Build the Tier 1 account list your ABM motion deserves

What you’ll do

You'll mine your closed-won corpus — including call recordings — for the fit variables that actually predict your wins, then score fit across your whole account universe with strict cost discipline. You'll build an engagement score from first-party signals with a 90-day decay, and set an order-of-magnitude contract value per account from your most similar closed-won lookalikes. Then you'll route everything through the engagement-by-fit 2x2: remarketing for the low/low quadrant, velocity sales for hot-but-smaller accounts, marketing-only warm-up for high-fit cold accounts (never cold calls), and full joint ABM for the high/high Tier 1s. Every score component lands in the CRM as a visible field with a tier reason, changes surface as a ranked report instead of chat noise, and accounts graduate between segments automatically as engagement rises.

The steps

1. 01
   Derive your fit variables from closed-won, not from opinion

   Week 1

   Before you score anything, figure out what actually predicts a win for you. Pull your closed-won opportunities from the last few quarters — and your closed-lost — enrich each account, and pipe in call recordings where you have them. Then run an AI analyst pass over the table with one question: what five features do the winners have in common? The answer is your fit model. Teams that do this exercise are routinely surprised: the variables that surface are rarely the ones the sales team would have guessed.

   • Fit variables are company-specific, not generic. A physical-security vendor scores store count; a payroll company scores the number of jurisdictions a prospect operates in; a fintech card scores funding events. Industry and headcount are table stakes — your three defining variables are the ones only your closed-won data can reveal.
   • Call recordings are the underused input. Win/loss reasons buried in transcripts ('we're replacing X', 'the new VP wanted this') become minable features once an LLM reads them at scale.
   • Mine humans too: ask your best sellers 'given fifty accounts, which do you call first, and why?' Meeting-takers also volunteer their trigger — a new exec hire, a fresh raise, a market entry. Write those down as candidate variables.
   • Spend more time on why you won than why you lost. Loss reasons are noisy (budget, timing, politics); win patterns are the statistical signal an AI pass is genuinely good at finding.
2. 02
   Score fit across the whole universe — with cost discipline

   Week 1-2

   Account fit is the static score: 'always true' facts about a company that no team owns and no campaign changes. Industry and motion match derived from your closed-won patterns, total headcount and sales-team headcount specifically, presence of the functions that buy you, a region you can support post-sale, funding, and technographics. Score it across every account you can see — but never enrich everything. Pre-filter the obvious non-fits on cheap fields first, run expensive enrichments conditionally, and only pay again when a value actually changed.

   • The order of operations is the whole cost model: employee-count floors and business-model checks are nearly free, so use them to cut a six-figure account universe down to an enrichable set before any deep research runs.
   • Public job descriptions are a technographics goldmine. A company hiring someone to administer a specific tool tells you their stack; a services firm's project postings tell you their client mix. All of it is free to read and minable at scale.
   • Fit data barely moves. Business model never changes; headcount and revenue are worth refreshing about twice a year; funding checks should only cost you anything when there's actually a new round. Write every result back to the CRM so you never re-run a build you already paid for.
   • Fit is a fact, not a preference. That framing matters later: when sales and marketing argue about an account, the fit score is the part neither side gets to negotiate.
3. 03
   Build the engagement score — and let it decay

   Week 2

   Engagement is everything fit is not: fleeting, first-party, and owned by marketing. Trial signups, webinar joins, demo requests, event attendance tracked as CRM campaigns, web traffic aggregated to the company level, social mentions of your category terms. The one rule that makes it honest: no new touches for roughly 90 days and the score decays to zero. An account that downloaded a whitepaper last spring is not engaged — a score without decay says it is, and your reps will burn mornings on ghosts.

   • 'Surging' is the trigger, not the absolute score: a period-over-period delta from zero to something is worth more attention than a big number that has been flat for a month.
   • One sharp example of a signal worth wiring in: an open job req for the role your product serves. The role existing on the org chart is a fit signal — the role being open right now is timing. Same data source, two different scores.
   • Track engagement down-funnel too. A late-stage deal gone quiet on email while account engagement stays high means they're still evaluating — that's an exec-to-exec nudge, not a closed-lost.
   • Keeping engagement separate from fit is the sales/marketing alignment mechanism. Fit is a fact; engagement is marketing's number to raise on good-fit accounts. That turns the attribution war into a data challenge, not a trust challenge.
4. 04
   Set potential contract value from closed-won lookalikes

   Week 2-3

   The third score answers one question: if this account buys, is that a $5K deal, a $50K deal, or a $500K deal? Order of magnitude only — precision is a trap here. Pull the twenty-five or so closed-won accounts most similar to the prospect on your fit dimensions (headcount, industry, funding stage, region) and read a percentile of their real contract values. That number decides where your expensive motions go and holds the SDR team accountable for spending calories on six-figure-shaped accounts instead of comfortable small ones.

   • This is internal-only math, deliberately rough. Never anchor a prospect — or your own sellers — with it. Prospects surprise you upward, and a CFO who hears your internal estimate will negotiate against it.
   • The funnel math from one of the sharpest GTM teams we track, running exactly this motion: 250,000 CRM accounts scored down to roughly 970 Tier 1 accounts per quarter, then cut to about 300 ABM targets by the contract-value threshold. That's the list an ABM budget can actually cover.
   • A simplified starter version — a formula over employee growth and revenue range — is enough to begin. Upgrade to true lookalike percentiles once your closed-won corpus is big enough to be a sample.
5. 05
   Route every account through the 2x2

   Week 3

   Plot engagement against fit and four plays fall out. Low fit, low engagement: remarketing and self-serve only — not bad accounts, just unaffordable ones. High engagement but lower fit: velocity sales — automation-first, remove friction, make it easy to buy, humans only on calls and meetings. High fit but cold: marketing-only warm-up, and this is the rule teams break most — never cold-call these. Community, events, education, awareness ads whose CTA is a useful conversation, never 'get a demo.' High fit and high engagement: that's Tier 1 — joint SDR-plus-marketing ABM, multi-threading five to six buying personas per account.

