Strategy & Planning

BSMS 26 - How to define and create your Ideal Customer Profile (ICP)


Today we get into the nitty gritty of Ideal Customer Profiles (ICPs). In episode 16 we talked about how to verify your ICP, but didn’t address the basics. After viewer feedback, today we revisit the topic to answer important questions about ICP. We discuss what it is, why you need it, how to come up with it, what it should look like, and how to use it.

Some specific conversations within the episode include how to use your existing CRM data to better outline your ICP, using segmentation models based on market and product maturity, and mistakes we often see marketers make with their ICPs.

If you enjoy the episode, check out Stijn's new book. Coming soon. 

Episode Transcript:


Welcome to episode 26 of B2B SaaS Marketing Snacks. I'm back. Thank you very much to Richard, my friend, my man, for holding down the fort the last couple of weeks. I appreciate you. Today, we are going to be talking about Ideal Customer Profiles again, except today's conversation is going to be much more tactical. So we're talking about a lot more of the nitty gritty specifics than in previous conversations. So in episode 15, we talked about turning customers down and how to nail your niche. In 16, we talked about how to verify your ICP, but we never really went into detail about how to define it. And we've gotten a lot of feedback from listeners that ICP's are something that is an interesting topic and they want to learn more. So here we are.

So today we talk about what your ICP should look like, how to use your existing CRM data to outline it, to define it better. How to use segmentation models, depending on the maturity of your product and your market to make your ICP look the way that it should, what it is, how you use it, mistakes we often see people make, how long it should be, how often you should revisit it. Things like this are kind of what we get into today.

Before we jump into it, I want to give another shout out to T2D3, which is the book that Stijn is writing that documents a lot of the playbook that we use at Kalungi for the clients that we have. If you go to, you will see a form to register for early access. In that form, there is a field that asks who referred you. If you type in B2B SaaS Marketing Snacks, BSMS, or the podcast, whatever you may please, you will get an extra entry to win a free book. So if that is interesting to you, have at it. Otherwise without further ado, let's jump into it.

So, in a previous episode, we talked about your Ideal Customer Profile, but we talked about how to confirm it using Go-to-Market tests. But we didn't really talk a lot about how to define it and get to it, why you need an ICP, what it is, how you can create one and yeah, all those kinds of components. So I wanted to center today's conversation around that. I have lots of thoughts, so I can maybe start off and you can tell me if you agree with my understanding of this. And maybe we can have a conversation about where we dissent a little bit, if that's the case.

So I guess the first kind of question that usually comes up when people are starting to look into their ICP, which a lot of cases, people will come actually search B2B SaaS ICP, which will lead them to one of our blog posts on our website, and the questions that people have actually direct messaged me on LinkedIn relating to them or to that topic is, "What is it and why do you need it?"

And the way that I define an ICP, there's a lot of definitions online, it varies greatly, but the way that I see it is it's a list of characteristics and identifiers that you use to outline the ideal account for your product. So the ideal account being one that has the highest likelihood of getting value from your product and staying with you long-term. So these are the ones that you have the best ability to service, the ones that are going to become happy customers and hopefully become great referrals for you down the road. Is that kind of in line with how you understand the ICP or would you define it differently?


No, I think that's a great initial definition, Mike. I think the word ideal of course also can change over time. So, when you are a young company, what's ideal today might not be ideal tomorrow. For example, there are big risks when you are a startup, when you're a small software venture, to get customers that are, for example, too big, right? That can be very demanding and could almost become a magnet that pulls you in the wrong direction. I've seen small software companies dedicate half of their R and D capacity just to service one client. Which of course was fantastic from a financial perspective to get some initial revenue in the door, but it did not necessarily mean that that was ideal long-term. So, Ideal Customer Profiles change over time.

The other thing is that Ideal Customer Profiles for targeting purposes, “who do you optimize your messaging for? Where do you go look for more prospects?” Can work in relatively more traditional attributes, like “how big are these customers? What industries are they in? What type of technologies do they use?” As soon as you also need to apply it for Inbound Marketing, from what content do you build, how you optimize the things you say and the messages and the nurture you do during the customer journey, things like the Jobs to be Done also become very important, right?

