Most B2B teams have an ICP document. Very few have an actual ICP. The document says "we sell to 50-500-person companies." The actual ICP, the scored, named-account list that outbound runs against, is missing. That is the gap that keeps reply rates stuck at 2-4%.
At Bring My Customer, every client engagement opens with a week-one ICP build using the framework in this guide. We score 500 named accounts across 6 dimensions, filter down to the 70+ tier, and only then does the specialist begin reaching out, one hand-written touch at a time, across LinkedIn, email, Twitter/X, DMs, SMS, and phone. That one discipline is why our clients see 8-13% reply rates while the rest of the industry sits at 2-4%.
Who this guide is for
Founders, revenue leaders, and marketing heads who have been running outbound for 3+ months with mediocre results. If your reply rate is under 5%, your ICP is almost certainly the problem. This guide gives you the Bring My Customer ICP framework in enough detail that you can rebuild yours this week.
What a real ICP is (vs what most people call an ICP)
Here is the honest version. An Ideal Customer Profile is a precise, scored definition of the type of company most likely to buy from you, close fast, retain long, and expand. It is not a slide in your pitch deck. It is not a sentence that says "mid-market B2B business." It is a 500-2,000-row list of named accounts, each with a 0-100 fit score, each with contact enrichment, each with recent trigger events tagged.
If a stranger could not pull up your ICP, pick one row at random, and tell you specifically why it scored 83, you do not have an ICP. You have a hypothesis.
ICP vs. buyer persona
- A persona describes a person. "Marketing Mary, 35, VP Marketing at a $20M business, cares about MQLs."
- An ICP describes an account. "50-200-person B2B business in North America, headcount growing > 15% YoY, hired a VP Sales in the last 90 days, running HubSpot + Salesloft."
Personas help you write the message. ICPs decide whether you send the message at all. B2B outbound lives on the ICP side of that line.
Why loose ICPs kill outbound
Loose ICPs are the silent assassin of B2B outbound campaigns. The pattern is predictable:
- Team defines ICP as "any mid-market B2B business."
- A broad database query returns 48,000 companies.
- Team scrapes contacts and blasts 10,000 generic cold emails.
- Reply rate: 1.8%. Booking rate: 40%. Show rate: 55%.
- Team spends $9K on tooling and 3 months of SDR salary for 12 meetings.
- CFO cuts the outbound budget.
Compare that with a tight ICP, 500 named accounts, each scored 70+ and researched by hand, each touch written one-to-one by a specialist. Reply rate: 10-13%. Booking rate: 65%. Show rate: 78%. 25-40 personalized touches a day across a focused book beats 2,000 bulk sends across 48,000 strangers, by orders of magnitude.
This is where BMC's manual-only approach wins
A tight ICP only pays off if the person working the list treats every account as a human relationship. Our specialists research every prospect by hand and message them individually across LinkedIn, email, Twitter/X, DMs, SMS and phone, which is why BMC clients see 8-13% reply rates while the industry sits at 2-4%.
A tight ICP sent an average message beats a loose ICP sent a brilliant message, every single time. ICP is the lever, messaging is the amplifier.
The 6-dimension framework
Here is the framework Bring My Customer runs for every single client. Six dimensions. Each weighted. Each with a specific score range. Sum them and gate outreach at 70+.
| # | Dimension | What it measures | Weight |
|---|---|---|---|
| 1 | Firmographic | Size, industry, geography, tech stack, revenue | 0-20 |
| 2 | Role & title | Buyer, champion, approver reachability | 0-15 |
| 3 | Pain profile | Evidence of the expensive problem right now | 0-15 |
| 4 | Trigger events | Recent change that opens the buying window | 0-20 |
| 5 | Buying signals | Observable intent and in-market activity | 0-15 |
| 6 | Fit score | Composite gate (sum of 1-5), 0-100 | Gate at 70+ |
Dimension 1: Firmographic (0-20)
Firmographic is the baseline, the stuff you can filter in a LinkedIn search in 30 seconds. It is necessary but completely insufficient. Teams that stop at firmographic match are the ones running 1.8% reply rates.
What to score
- Employee count. Pick a 2-3x range (e.g. 50-200). Not 10-10,000.
- Business category. Specific SIC or NAICS codes, not a vague label.
- Revenue band. Match your historical best-fit customers' revenue, not aspirational.
- Geography. Where can you sell, deliver, and comply? Only those regions.
- Tech stack. Presence or absence of specific tools correlates with buying readiness.
Score each account 0-20 based on how closely it matches your historical top 20% of customers, not your total customer base, your top quintile. Those are the accounts you want to clone.
Dimension 2: Role & title (0-15)
Firmographic match means nothing if the right person inside the account is unreachable. Dimension 2 scores how clearly the account exposes a buyer, a champion, and an economic approver you can get to via cold outreach.
The three roles to map
- Buyer. The person whose budget gets spent. Usually a VP or Director.
- Champion. The person who feels the pain and will shepherd you through procurement. Usually a Senior Manager or Director.
- Economic approver. The person who signs. Usually a C-level or senior VP.
