The Science of Scaling · 23 Apr 2025
The Critical Importance of Focusing Your ICP w/ Dan Sperring (Founder, AlignICP)
These are notes on the conversation, checked against its transcript. The episode itself has the full discussion.
In brief
Mark Roberge hosts this HubSpot Original episode and interviews Dan Sperring, founder of AlignICP, about focusing on an ideal customer profile. They discuss how product success often depends on the specific use case a customer buys for rather than demographics, and why founders should find product-market fit in their existing customer base before broadening their target market. The conversation also covers turning an ICP into a CRM-tracked program, with benchmarks on pipeline and install base share, and briefly touches on how AI may change ICP definition and CRM data. The most important argument is to be the best in the world at one use case first, then earn the right to expand.
For founders
- Be the best in the world at one use case first, then earn the right to expand by solving more use cases.
- Before expanding your target market, look at your existing customers to find where product-market fit is strongest, using lifetime value, retention and net revenue retention.
- Codify product-market fit by defining a leading-indicator event and measuring the percentage of customers who reach it, then check a year later whether those customers retain better.
- Mark argued that a seed-stage company does not need to prove its vision with every customer in every segment, and that building and selling to a huge market from the start risks failure.
- Mark argued that product-market fit is about where customers see value that drives retention and expansion, not about where you close deals.
For revenue leaders
- Operationalize your ICP by writing its definitions and tagging accounts in the CRM with those labels, so pipeline quality can be measured.
- Mark's team listened to 20 discovery calls and found tailored demos fell into three buckets, which reps were then trained to choose between.
- Dan said that in his client work, less than 20% of pipeline is typically with ICP customers, and less than 30% of a horizontal SaaS install base is.
- Dan cited a HubSpot stat that organizations with well-defined ICPs have 68% higher win rates, and that fewer than 7% of salespeople feel marketing creates high-quality leads.
- Mark said onboarding can be configured by use case so it feels custom while still scaling as services.
What was said 24, most useful first
Onboarding outcomes at an early-stage SaaS company tracked which customers bought, not how they were onboarded Listen
Dan joined the company in 2013 and eventually became a customer success leader there, running quarterly win-loss analysis on customer outcomes. Some customers were up and running within weeks and tended to grow and expand, while others still had not implemented the software six months after signing, even though the same onboarding and implementation teams worked with both groups. Dan concluded the difference lay in the customer and the personas using the software rather than in the onboarding.
“Conclusion was it wasn't really the variance in how the customer was onboarded. It was more of a variance about who the customer was.”
Listen to the episode Retention & customer success Link to this Report a problem
Be the best in the world at one use case first, then earn the right to expand by solving more use cases Listen
Dan advised staying a point solution focused on one use case rather than pursuing a broad market, and expanding only after that use case is won. He said this matters a lot for your persona, and that focusing on the use case is how you earn the right to expand.
“be best in the world at that one use case. Earn the right to basically expand by focusing on solving use cases.”
Listen to the episode Strategy & market Link to this Report a problem
Listening to 20 discovery calls showed HubSpot's tailored demos fell into three buckets that reps could be trained to choose between Listen
Mark did this with HubSpot's first sales trainer, Andrew Quint. They spent a day listening to 20 discovery calls and crafting a demo for each customer, then found those demos fit three buckets. Rather than asking reps to tailor every demo, they trained account executives to work out what the buyer needed and choose one of three demo options.
“We sat in a room for a day and we listened to 20 discovery calls of 20 different customers.”
Listen to the episode Sales process & deals Link to this Report a problem
Before expanding your target market, identify where product-market fit is strongest in your existing customer base Listen
Dan said most founders who want a different ICP can't say what share of their market they have already penetrated. He suggested starting with the existing customer base and looking at customer lifetime value, retention and net revenue retention to find where fit is strongest before going to adjacent markets.
“Start by looking at your customer base and saying, you know, of our 500 customers, where do we have the strongest product market fit?”
Listen to the episode Strategy & market Link to this Report a problem
Product-market fit may already exist in segments a company never targeted, so doubling down there can beat entering an adjacent market Listen
Dan said that companies that haven't done this analysis will likely find fit in segments they never pursued. As an example, they may have organically acquired 35 to 40 customers in a segment they weren't proactively targeting that are growing at a healthy rate, being retained and giving positive NPS scores. He advised doubling down on these existing customer types rather than entering an adjacent market to find fit.