It's a mistake to think of ICP only in the form of demographics, firmographics, technographics, you have to add this “what do the individual people actually care about” component. And sometimes it's done by personas as an extra layer on top of the Ideal Customer Profile definition. But you more and more see the Job to be Done framework, which is more about the work and less about what the individual people look like. That is also very helpful. So I think, let's focus on ICP today, but there's this richer other topic around it, around the people behind the ICP, who work there, that you also want to think about.


A hundred percent. And so you talked about something where I think a lot of early stage companies tend to make a mistake on, which is they define their ICP as what they... It defines this account that they think they want to go after. So they think they want this big enterprise because it's going to bring them a lot of money. It sounds really nice on paper, but I think there's two kinds of scenarios where you're creating your ICP.

One is where you don't have any data to start with. Right? So you're really just getting started. You've got an MVP, you're kind of testing it out in the market, and you're doing a little bit more guessing at what your ICP is, what those characteristics are that define your ideal accounts. And then the second scenario is you're kind of beyond the MVP stage and you're moving towards Product Market Fit and you do have a little bit of data and you do have a database of customers that you already have, that you can dig into and look to for insights here.

So what I think is really important when you're just getting started, you kind of have to make some guesses, make some bets a little bit when you're defining your ICP. But when you do have some data, there's no reason not to use it. So I think you have to look at the existing customer database, do a full export of it, highlight only the accounts that are really good fits for you. The ones that cost, as you were saying, Stijn, the least to service that pay you the most. That are the happiest. Just look at that group and try to find patterns between them.

So, what are the characteristics that they share that you can then take and use as filters to stamp onto your total market, to find more of them that look like them. And I think when we work with clients, even when we're talking to prospects on sales calls, that tends to be the biggest misunderstanding of what an ICP should be. It's not the account that you want to go after, it's the account that you've already identified as being the best possible fit. So now you need to go find more that look like them.


Yeah. It's really important that you base it on real data. Ideal Customer Profiles are best grounded in what has been successful so far, instead of what you hope would be successful. You touched on something else, the cost to service, the cost to acquire a customer versus the value of that customer. I think that's a really important point Mike, that what ideal means for you as a company changes given where you are in your growth maturity journey.

And we have these four stages that we like to think about, you're getting to MVP, Minimum Viable Product. Then you go to Product Market Tit and after Product Market Fit, you try to hit exponential growth, often following the T2D3 model, tripling your ARR two years in a row, doubling it three more years in a row. And then finally you get to what we call the rule of 40 land, where companies need to turn from high-scale to high profitability and to turn that corner of becoming what's called a 40% company that has a combined growth rate and profitability of 40%.

So for those four stages, Mike, here are some thoughts on what the Ideal Customer Profile could look like, right? For the first one, for an MVP, an ideal customer gives you a ton of feedback, gives you fantastic inputs to improve the way your product is fitting the market, so you can then build for Product Market Fit. So an Ideal Customer Profile when you're still in MVP stage is not about how big they are, or even maybe in what industry they are, it's about how willing are they to give you a lot of feedback to test your product, to participate in building the product with you. That's an Ideal Customer Profile tribute when you're in the early stage.

When you're getting to Product Market Fit, now you want the customers who provide you testimonials and give you great quotes. Even if they're not that big, you want the ones that help you, what's called cross the chasm, right? The problem with getting the Product Market Fit is really to go from the early adopters who are going to be super excited about your product, but they may not be willing to talk about it, with others, to get some of this early majority and who will provide you with testimonials that will help others who look like them get to trust your product and help you build credibility. So in that Product Market Fit stage, your Ideal Customer Profile includes the ones who will provide testimonials, or the ones who you're able to service so well that it will yield testimonials, right?




And then the Ideal Customer Profile when you have to hit that exponential growth curve at T2D3 is about lifetime value of the customer. What's the total value that this customer will bring to you, divided by the cost to acquire those customers? That's what you do when you want to hit T2D3 scale. You want to get as many customers with as high value as possible, high ARPU, low churn, for the lowest Customer Acquisition Cost.

And then in the last phase when you go to the 40% rule, to turn from high-scale to profitable growth, you now need to find customers with a high lifetime value. Those are still high ARPU, low churn, but now with the lowest cost to service. Not just low acquisition costs, but also low cost of service. So, Ideal Customer probably is different things based on where you are in that growth journey. And then to your point, it's those attributes that will lead you to help you find more of those.


