So, the exercise, it's important to have sales and success come to the party with this because they're the ones speaking. They are on the front lines all the time. You need to have data; you need a product. So, an exercise, if you're thinking about doing it, it really needs to be like a company-wide initiative and there can be teams like user research leading in itself. But having this as a company alignment and bringing things together is like one of the most powerful things, I think you can do in your business.
Erin - 00:36:40: Fantastic. That tracks, thanks so much for joining us. It's been a lot of fun.
JH - 00:36:43: Cool. Thank you.
Andrew - 00:36:44: Thank you very much for having me. It's been great chatting.
Erin - 00:36:49: Thanks for listening to Awkward Silences brought to you by User Interviews.
JH - 00:36:54: Theme music by Fragile Gang.