“you've organically acquired like 35, 40 customers in a segment that you haven't been proactively targeting that are growing at a very healthy rate”
Listen to the episode Strategy & market Link to this Report a problem
Codify success as the percentage of customers who reach a defined leading-indicator event, using examples such as Slack's 70% sending 2,000 team messages a month Listen
Mark suggested calling the leading indicator retention, expressed as the percentage of customers who achieve a specific event every time. He gave examples as he stated them: Slack, 70% of customers send 2,000 team messages every month; Dropbox, 85% back up the device every day; HubSpot, 80% use five more features in its 25-feature platform every month. He said the event is not the same for everybody, but it can be found and then measured.
“we can also call the leading indicator retention, which is P% of customers achieve e-event every time.”
Listen to the episode Metrics & finance Link to this Report a problem
Operationalize an ICP by writing its definitions and tagging accounts in the CRM with those labels, so pipeline quality can be measured Listen
Dan said many companies have no programmatic process to operationalize their ICP, which is why it fails demand gen and sales. He said the starting step is creating the definitions and tagging accounts in the CRM, after which you can measure your ability to build quality pipeline and close deals with accounts more likely to renew and expand.
“And it starts with us creating these definitions and then tagging our accounts within the CRM with these labels.”
Listen to the episode Pipeline & demand generation Link to this Report a problem
In Dan's client work, less than 20% of pipeline is typically with ICP customers, and less than 30% of a horizontal SaaS install base is Listen
Dan said that for vertical SaaS companies it is closer to 50% of pipeline and around 60% of the install base. He gave these as what he typically sees when working with clients.
“As we work with clients, what we typically see is less than 20% of the pipeline is with ICP customers.”
Listen to the episode Pipeline & demand generation Link to this Report a problem
Companies with a very tightly defined ICP that is operationalized on the front line spend 50% less on sales and marketing, with CAC payback 24% shorter than the mean Listen
Mark read this data in the episode introduction, along with the claim that ICP customers have a 425% higher expansion rate in their first 12 months than customers outside the ICP.
“Companies with a very tightly defined ICP that is operationalized on the front line spend 50% less on sales and marketing.”
Listen to the episode Metrics & finance Link to this Report a problem
Software built for a handful of use cases loses its ability to deliver value as more use cases are added, because jobs to be done differ by use case Listen
Dan said products are maybe good at one to four use cases, and as they reach the eighth, ninth or tenth, customers' ability to drive value from each one is compromised. He gave retail and media as examples whose jobs to be done and the features they use can be foundationally different.
“But as you start to get to like the eighth, ninth, tenth use case, the ability for customers to actually drive value from that use case gets compromised.”
The amount of services included in a deal can significantly affect whether the customer derives value from the purchase Listen
Dan described pricing and packaging as a second dimension alongside the use case. He said how a deal is structured, particularly how much services are included in the quote, can have a dramatic impact on the client's ability to derive value from the purchase.
“The way you structure the deal, like the amount of services you include on it can have a dramatic impact on the client's ability to actually derive value from their purchase.”
Listen to the episode Pricing & packaging Link to this Report a problem
A new customer's use case is foundational to whether they derive value, which Dan said lets you predict who will become a long-term customer Listen
Mark asked whether a short interview at signup about the use case behind the purchase could predict a long-term customer, and Dan answered without a doubt. He said the use case is foundational to a client's ability to derive value.
“Without a doubt. And so the use case is foundational to a client's ability to derive value.”
Listen to the episode Strategy & market Link to this Report a problem
Onboarding can be configured by use case so it feels custom while still scaling as services Listen
Mark said that if a prospect is sold on their use case and then goes through a generic onboarding, the continuity breaks. He said tailoring onboarding to the use case, as the demo was tailored, feels custom but is configured, so it still scales in services.
“It feels custom, but it's just configured. So it still scales in services.”
Listen to the episode Retention & customer success Link to this Report a problem
Building for a huge market and selling to all of it from zero to one can mean boiling the ocean and failing Listen
Mark said entrepreneurs often paint a big picture for their team, then build and sell to that huge market through zero to one and one to ten million dollars, which he said leads to failure. He argued that a seed-stage company does not need to prove its vision with every customer in every segment, and should instead find a first minimal viable product and market.
“you end up boiling the ocean and you fail”
Listen to the episode Strategy & market Link to this Report a problem
A company can exploit a simple use case while testing new markets or products, expecting most tests to fail and one to fund the next stage of growth Listen
Mark gave a hypothetical in which a company exploits its simpler use case from zero to one, one to five and five to 15 while running a few tests, such as enterprise or a tangential product, set up so they don't exhaust resources. He said two may fail and one may succeed, and the winner gives time to grow from 15 to 30 to 60.