And so, as you're going through the typical journey of growth as a company, how often should you revisit your ICP to make sure it's updated? The article that I have on our site, one of the last kind of next steps is keep it fresh, is I think what I said. Is to keep updating it and I've looked at it I think maybe once a year, but I'm curious to hear what your thoughts are on how often you should be revisiting it. I think revisiting it too often just causes confusion and doesn't give you enough time to actually put into practice what you've defined and learn anything from it to see if it's actually right or wrong. But I'm wondering if you have any thoughts on that?


Yeah. If you think of the four roles in the customer journey for a SaaS company, it's really marketing, initial sales, driving new ARR,  marketing drives MQL's, just simplifying it a little bit now, sales drives new ARR. Then you have account management, people who manage those existing account relationships to improve basically the lifetime value of these customers, mostly through ARPU, expansion of these accounts. Of course, reduce churn but then also improve ARPU. And then sometimes that’s called customer success when you really drive focus on that churn reduction.

I think that's where your answer lies. I think as soon as you see churn numbers indicate that your ICP needs updating, that's a great moment to do so. Another time is when your marketing and sales team feel they're running out of markets to go after, right? And that is, of course, a very sensitive discussion because marketing and sales will typically want to go broader than you may want to, as an early stage company. So you want to not unleash all that, stay really focused on the clients you can service well. But especially when marketing and sales have done a good job nailing that niche and you're a market leader in your space, then of course you need to expand that. But then let customer success and the churn numbers tell you when you're doing that in the wrong places. If it doesn't lead to great happy customers and testimonials, then you need to revisit that as well.


Sure. So the longer your sales cycles essentially, the higher the ACV you have, generally the more or the longer it's going to take for you to get enough feedback to actually determine whether or not you should be widening. I guess… what's the opposite? Broadening or specifying.


Yeah. That's a great point. The size of your typical client, whether you're chasing rabbits or hunting deer or signing up elephants, will make it a little different in how quick your feedback loop is. What the type of sale motion that you're running is, do you actually have people in your sales team talking with customers that allow you to get that feedback? Or is it more of a self-service signup, product led growth engine? So, that will make it a little harder sometimes to get that information.


Cool. And there's one more big thing that I wanted to talk to you about, which was the segment - what the ICP should look like. And then also what the segmentation models are, depending on the maturity of your market. I think it really depends. If you look up ICP online, there's a lot of different resources that tell you it should be this long drawn out document. There are some that say it should be really short. I tend to lean more towards the short side. I like to say you should be able to fit your ICP onto a single slide, and it should be very simple and very clear and straightforward. Even just some bullet points that tell you what the firmographic, technographic filters are, with a few signals that tell you if they're a better fit versus others, I think is enough. As long as it gives you filters to take and apply on to your Total Addressable Market and make it smaller. That tends to be my approach towards it.

If you go onto our site and search in the blog ICP, you'll see the article, which is "How to define your ICP" and attached to that is a template. And there's a slide that actually helps you build your ICP. But you, Stijn, in the book you have this interesting kind of model for defining which kind of... It's like different segmentation methods, depending on how mature your market is. And it's basically you take Geoffrey Moore’s adoption curve, which is you have these five stages. You have innovators, early adopters, early majority, late majority and laggards, and it plots it on a chart. Then you have these lines that cross through it over time, which is an area chart. And it basically tells you how you can segment the market at each one of those stages, and generally get a good idea for who your ideal customer is. I don't know if you can explain that over the podcast, but I'd love to see if we can try to do that, if that makes sense?


Yeah. I gave this a lot of thought, Mike, over the years. It is hard to split two parts of market segmentation. How do you segment companies, businesses, if you're in a B2B space? You target certain types of organizations. That's really what we mean when we say Ideal Customer Profile. But then you also have the individuals because when we're doing marketing, we're of course marketing to people, to the people who work within those businesses. So you always get this mix of things like firmographics that are technographics, type of verticals, industry size of companies. And you have to do more persona related segmentation, demographics, Job to be Done is often very similar to that, it's more about how individual people are trying to get a certain job done.

And so there's a reason to split those up and think about personas and jobs to be done more when you're thinking of content for certain segments of the market, and to use more the company segmentation models when you need to understand what type of lists to build for your outbound campaign. Or maybe which type of analyst to try to market to, or to influence, covering a certain market segment that's typically the type of companies they cover.

