How to Select and Tier Your ICP Accounts

You can't ABM everybody. You pick 50-500 accounts that are most likely to buy, most likely to generate revenue, and most likely you can reach.

But which ones?

Most companies define an ICP (Ideal Customer Profile) by firmographics: company size, industry, region. Then they filter down to 50k companies that match. Then they realize that's not a workable list.

Real account selection is more nuanced. You're not just looking for "companies that match our ICP." You're looking for "companies that match our ICP, where we can win, in a reasonable timeframe, with available resources."

This guide shows you how to select and tier your accounts.

Define Your ICP (Narrowly)

ICP is not "companies with 100-5000 people in tech." That's too broad.

ICP should be specific and defensible:

Good ICP definition: "B2B SaaS companies with 150-2000 employees, founded 2010-2020, based in North America, with dedicated go-to-market teams and annual revenue ops budgets over $500k."

Why this is good: - Company size: 150-2000 (specific, not "mid-market") - Customer type: B2B SaaS (specific, not "tech") - Location: North America (specific) - Maturity: Founded 2010+ (specific - rules out super early stage) - Use case indicator: GTM teams and revenue ops budget (shows they care about your problem)

Bad ICP definition: "Mid-market companies in tech."

Why this is bad: - "Mid-market" is 500-5000 people (too broad) - "Tech" includes hardware, developer tools, enterprise software, everything - No signal about your use case or budget

Your ICP should be narrow enough that you can actually list the companies, not so broad it includes 100k+ companies.

Identify the Universe

List all companies that match your ICP.

How many are there?

• If the answer is 500k+, your ICP is too broad. Narrow it.
• If it's 50k-500k, that's realistic. You can layer in scoring to narrow further.
• If it's under 5k, your ICP might be too narrow.

Example: "B2B SaaS in North America with 150-2000 people"

Google: ~15,000 companies. That's your universe.

Can't be that exact? Use a data platform (Apollo, ZoomInfo, Hunter) to refine: - Company size: 150-2000 - Industry: Software/Saas - Region: North America - Last funding: within 10 years

Refine to ~5-20k candidates. That's workable.

---

Score on Three Axes

For a realistic target list, you need 50-200 accounts (depends on your sales team size).

From 5-20k candidates, score on three axes. Each is 0-10:

Axis 1: Business fit (Does your product solve their problem?) - 10: Exact use case match (you've sold to similar companies, you know they care) - 7-9: Strong fit (they have the problem, they know it, they're actively solving it) - 4-6: Moderate fit (they might have the problem, but they're not focused on it) - 1-3: Weak fit (they have the problem, but it's not a priority) - 0: No fit

For a revenue ops software: A Series B SaaS with high sales churn = 10 fit. A manufacturing company with no sales team = 0 fit.

Axis 2: Budget signal (Do they have budget to buy?) - 10: Recent funding, public company with disclosed IT budgets, job postings for relevant roles - 7-9: Raising capital soon, parent company has disclosed budgets, strong revenue growth - 4-6: Profitable, growing, but no strong signal - 1-3: Bootstrapped, flat growth, unlikely to invest - 0: Struggling financially, lay-offs, no hiring

Funding: good signal. Hiring on your area: good signal. Lay-offs: bad signal.

Axis 3: Accessibility (Can you reach a decision-maker?) - 10: Warm intro from customer or partner - 7-9: Decision-maker on LinkedIn, likely to take your call - 4-6: Cold prospecting, decision-maker exists but hard to find - 1-3: Highly gatekept, lots of screening, hard to reach - 0: Impossible (government contract, hyper-regulated, no inbound possible)

Calculate Your Score

Score = (Business fit × 2) + Budget signal + Accessibility

Weight business fit double because if they don't have the problem, the other factors don't matter.

Example:

Acme SaaS: - Fit: 9 (perfect match, Series B, high churn) - Budget: 8 (just raised $20M) - Accessibility: 9 (warm intro from customer) - Score: (9 × 2) + 8 + 9 = 35

Beta Inc: - Fit: 5 (might need it, not priority) - Budget: 6 (stable, growing) - Accessibility: 5 (LinkedIn connection possible) - Score: (5 × 2) + 6 + 5 = 21

Gamma Corp: - Fit: 8 (strong fit) - Budget: 3 (bootstrapped, small) - Accessibility: 4 (hard to reach) - Score: (8 × 2) + 3 + 4 = 23

Acme (35) is your top priority. Gamma (23) is longer sales cycle. Beta (21) is even harder.