“Two will fail, one will succeed, and that one that succeeds gives you the next three years to go from 15 to 30 to 60.”
Listen to the episode Strategy & market Link to this Report a problem
Product-market fit comes first, and go-to-market efficiency metrics are the second most important driver of company valuation Listen
Dan said the belief behind his work is that a company starts by understanding and optimizing for product-market fit. He named LTV to CAC ratio and the price of acquiring ARR as efficiency metrics, and described the path as iterative, starting in a market where you can help customers derive immense value.
“The belief that we have is it starts by really understanding and optimizing for product market fit. The second most important variable to drive company valuation is your efficiency metrics related to go to market.”
Listen to the episode Metrics & finance Link to this Report a problem
Product-market fit depends on where customers see value that drives retention and expansion, not on where you close deals Listen
Mark said choosing an ICP based on where you are closing customers is market-message fit, which is a different thing, and gave the idea that you can sell ice to Eskimos. He said product-market fit is about where people see value from the product, which leads to retention and expansion.
“It's product market fit is about where people are seeing value from your product that ultimately leads to retention and expansion.”
Listen to the episode Strategy & market Link to this Report a problem
Check a leading indicator by comparing retention a year later for customers who hit it against those who didn't, and mine logs if there is no difference Listen
Mark said that after a year you can compare cohorts. If customers who hit the indicator have high retention and those who don't have low retention, the indicator is validated. If there is no variation, he suggested analyzing customer and user logs and running correlations, for example to test whether the number of messages or the number of users matters more.
“those that did hit the leading indicator have very high retention, and those that do not have very low retention.”
Listen to the episode Metrics & finance Link to this Report a problem
A HubSpot stat Dan cited says organizations with well-defined ICPs have 68% higher win rates Listen
Dan cited this as a HubSpot stat and said it is probably not surprising, since companies that know their customers best probably do much better on the go-to-market side.
“organizations that have well-defined ICPs, they have 68% higher win rates”
Listen to the episode Sales process & deals Link to this Report a problem
Fewer than 7% of salespeople feel marketing is creating high-quality leads, according to a HubSpot stat Dan cited Listen
Dan cited this stat and attributed the gap to the lack of a programmatic process to operationalize ICPs. He asked how to turn an ICP into a program that guides go-to-market activity.
“less than 7% of salespeople feel like the marketing organization is creating high quality leads.”
Listen to the episode Pipeline & demand generation Link to this Report a problem
Picture the target market as a dartboard with the ICP as the bullseye, the small subset of best-fit customers that renew, expand and grow Listen
Dan said best-fit customers typically represent a small subset of the install base, and the ICP's purpose is to focus the organization on the accounts that will drive the business forward. He said that for demand gen teams with a $3 million budget to create great leads, the ICP typically fails them because it doesn't give them the information they need, so they end up leaning into quantity over quality.
“so your ICP is the bullseye of the dartboard.”
Listen to the episode Strategy & market Link to this Report a problem
Dan predicted AI could help keep CRMs focused on the accounts that matter, with contacts licensed and updated as buying committees change Listen
Dan said a CRM may hold around 30,000 accounts, with a small subset active customers and probably around 3,000 in the ICP. He said people change roles every 15 to 24 months, so a contact has a shelf life, and he sees a day when contacts are licensed and constantly updated rather than bought in fixed batches. He framed this as an expectation of the future rather than something done today.
“And so there is a shelf life in terms of the value of a contact.”
AI could arguably define the ICP from customer attributes that correlate with lifetime value once it reaches a mature state Listen
Mark said that at a mature state, AI could do a better job than humans, who work with limited simplicity and error, defining the ICP by use case, persona, demographics or preference. He said right customers could then be identified before a sales conversation and the ICP aligned with demand gen strategy and investment.
“AI arguably at a mature state can do a much better job of defining the ICP according to something about the customer”
A company does not have to pick a vertical; it can pick a use case that is prevalent across multiple horizontal markets and be best at it Listen
Mark raised the concern that choosing a vertical hands other verticals to competitors who might outmaneuver a company with more capital. Dan's answer was to pick a use case that is very prevalent across multiple horizontal markets and be best in the world at it as a point solution.
“I'd argue you don't have to actually pick a vertical, pick a use case.”
Listen to the episode Strategy & market Link to this Report a problem