   • The high-fit/cold quadrant is where discipline pays. A dialer pointed at a cold whale burns the account that could be your biggest customer in a year. Warm it until it surges, then sell.
   • Velocity accounts get pre-built, templated outreach — SDRs shouldn't hand-craft emails to accounts the model says are opportunistic. Save the craft for Tier 1.
   • Nobody signs up for a webinar alone and then buys a quarter-million-dollar contract. Tier 1 treatment means deliberately creating hand-raisers across the whole buying committee, not working one champion.
   • Watch for brand-name gravity: left unmanaged, sellers work recognizable logos instead of engaged accounts. The 2x2 exists precisely to override that instinct.

   ↳ Reference: The ABM gifting playbook this Tier 1 list feeds
6. 06
   Push every score component into the CRM — no black boxes

   Week 3-4

   A score sales can't interrogate is a score sales won't trust. Every component — each fit variable, each engagement channel, the contract-value estimate — lands in the CRM as a visible field, alongside a written tier reason a human can read: 'large sales team, heavy ops function, recent raise, supported region.' And humans own the weights. Let AI propose a baseline weighting from the data — that's a genuinely good starting point — but you do not ship a vibe-coded account score to production. Review it, argue with it, freeze a version you trust.

   • The debugging loop is staring at incongruities: an account with a huge sales team and a low fit score means either the model is wrong or the account is genuinely unusual. Every one you resolve makes the weights better.
   • Expect to tweak weights frequently during the build, then freeze a baseline and stop. After that, only the underlying data refreshes — a score that changes meaning every week is as useless as a black box.
   • This is also your escape hatch from opaque intent platforms: when a bought score breaks, you can't see why. When your own score breaks, the broken component is a visible field you can fix.
7. 07
   Surface changes as a list, and graduate accounts automatically

   Ongoing

   The output of this system is not a stream of alerts — it's a ranked report your team reads on a rhythm. A daily or weekly CRM view sorted by fit and engagement, with reps validating the components behind each surge before they act. Chat alerts get noisy fast, and strong signals drown under weak ones; reserve pings for the rare, unambiguous triggers. Meanwhile the segments maintain themselves: fit is static, engagement is the mover, so accounts auto-graduate from the warm-up pool into worked patches the moment their engagement crosses the line.

   • The rep workflow off the list: back out accounts with open opportunities, sort by fit plus engagement, validate the surge by reading the component fields, then hydrate the account with decision-makers across five to six personas and multi-thread.
   • Because marketing carries awareness in cold territories, volume stops being the crutch — teams running this motion cap sends per rep per day and see deliverability problems disappear, because every send is aimed at an account the model already justified.
   • Define your event-driven exceptions up front: a product-qualified signal at an enterprise-ready account can route straight to a rep regardless of quadrant. Exceptions are fine — silent exceptions are not.
   • The graduation loop is what makes this a system instead of a one-time sort: this quarter's marketing-only accounts are next quarter's Tier 1, automatically, with no re-scoring meeting required.

What goes wrong

The failure modes that catch most founders.

• You blend fit and engagement into one magic number

  One blended score hides everything you need to know: a hot-but-tiny account and a perfect-fit-but-cold account can score identically and demand opposite plays. Worse, blending re-ignites the sales/marketing trust war, because nobody can see whose number moved. Keep the scores separate; the separation is the alignment mechanism.

• You buy a black-box score sales can't interrogate

  When an opaque intent score stops working, you can't see why — and long before that, your sellers will quietly stop believing it. Every component should be a visible CRM field with a written reason. If a rep can't answer 'why is this account Tier 1?' by reading the record, the score isn't real.

• You enrich everything before you filter anything

  Running deep enrichment across your entire CRM is unreasonably expensive and mostly wasted — a huge share of any account universe fails a free employee-count check. Cheap filters first, conditional enrichment second, write-back always, refresh only when values change. The scoring formula itself should cost nothing.

• You let engagement scores live forever

  Without decay, engagement becomes an archaeology of every touch an account ever had, and reps chase companies whose interest died two quarters ago. No new activity in about 90 days means the score goes to zero. The decay is what makes 'surging' a real, actionable state instead of noise.

• You pipe every signal into chat

  Alert channels degrade fast: weak signals bury strong ones, the team mutes the channel, and the system dies of noise. A stacked report read daily or weekly beats real-time pings for nearly everything. Reserve alerts for the handful of triggers strong enough that you'd want a phone call about them.

• You cold-call high-fit accounts that have never heard of you

  The cold whale is the most tempting mistake in the quadrant map. A premature cold call converts at rock bottom and burns an account that, warmed properly, could be your largest customer next year. High-fit/cold accounts get marketing-only treatment — community, education, awareness — until their engagement says otherwise.

Want the technical depth?

The chapters with the full reference detail.

• → ABM gifting and direct mail playbook— The campaign this Tier 1 list feeds
• → Gifting & Direct Mail reference— The full-motion chapters for your Tier 1 quadrant
• → Post-raise outbound playbook— Funding events as a fit and timing signal