So this model that's on page 94 in the book, or in chapter five, is a little bit of a combo, but it comes at it in a different angle. So let me try to explain it. When your category is maturing, at some point in the B2B space, there will be someone almost accountable in your customer's organization to buy maybe the type of solution that you're selling. Let's say you're in a very mature market and you're selling a spend management software, some type of solution to match invoices with purchase orders, with delivery, no late bills and things like that. If that's a mature market, there's probably someone in an organization that has the title procurement manager, so now demographics that include job title will work in that relatively mature category.

If the segment, the category is even more mature, then there might even be companies that are in more need of these types of solutions than others, and you can find those. You notice certain verticals. So firmographics and demographics and things like job titles are very suitable for a mature category where your solution is so well understood that there will be people that have job titles that literally are looking to buy what you have to offer.

If the category is not so mature, then at some point the people you want to target have completely different job titles, you have no idea. Project manager this, or a role that just happens to be having this problem assigned to them, but you have no idea what the title for that job is. Maybe you can find when it's Job to be Done, if that's your segmentation model, you can find people who are assigned working on that problem. For example, what they're describing they're working on in their LinkedIn profile. Might not be the job title, but there might be, "Hey, I'm in charge of thinking about using artificial intelligence to improve our customer success from customer service," or something like that. It might not be a job title for that person yet, but at least you know what they're working on. So Job to be Done might be more suitable if the category is less mature.

If you're in a completely new category where you're really targeting innovators, because you're not only creating a new category, you're looking for other people to help you create that category. Then of course, things like psychographics would become more suitable because now you want to look for people who are willing to try new things. People who hang out with other people who try new things. People who have self-declared to being innovators. And these are typically the early adopters that hang out in maybe Reddit sub-groups that talk about a problem that you're trying to solve. The problem doesn't have a name yet, but you find those people who are excited about solving that early on.

So I think that's what this graph in the book does. It talks about in the early stage of a category, innovation, early adopter psychographics are probably a great segmentation model. Job to be done might be super helpful. And then as a category matures and you get the early, late majority, then demographics might be relevant because now people have job titles that relate to the problem that you're solving, that your solution is addressing. And if the category gets even more mature, then you'll be able to say, "Hey, this vertical, this industry, this type of customer has this need more than others, and that's where I should target my segmentation." Does that make sense?


That makes perfect sense, yeah. Love it. That's really helpful. So the last component that you really have is the difference between filters and signals. So on an ICP slide, on the left side, I generally like to put the filters, which are those things that you're talking about which are firmographics. So for instance, they have an annual revenue that's greater than one million, or they have a company headcount that is between 20 and 100, or they have a person with a job title that is equivalent to X, Y, or Z.

And then on the right side, I like to have signals. So if you think about your filters, I like to think about your Total Addressable Market. Think of it as just a giant sheet of cookie dough, and your ICP is a cookie cutter. You're basically taking your cookie cutter and you're stamping it on this giant sheet of cookie dough, and now whatever that cuts out, let's say it's a heart. That heart is your ICP.

Now within that heart, there are going to be pieces of that cookie that are going to be more desirable to you than others. Maybe when you cook it, when you bake it-


Get the pieces of chocolate out. 


Exactly, yeah. There's pieces of chocolate in it. When you bake it, the edges get crispy and the center is a little soft. Maybe you like the softer part. So the signals are what you're going to look at to tell you within that ICP, which parts of it do you want first? So when you go and build your list of accounts and contacts to do ABM, who are you going to actually message first? Who is going to be tier one? Who's tier two?

So those are the things that you use as signals. And they're usually, at least what I've noticed, is that they're often not as public or easy to find. So they're kind of like the things that are hiding, as you were mentioning Stijn, in somebody's LinkedIn profile or it's noted somewhere in maybe like a press release that they received funding.


Yeah an announcement they've made.


Yeah. It's not usually like a public data point. It's something that you kind of have to identify through conversation or a little bit of research. So some examples would be like, for us, for Kalungi's ICP, we have a signal that if there is a recent change in marketing leadership, it's a good signal for us because it means that maybe they had a marketing leader who wasn't a good fit. We kind of serve that function in a lot of cases. Another good signal for us is a company who has recently gotten funding. It's not necessarily indicative, it's not a filter, but it's a really good signal for us. So I just wanted to make that distinction as well, that there's a difference between the two and your filters generally will be fewer and your signals generally more.


