Why ICP Workshops Fail to Fix an Unclear ICP
A founder on Indie Hackers wrote: “Sent 170 cold emails to companies with 30-200 employees. Zero replies.”
That is why ICP workshops fail. They produce a description of who could buy: size, industry, title. They never measure what state the buyer is in right now. Two companies with the same profile can sit in different states, and only one is ready to move. A workshop settles the profile and leaves the state unresolved.
Median B2B SaaS spend: about $2 to acquire $1 of new ARR (Stijn Hendrikse via T2D3, 2025).
Why does an ICP workshop look like the right fix?
Because the pain it targets is real. Sales says the leads are bad. Marketing says sales can’t close. Nobody agrees on who to go after, and the team is growing, so every new hire brings a new opinion.
A workshop ends that argument. It puts founders, sales and marketing in one room. It forces a written answer to “who is this really for.”
That has value. A shared definition beats five private ones. The profile also does one job well: it filters out companies that cannot buy. Wrong size, no budget, wrong sales motion. Cut them.
Smart teams pick workshops for good reasons. The method is not wrong. It is measuring one variable and calling it the whole picture.
Where does the workshop break once the team starts selling?
It breaks at handoff. The deck gets approved, the list gets built, and reps start sequences. Results come back mixed, and the room is quiet because the workshop already “solved” targeting.
So the team blames the layer below. Messaging gets rewritten. A new SDR gets hired. Spend goes up.
Meanwhile the numbers move the wrong way:
- New customer acquisition costs rose 14% year over year (Pavilion 2025 B2B SaaS Benchmarks).
- Median CAC payback extended to about 20 to 23 months, up from a historical 12 to 14 (KeyBanc / industry synthesis, 2024-2025).
Every month of state-blind outreach is paid for against a longer payback window. That is the cost of a profile that cannot tell ready from not ready.
For example, a VP Sales joins a $5M ARR team after a workshop. The agreed ICP reads: Series A to B software company, 50 to 200 employees, VP Marketing as champion. The list holds hundreds of matches. This is a composite of a common pattern, not a client case.
Three accounts on that list look identical on paper.
The first VP Marketing just inherited a missed quarter. Their public posts say “we’re getting traffic but no pipeline.” The second renewed a competitor for two years last month and is browsing for a board slide. The third is a new team lead who has no budget until next quarter.
All three get the same sequence. One replies. The rep concludes the sequence is weak. The founder concludes the list is weak. The workshop deck says all three are ICP.
Scale the team and the pattern multiplies. More reps send the same state-blind message to more mixed accounts. CAC rises with headcount. This is the friction behind “we’re seeing CAC increase as we scale the team.” The signals conflict because the targeting has no way to sort them.
What variable does the workshop ignore?
It ignores the buyer’s internal state at the moment of contact.
Signal Resolution’s Psychographic Resolution™ is the layer that reads it. Signal Resolution defines Psychographic Resolution as the process of identifying buyer state using real observed language, not assumed personas. It is the third layer of Signal Resolution’s ICP Resolution System™, and it runs after the profile filter, not instead of it.
The inputs are words buyers actually used. Sales call transcripts. Lost-deal notes. Public posts. Review sites. Phrases like “nothing is converting,” “CAC is rising and we don’t know why,” and “we don’t know who this is really for.”
Each phrase marks a different state. The first VP in the example is urgent and blaming the wrong layer. The second is comfortable and shopping for cover. The third is blocked. The workshop profile sees one company type. The language sees three different buyers.
This gap shows up in the data on how sellers and buyers frame the deal:
- 54.5% average misalignment between how sellers and buyers define the core problem to solve (Emblaze, 2024).
- When seller and buyer align on the problem definition, win rates improve 38% (Emblaze, 2024).
A workshop aligns your team with itself. It does not align your team with the buyer’s problem. Our guide to buyer states shows how these states differ in practice.
What should you do instead of another workshop?
Keep the workshop output. Add the layer it skipped.
- Keep the profile as a pass or fail filter. Companies that cannot buy never reach outreach.
- Pull the exact words from your last closed-won and closed-lost deals. Quote them. Do not summarize them.
- Sort those phrases into states. Group by what the buyer was feeling and why they acted, not by title.
- Write one message per state. A message built for urgent buyers will read as pressure to a comfortable one.
- Point reps at the state that can move this quarter. Park the rest.
Other approaches stop short of this for a founder-led team. Intent platforms are fast and surface accounts that are researching. They are also account-level, and 50% of companies using third-party intent data report too many false positives (Forrester, cited by Prospeo, 2026). Research activity is not readiness.
Timing also works against them. 95% of the time the winning vendor was already on the buyer’s Day-One shortlist (6sense, Buyer Experience Report, 2025). The frame is set before the surge shows.
Outbound stacks execute well. They assume you already know who to reach and what is broken, so they run the wrong motion faster. Fractional CMOs and agencies bring senior judgment. The buyer understanding lives in one head and leaves when the contract ends.
Resolve the state first. Then decide what to buy or automate.
FAQ
Do ICP workshops have any value?
Yes. A workshop gives the team one shared written profile and filters out companies that cannot buy. That is a real gain. The problem starts when the profile is treated as the finished ICP. Keep it as the first filter, then add buyer-state language on top of it.
Why do users love the product but buyers won’t pay?
The person using the product is often not the person with the budget or the urgency. A profile built from user interviews describes the user. Buyer state comes from the language of whoever signs, and from what pressure they are under right now.
Why do we get traffic but no pipeline?
Traffic fills the top with mixed states. Some visitors are ready, some are researching for a slide, and some cannot buy this quarter. A profile treats them all as equal. Sorting by state shows which group your message actually fits and which one it quietly repels.
How long does it take to resolve an ICP?
It depends on how many deals and conversations you can pull language from. A founder can start this week with closed-won and closed-lost notes. The first pass is rough. It improves as more real phrases are sorted by state.
What to check before the next workshop or tool
If you are about to book another workshop or buy another tool, test the ICP you already have. Run the Signal Diagnostic as the self-evaluation step before choosing any approach. It shows which layer of your targeting is breaking.