Score 0-15: 15 if all three are identifiable and reachable (verified email + LinkedIn). 10 if two of three. 5 if only one. 0 if none can be identified. Below 10, archive the account.
Dimension 3: Pain profile (0-15)
The question here is simple: what expensive problem is this account living with right now that my offer solves? If you cannot name it in one sentence per account, you are shipping guesses.
How BMC scores pain profile
- 15: Account has publicly expressed the exact pain (LinkedIn posts, job descriptions, press interviews, podcast appearances).
- 10: Account has indirect evidence (hiring for a role that signals the pain, product gaps, lagging benchmarks).
- 5: Account structurally has the pain by virtue of their industry + size, but no fresh signal.
- 0: No evidence the pain exists at this account.
Dimension 3 is where most teams quit. It is manual, it does not scale cheaply, and it is also the single highest-leverage lift in reply rate. Our Bring My Customer specialists spend 40% of week-one time on this dimension alone.
Dimension 4: Trigger events (0-20)
Triggers are the most under-utilized dimension in B2B outbound. They answer the question: why now?
Firmographic match tells you an account could buy. Triggers tell you an account is likely to buy right now, this quarter, this month. Reply rates on trigger-tagged accounts run 2-3x higher than reply rates on firmographic-only accounts.
Triggers that score 15-20
- Funding round in last 90 days. Cash to spend, mandate to grow.
- Leadership change in your buyer role. New VP of Sales = tooling review within 60 days, 80% of the time.
- Hiring for a role your offer makes unnecessary. They are already feeling the pain.
- Product launch or expansion. New revenue goals, new systems needed.
- M&A activity. Integration = tooling consolidation = buying window.
Triggers that score 5-10
- New office or geographic expansion.
- Public RFP posted.
- Competitor churn mentioned on review sites.
- Technology deprecation (e.g. retiring an incumbent tool).
At BMC, triggers are spotted by hand: the specialist checks LinkedIn profiles, company "News" pages, podcast appearances, and recent press. No scraping tool reliably catches "the new VP of Sales mentioned their first 90-day priority on their third post", a human does. Fresh triggers (under 60 days) score highest. Stale triggers (over 180 days) do not count.
Dimension 5: Buying signals (0-15)
Buying signals overlap with triggers but focus specifically on observable, in-market behavior. Where triggers say "something changed," signals say "they are actively looking right now."
- Public evaluation activity. G2/Capterra category visits, review browsing, or buyer comments comparing vendors.
- SDR / BDR hiring. They are investing in outbound, they understand the game.
- LinkedIn activity. Buyer actively commenting on content in your category.
- Tech-stack changes. Just adopted or churned a competitor's tool.
- G2 or Capterra reviews. They are evaluating alternatives publicly.
Score 0-15 based on strength and freshness. Two or more live signals? 15. One live signal? 10. Historical signals only? 5. Nothing? 0.
Dimension 6: Composite fit score (gate at 70+)
Sum dimensions 1-5. Possible range: 0-85. (Dimension 6 is the gate itself, not an added weight.) Normalize to 0-100 if you prefer. Then set the gate: only accounts scoring 70+ enter the specialist's active book for manual outreach.
Everything below 70 goes to one of three buckets:
- 50-69: Hold list. Re-enrich in 60 days. Trigger events may push them into the 70+ tier.
- 30-49: Nurture list. Light-touch content, no cold outreach.
- Under 30: Archive. Do not touch.
Want BMC to build your 500-account ICP for you?
Every Bring My Customer engagement opens with a dedicated ICP build in week one. We score 500 named accounts across all 6 dimensions, enrich triggers, verify emails, and hand you a scorecard before any outreach starts, all included in the $320/month base.
Book Free Strategy CallHow to build the 500-account list
Here is the operational version. Below is a sample ICP scorecard for a hypothetical business, let's say a B2B revenue-intelligence business called "RevSignal" selling to Heads of RevOps at 100-500-person B2B companies.
Sample ICP scorecard: RevSignal
| Dimension | Criteria | Score band |
|---|---|---|
| Firmographic | B2B, 100-500 employees, $10-$75M revenue, North America, running Salesforce | 18-20 if all five match |
| Role & title | Head of RevOps + VP Sales + CRO reachable with verified emails | 15 if all three |
| Pain profile | Forecasting accuracy < 70% (per public earnings) or VP Sales turned over in last 180 days | 10-15 |
| Trigger events | Funding round in last 90 days OR new VP Sales hired OR Salesforce admin role posted | 15-20 |
| Buying signals | G2 category visits + hiring an SDR + LinkedIn RevOps content engagement | 10-15 |
| Composite | Gate at 70+ | Enter specialist's book |
The 7-step build process
- Day 1: Pull historical top-20% customers. Reverse-engineer the firmographic signature.
- Day 2: Build a candidate universe of ~2,000 accounts, hand-selected from LinkedIn, industry roundups, and buyer-adjacent writing.
- Day 3: Score dimension 1 (firmographic) on all 2,000. Keep top 1,000.
- Day 4: Score dimensions 2 (role) and 3 (pain). Keep top 700.
