TemplatesResearchOpsWyse

ICP Research Brief Template

Most ICP docs are one paragraph that says 'mid-market SaaS.' This is the nine-section brief that defines an ICP precisely enough to filter a lead list.

2 to 3 hours first time, 30 minutes quarterly after thatDifficulty: mediumFree, no signup

Why this matters

1A precise ICP cuts your top-of-funnel volume by 60-70 percent and your meeting-to-close rate up by 2 to 3x. The math always favors precision.
2AI agents (Wyse, Apollo Agent, Clay) can filter a lead list against a precise ICP in seconds. They can do nothing useful with a vague one.
3Most SaaS companies have an ICP doc that says 'mid-market SaaS in North America.' That isn't an ICP, that's a TAM slice. This template forces precision.
4ICP-fit deals close 4.2x faster than non-fit on average. Speed compounds across the year more than win rate does.

The template, step by step

01
Section 1: Firmographics, hard filters only
Industry (use NAICS codes if you can), company size band (employee count), revenue band, geo. These are the filters where 'maybe' is wrong. If you're not sure whether to include 51-100 person companies, you don't have a tight ICP yet.
02
Section 2: Stage and growth signals
Funding stage, growth rate, recent hiring patterns, recent product launches. These are the leading indicators that a company is in the buying window for your category. Most ICP docs skip this entirely. It's often the most predictive signal.
03
Section 3: Tech stack
What they currently use that signals they're in market for you. Specific tools (HubSpot, Pipedrive, Outreach), not categories ('CRM'). Tools they use that signal they're NOT for you (lock-in incumbents). AI tooling adoption signals.
04
Section 4: Trigger events
What specific event in the company makes them ready to buy this quarter? New hire in a specific role, recent funding round, recent layoff, recent product launch, recent leadership change, recent SOC 2 announcement. List 5-8 trigger events.
05
Section 5: Buyer persona (decision-maker)
Title family (e.g. 'Head of Revenue / VP RevOps / Director of Growth'). Years in role band. What they read (newsletters, podcasts, books). What conferences they attend. What they're measured on quarterly. What 'success' for their year looks like.
06
Section 6: Buyer persona (champion)
The person inside the company who will recommend you and run the trial. Often a different person from the decision-maker. Title family. Years in role. Pain points that are urgent (not just nice-to-have). What they could do with your tool that would make them a hero internally.
07
Section 7: Anti-ICP
Who looks like ICP but isn't. Companies that match the firmographics but never close. Patterns that signal a wasted demo. Industries you don't sell to even though they could afford it. This section saves more time than every other section combined.
08
Section 8: Win and loss patterns
Pull your last 20 closed-won deals: what 3-5 things did they have in common? Pull your last 20 closed-lost: what 3-5 patterns did they share? Add to ICP / anti-ICP respectively. Update this section every quarter.
09
Section 9: Verification list
5-10 specific company names that match the ICP exactly and are not yet customers. If you can't name 5, your ICP is too narrow or too vague. Use these as the testing ground for the sequence templates.

Nine-section ICP brief

1. FIRMOGRAPHICS (hard filters)
   Industry / NAICS: ____________
   Employee count: ____________
   Revenue band: ____________
   Geo: ____________

2. STAGE + GROWTH SIGNALS
   Funding stage: ____________
   Growth rate band: ____________
   Hiring patterns: ____________
   Product velocity: ____________

3. TECH STACK
   Tools used that signal fit: ____________
   Tools used that signal NOT fit: ____________
   AI adoption signals: ____________

4. TRIGGER EVENTS (5-8)
   ____________
   ____________
   ____________

5. BUYER PERSONA (decision-maker)
   Title family: ____________
   Years in role: ____________
   Reads / listens to: ____________
   Conferences: ____________
   Measured on: ____________
   "Win the year" looks like: ____________

6. BUYER PERSONA (champion)
   Title family: ____________
   Pain points (urgent): ____________
   Hero moment with your tool: ____________

7. ANTI-ICP
   Look-alike but never closes: ____________
   Demo-killer patterns: ____________
   Industries we don't sell to: ____________

8. WIN + LOSS PATTERNS
   Last 20 won: 3-5 commonalities: ____________
   Last 20 lost: 3-5 patterns: ____________

9. VERIFICATION LIST
   5-10 specific companies that match exactly: ____________

Common mistakes

Treating ICP as a one-time exercise. ICP shifts every quarter as your win/loss data changes.
Skipping section 7 (Anti-ICP). Most ICP docs only describe who you want, not who looks similar but isn't. The look-alikes eat 30 percent of your top-of-funnel time.
Not naming verification companies. Without 5-10 named examples, the ICP is theoretical and nobody can use it to filter a lead list.
Confusing ICP with persona. ICP is the company. Persona is the person inside the company. Both matter, neither is the other.
Letting marketing and sales have different ICPs. Top symptom: marketing's MQLs aren't sales' SQLs. Fix at the ICP layer, not the qualification layer.

Skip the manual work

Let OpsWyse run this template on autopilot.

Wyse drafts every input, every personalization, every follow-up in your brand voice. You approve before anything goes live.

Try OpsWyse See pricing

Questions

How long does the ICP brief take to fill out?

2-3 hours the first time if you have 6+ months of closed deal data to mine. 30 minutes for the quarterly refresh after that. Most of the time goes into section 8 (win and loss patterns), which is the highest-value section.

How often should the ICP be revisited?

Quarterly minimum, monthly if you're under $1M ARR and your customer base is shifting fast. The trigger events and anti-ICP sections shift the fastest. Firmographics and persona shift the slowest.

What if I don't have 20 closed deals yet to mine for patterns?

Use closed lost as a starting signal: any patterns there? Use prospect-to-meeting conversion rates as a proxy for ICP-fit until you have closed data. Better yet, treat the early ICP as a hypothesis and revisit every two months with whatever data you have.

Can AI build the ICP for me?

An AI can synthesize an ICP if you feed it your closed-won and closed-lost deal records, your CRM data, and your demo recordings. Wyse does this as a starting point and asks you to verify each section. AI without your data builds an ICP that fits no specific company.

Should the ICP be in the CRM?

Yes, as filterable fields on the lead and account objects. Otherwise the ICP lives in a Notion doc nobody opens during prospecting. The whole point is to filter a lead list against it automatically. If it can't filter, it doesn't matter.

Related templates

Sales