ABM Account Selection

Account selection is where ABM succeeds or fails. Bad accounts with perfect execution produce zero pipeline. Good accounts with mediocre execution still convert. Spend more time here than on the campaign itself.

The Selection Framework

Account selection runs through three filters, in order. Each filter is a gate. An account must pass all three.

An account that passes ICP Fit but has no signal goes to nurture. An account with a signal but no accessibility goes to inbound/content plays. Only accounts that pass all three enter ABM.

Filter 1: ICP Fit

Score every candidate account on firmographic and technographic fit. Binary pass/fail on hard criteria, weighted score on soft criteria.

Hard criteria (must pass all)

These are non-negotiable. If an account fails any one, it's out regardless of signal or intent.

• Company size: Within your serviceable range. Define by employee count, ARR, or both. "50-500 employees" or "$5M-$100M ARR"
• Geography: In a region you can sell to and support
• Industry/vertical: In a vertical your product serves. Adjacent verticals need a clear use case before inclusion
• Tech stack compatibility: Runs infrastructure your product integrates with. If they're on a stack you don't support, they can't buy

Soft criteria (weighted score)

Score each on a 0-3 scale. Total ≥ 7 to pass.

ICP fit rules

• Never include an account because "sales asked for it" without running it through the filters. Political accounts waste ABM budget
• Never include an account because of brand recognition alone. Logos don't pay back campaign cost
• Revisit hard criteria quarterly. Markets shift. Your serviceable range changes as the product evolves
• If more than 30% of candidate accounts fail ICP fit, the sourcing method is wrong. Fix the top-of-funnel list, not the filter

Filter 2: Signal Presence

A signal is a specific, recent, observable event that creates urgency. No signal = no reason to act now = no deal velocity.

Tier 1 signals (strong intent, act within 2 weeks)

• Funding round announced in the last 90 days. New money = new initiatives = new budget
• Key leadership hire: New VP Sales, VP Marketing, CRO, VP RevOps in the last 60 days. New leaders buy tools in their first 90 days
• Active evaluation: Visiting your pricing page, requesting a demo, engaging with comparison content, showing up in intent data from 6sense/Bombora/G2
• Competitor contract expiring: Known renewal window for a competitor. Rare to have this data, high-value when you do
• Tech stack change: Ripped out or added a tool adjacent to yours in the last 90 days. Signals the team is rebuilding that layer

Tier 2 signals (moderate intent, act within 30 days)

• Job postings for roles that use or buy your product (RevOps, Growth, Marketing Ops)
• Product launch or major release that creates new operational complexity
• Expansion signals: New office, new market, international launch
• Conference attendance or speaking at events in your space
• LinkedIn activity: Key contacts posting about problems your product solves

Tier 3 signals (weak, use for prioritization within a tier only)

• Company blog posts about relevant topics
• Glassdoor reviews mentioning relevant pains
• General industry tailwinds affecting the vertical
• Competitor mentions in their content

Signal rules

• Every ABM account must have at least one Tier 1 or Tier 2 signal before entering a campaign. Tier 3 alone is not enough
• Signals decay. A funding round from 12 months ago is not a signal. Define a freshness window for each signal type
• Document the specific signal for each account. "They're growing" is not a signal. "Posted VP RevOps role on March 15, still open" is a signal
• Signal stacking: accounts with 2+ signals get higher priority within the ABM list. A funding round + a key hire = move to the top

Filter 3: Accessibility

Can you actually reach the people who buy?

Accessibility criteria

Accessibility rules

• An account that passes ICP fit and has a strong signal but fails accessibility goes into a "build access" track: content, ads, community engagement. Not direct ABM outreach
• Don't confuse "I can find their email" with accessibility. If every email you find is a generic role-based address, the account isn't accessible enough for ABM
• Prior relationship is the single strongest accessibility signal. One warm intro outperforms 10 cold emails. Prioritize accounts where your team has a connection
• If your enrichment tools can't find contacts at an account, that's a data quality signal, not a reason to skip the account. Try LinkedIn Sales Nav, conference attendee lists, podcast guest lists before disqualifying

Tiering Selected Accounts

After all three filters, sort passing accounts into ABM tiers:

Tiering rules

• Tier 1 should never exceed 10 accounts. If more than 10 pass Tier 1 criteria, raise the bar. ABM quality degrades when the "top tier" is too large
• Tier 2 accounts must cluster. Group them by shared trait (vertical, stage, signal type) before assigning to a 1-to-few campaign. Random assortments of Tier 2 accounts aren't a cluster
• Tier 3 is the overflow. If an account barely passes all three filters, it's Tier 3. Run targeted outbound, not ABM. Be honest about the line between ABM and outbound-with-research

Data Sources for Account Sourcing

Where to find candidate accounts before running them through filters:

Sourcing rules

• Start with CRM data. Closed-lost and stalled opps are the highest-quality source because you already have context, contacts, and history
• Layer intent data on top of, not instead of, firmographic fit. Intent data alone produces low-quality lists because the signal is noisy
• Never build an ABM list solely from a purchased data set. Purchased lists have no relationship context and high data decay
• Refresh the candidate pool monthly. New signals appear. Old signals expire. A static list degrades within 60 days

Account Selection Cadence

Cadence rules

• Sales and marketing must agree on the list. A list built by marketing alone will be ignored by sales. A list built by sales alone will be too political. Joint ownership or it fails
• Remove accounts that completed a full ABM campaign with zero engagement. Don't recycle them immediately. Wait 6 months before re-evaluating
• Cap total active ABM accounts at a number your team can actually execute against. 10 accounts per ABM marketer for 1-to-1. 50 accounts per marketer for 1-to-few. Exceeding capacity dilutes every campaign

Anti-Pattern Check

• Selecting accounts based on logo prestige instead of ICP fit. Enterprise logos feel good in slides but burn budget if they can't buy
• Selecting 200+ accounts and calling it ABM. That's demand gen with a target list. Real ABM requires per-account investment that doesn't scale past 50-100 accounts
• Skipping the signal filter. "They fit our ICP" is necessary but not sufficient. Without a signal, there's no urgency and the campaign will produce meetings that don't close
• Letting one stakeholder (usually a sales leader) dictate the list without data. Political accounts are the #1 source of wasted ABM budget
• Never refreshing the list. Signals decay. Contacts leave. A list built in January is stale by April
• Running all accounts through the same motion regardless of tier. Tier 1 and Tier 3 require fundamentally different investment levels. Treating them the same wastes money on Tier 3 or under-invests in Tier 1
• Scoring accounts once and never re-scoring. An account that was Tier 3 six months ago may have a new signal that makes it Tier 1 today