- Day 5: Overlay dimensions 4 (triggers) and 5 (signals), each spotted by the specialist. Keep top 500.
- Day 6: Confirm every contact by cross-checking LinkedIn and the company site. Discard anything that doesn't line up.
- Day 7: Score composite. Confirm 500 at 70+. Hand the list to the specialist who will work it by hand.
This is exactly the sequence a Bring My Customer specialist runs for every new client. It is boring, it is manual, and it is why reply rates post-launch sit at 9-13% versus 2-4% industry baseline.
Common mistakes that kill the ICP
- Building ICP from "any customer" instead of top-quintile customers. Your worst customers are not your ICP. They dilute the signal.
- Skipping trigger events. Firmographic match without triggers = 2% reply rate. With triggers = 9%+.
- Letting Sales define ICP without Marketing (or vice versa). ICP is cross-functional. Both teams must agree.
- Writing the ICP once and never revisiting. Markets move. Rebuild every 6-9 months. Audit monthly.
- Making the ICP too broad because "we don't want to miss anyone." You will miss them. Outbound is a sniper rifle, not a shotgun.
- Confusing ICP with messaging. ICP decides who gets the message. Messaging decides what the message says.
- Not scoring at all. If you cannot explain why Account A scored 82 and Account B scored 61, you do not have an ICP, you have a wishlist.
Frequently asked questions
What is an Ideal Customer Profile (ICP) in B2B?
An Ideal Customer Profile is a precise, scored definition of the type of company most likely to buy from you, close fast, retain long, and expand. It is not a buyer persona (that describes a human). An ICP describes an account on measurable dimensions: firmographics, role, pain, triggers, buying signals, and fit score. At Bring My Customer, every client gets a 500-account ICP scorecard in week one before any outreach is sent.
How is a B2B ICP different from a buyer persona?
A persona describes a person, "Marketing Mary, 35, VP of Marketing, cares about ROI." An ICP describes an account, "50-200-person B2B business in North America with a VP of Sales who was hired in the last 90 days." Personas help with messaging. ICPs decide whether you contact the account at all. B2B outbound lives and dies on the ICP, not the persona.
What are the 6 dimensions in BMC's ICP scoring framework?
Firmographic (size, industry, geography, tech stack), Role & Title (who buys, champions, signs off), Pain Profile (the expensive problem they're living with), Trigger Events (recent changes that make now the right moment), Buying Signals (observable intent data), and Fit Score (the 0-100 composite that gates outreach). Only accounts scoring 70+ enter the specialist's active book at Bring My Customer.
How tight should a B2B ICP be?
Tight. A good ICP produces 500-2,000 named accounts, not 50,000. If your ICP returns more than 5,000 companies, it is too loose. If it returns fewer than 300, it is too tight. The 500-2,000 range is where reply rates compound and messaging stays specific. BMC's internal benchmark across 200+ clients is 500 accounts scoring 70+.
How often should I update my ICP?
Review quarterly. Rebuild every 6-9 months. Markets move, your product evolves, and the ICP that worked in Q1 is usually not optimal by Q4. At Bring My Customer we also run a mid-cycle "kill list" every 30 days, the bottom-performing 20% of ICP segments get cut and replaced with new accounts.
What's the biggest mistake teams make building a B2B ICP?
They build it from their best customers' demographics instead of their best customers' trigger events. Firmographic match tells you "who could buy." Trigger events tell you "who is likely to buy right now." Teams that skip triggers end up targeting 10,000 technically-qualified accounts and getting 2% reply rates. Teams that add triggers target 500 right-now accounts and get 11%+.
Do I need a separate ICP for different products or segments?
Yes, if the buyer profile actually differs. One product, two buyer types (e.g. selling to Heads of Sales and Heads of RevOps) usually means two parallel ICPs with separate messaging, cadences, and reporting. Running them as one blended ICP suppresses reply rate on both. Bring My Customer splits any account where this is true from day one.
Can Bring My Customer build the ICP for me?
Yes. Every Bring My Customer engagement starts with a dedicated ICP build in week one, using the exact 6-dimension framework in this guide. We score 500 named accounts, enrich them with trigger events, verify contact data, and hand you the scorecard before any outreach starts. The ICP build is included in the $320/month base engagement.
Next steps
You now have the entire Bring My Customer ICP framework in your hands. You can print it, hand it to a BDR, and build your 500-account list this week. If your reply rate jumps from 3% to 9%+ in the next 30 days, that is the framework working.
Most teams, however, do not have 40 hours a week of senior labor to burn on an ICP build. They want the output without owning the process. That is why Bring My Customer bundles the ICP build into every engagement, the $2/hour specialist runs weeks 1-2 of the framework for you, hands you the scored 500-account list, and launches outreach by day 14. 2,000+ qualified B2B client meetings booked. 200+ businesses. Same ICP framework, every time.
If you want to see what the ICP would look like for your specific business, book a free 20-minute strategy call. We will map the 6 dimensions to your market, give you a rough estimate of your 500-account universe, and tell you honestly whether BMC is the right fit.