Tier Your Accounts

Rank all candidates by score. Divide into tiers:

Tier 1 (Top 25%): Score 25+ - These are your hot accounts - You should have 50-100 here (if you have 5k candidates) or 10-30 here (if you have 1k candidates) - Focus your ABM efforts here - You expect 30-40% of quarterly pipeline from Tier 1

Tier 2 (Next 25%): Score 15-24 - Good accounts but not hot - 50-100 accounts - You work these if Tier 1 is covered or if Tier 1 account goes cold - You expect 20-30% of quarterly pipeline from Tier 2

Tier 3 (Next 25%): Score 5-14 - Possible accounts but lower priority - 100+ accounts - You nurture these lightly (marketing automation, no sales outreach) - You expect <10% of quarterly pipeline from Tier 3

Tier 4 (Bottom 25%): Score <5 - Not a fit or not addressable right now - Archive them - No effort

---

Seasonal and Trigger-Based Re-Tiering

Accounts move between tiers as signals change.

Triggers that bump accounts up: - They just got funding announced - They posted a relevant job opening - They acquired another company - They changed leadership in your function - They just launched a new product/feature in your space

Example: Gamma Corp was Tier 3 (score 23). They just raised $30M. Update their budget signal from 3 to 9. New score: (8 × 2) + 9 + 4 = 29. Promote to Tier 2.

Triggers that bump accounts down: - Major lay-offs announced - They're going through bankruptcy or acquisition - They explicitly said "no budget" - They're locked into a competitor for 3+ years

Example: Acme had score 35, was Tier 1. They just signed a 3-year contract with your competitor. Update their fit to 0. Demote to Tier 4.

Review your tier list monthly. Signals change. Your list should reflect that.

Picking Your Initial Tier 1

For first ABM deployment, you don't start with 100 Tier 1 accounts. That's too much.

Start with 20-30 Tier 1 accounts that meet all three criteria: - You have a warm path (existing relationship, customer referral, partner intro) - They have clear budget signals (recent funding, hiring, growth) - They're actively in a buying process or will be in next 90 days

These 20-30 are your "hot 20." Spend the next 12 weeks winning 3-5 of them. Use the playbook you develop to scale to your next 30.

Account Selection Meeting (Quarterly)

Once a quarter, sales and marketing should sit down and update:

1. ICP definition: Did we learn anything? Does our ICP need to shift?
2. Tier 1 composition: Which accounts are we actually winning? Are they the ones we predicted?
3. Signals that matter: Which signals predicted wins? Which were noise?
4. Tier 1 refresh: Remove accounts we've lost. Add accounts that just became hot.

Example learnings:

"Our Tier 1 accounts in California are closing 30% faster than Tier 1 in Texas. Let's weight geography more heavily."

"Accounts with recent funding are closing 2x faster. Let's increase the budget signal weight."

"We thought manufacturing would be a fit, but we're 0/5 on them. Let's remove manufacturing from our ICP."

Use this quarterly meeting to improve your targeting.

---

Tools for Account Selection

• Data platforms: Apollo, ZoomInfo, Hunter (find companies matching ICP)
• Intent data: 6sense, Demandbase (identify active buying signals)
• LinkedIn Sales Navigator: Identify decision-makers in target accounts
• Spreadsheet or CRM: Score and tier accounts, track signals

Common Selection Mistakes

Mistake 1: ICP is too broad "Mid-market tech companies" = 500k companies. You can't ABM 500k companies. Define narrowly.

Mistake 2: You pick accounts for the wrong reasons Sales says "They're big, let's pursue." But big ≠ winnable. Stick to the scoring model, not gut feel.

Mistake 3: No re-tiering An account stays in Tier 1 for 6 months with zero engagement. Time to demote them.

Mistake 4: Ignoring accessibility You picked 100 Tier 1 accounts, but 80 of them are impossible to reach (heavily gatekept, no LinkedIn path). Accessibility matters.

Mistake 5: Wrong signals You thought "raised funding" was your signal. But it turns out "changed CRO" is the real signal for your product category. You're chasing the wrong metrics.

Bring It Together

Define a narrow ICP. Identify your universe (5-20k companies). Score on fit, budget, and accessibility. Tier into four buckets. Start with a hot 20. ABM that list hard. Win 3-5. Learn. Tier up 30 more.

Good account selection removes the biggest variable in ABM: whether you picked the right accounts.