Yeah. Your filters are really meant to filter out all the customers and prospects that you know are not going to be ideal and they actually are even problematic. Or you're not going to be able to service them well. I think you mentioned Total Addressable Market. I think that's where you go from Total Addressable Market, to the SAM, the serviceable part of that addressable market. You're basically saying, "This is a group of customers that I can service very well. And these others, I really should not target them because I cannot service them that well." So an industry would be a good example, or certain size of company, et cetera.

And then within that Serviceable Addressable Market, you will have a preference of certain customers over others, and that's where you use the signals. It could be a preference because they have more appetite to, for example, buy your solution faster. You have limited resources, so ideally you rank these prospects in a way that is helpful for you to get the most ideal customers first. That could be because they're hiring for a certain role or they just created budget for it. Or they're in an area of regulatory impact where there is suddenly a certain amount of urgency. Those are all great signals to help you go from, "What part of the market can I potentially service, and what part of that market is for me the easiest to obtain?" Now you go from SAM to SOM to the Serviceable Obtainable Market, given you have limited resources, limited time, and it helps you to drive that extra focus.


One last technographic that I really like using is if you can find it, it can be really helpful for a lot of software products. You often are combining, or there's a signal that a company is a good fit because they use a certain type of software, or you integrate well with a certain type of software. Oftentimes if you can search somebody's profile and see that they've listed they're an expert let's say in Workday, or they're an expert in HubSpot or Pardot and you integrate with them, or because they use that, you maybe can replace that software or enhance it. That can be another really good filter to look for as well.


Filter or signal. It's like in a persona, when you see people wear certain things it tells you, "Hey, are they trying to be cool? Are they trying to be innovative?" And companies do the same thing. If you see in their technographic profile that they're using certain technologies, it tells you two things. It tells you, are they a good fit, a filter, because maybe your software works very well with certain other technologies or not. Or it tells you, "Hey, they are innovative. They like to try new things. They're investing in cutting edge solutions." Or the very reverse. They're very conservative in the type of technology, which might also indicate a great fit for what you do.

So I think the technographic footprint of a company if you can find it, gives you a ton of insights. You can buy technographic data from selling companies and things like that. There's all kinds of software that scans for cookies and things that happen on websites that tells you something about the technology companies use. But also to the earlier point, you can go on LinkedIn and find out, not based on the typical standard LinkedIn profile data, but what people say in their profiles, what type of technology they're using in a certain company.


Yeah, absolutely. Cool. Anything else on ICP?


Well, ICP in the end is about carving out the part of the market you want to go after, you want to focus on. It's also a very strategic question. It's not just about, "Okay, which customers have I serviced well in the past? Who are the happiest? Who have given us the most testimonials, lowest churn?" Et cetera. It's a great input, the historic data. But I think maybe as a final point don't forget about, what's your strategy here? Are you trying to dominate a certain market?

And then of course you want to see, "Hey, what's my beachhead?” How have I started and how do I expand in the most economic way, in getting the next set of clients that are going to be very happy? But if you're starting to differentiate, if you're in a very crowded space with a lot of competition and it's like a red ocean, so to speak, then it's all about niching down and finding that part of the market that may be under-serviced, that doesn't get exactly what they need. And can you find this part of the market where you can be the unique, best provider of a solution for a certain type of industry, for a certain sized customer, for a certain job to be done.

Or are you trying to disrupt the market? Are you trying to find customers who are over-serviced, who get more than they need? You could come up with a complete new positioning and new pricing, et cetera. And both those things should go into your ICP. What's your historical success, but also what's your strategy? Do you want to dominate a certain market, differentiate and carve out a new niche? Or disrupt an existing market and maybe take a completely different approach to it? Because that would mean a different ICP potentially in the future, than what you have serviced so far.


Great point. And keep it simple.


Keep it simple. Yeah, exactly. Until it's too simple, and then it's not valuable anymore.

B2B SaaS Marketing Snacks - More Episodes

Listen to more episodes

Head back to the B2B SaaS Marketing Snacks home page for more.

More episodes

Similar posts

Get notified on new marketing insights

Be the first to know about new B2B SaaS Marketing insights to build or refine your marketing function with the tools and knowledge of today’s industry.