Account Scoring for ABM: How to Rank Target Accounts That Actually Close

Your sales team has a list of 400 target accounts and 12 working days a month to chase them. Without a way to rank that list, reps default to whatever moved last: the inbound form fill, the loudest stakeholder, the logo they recognize. Plenty of revenue sits in accounts nobody touched because they never raised a hand.

Account scoring fixes the ranking problem. Instead of scoring individual people, you score the whole buying group at a company, then point sellers and ad budget at the accounts most likely to buy and most worth winning. This guide walks through the three signal types that drive a useful score, a point system you can build in a spreadsheet this week, and how to turn raw numbers into tiers your team will follow.

A caveat before the math: scoring tells you where to spend attention, it does not close deals. Treat the model as a prioritization tool, recalibrate it every quarter, and expect to throw out your first version.

Account scoring vs lead scoring: why the unit changes

In a classic lead scoring model, the unit is one person. Someone downloads a guide, opens three emails, visits the pricing page, and their points climb until they cross a threshold and get routed to sales. That works for high-volume, single-decision-maker sales.

B2B deals with five to ten people in the buying group break that model. A VP who never fills out a form can be the economic buyer. A junior analyst who downloads everything may have no budget authority. Score people in isolation and you misread the account.

Account scoring rolls signals up to the company. You still track individual behavior, but the score that drives action describes the account: how well it fits your ideal profile, whether the buying group is showing intent, and how deeply people there engage with you. If you are weighing the broader strategic choice, our breakdown of ABM versus demand generation covers when account-level targeting earns its overhead.

The three signals every account score needs

A score built on one signal lies to you. Fit alone tells you who you want, not who wants you. Intent alone surfaces tire-kickers and competitors doing research. You need three inputs working together.

Fit: how closely the account matches your best customers

Fit answers a blunt question: if this account bought, would they stay, expand, and refer? Pull your closed-won deals from the last 18 to 24 months, look at the ones with strong retention and healthy margins, and find what they share.

Common fit attributes:

• Firmographics: industry, employee count, revenue band, region.
• Technographics: the tools they run that signal a fit or a gap (a CRM you integrate with, a competitor you displace).
• Business triggers: recent funding, a new executive hire in your buyer's function, a hiring spree for roles your product supports.

Fit is the most stable signal and the cheapest to maintain. A 600-person logistics company in a region you serve will still fit next quarter. Score it once, refresh when firmographics change.

Intent: signs the account is in a buying window

Intent data captures research activity that suggests an account is actively looking. It comes in two flavors. First-party intent is behavior on your own properties: repeat visits to a solution page, a pricing-calculator session, a demo request. Third-party intent comes from providers who detect surging research on relevant topics across the web and attribute it to a company.

Third-party signals are directional, not gospel. A spike in "warehouse automation software" reads from an account tells you someone there is researching, it does not tell you who or how seriously. Weight first-party intent higher because it is yours and it is verifiable. Our guide to intent data in B2B goes deeper on sourcing and accuracy.

Engagement: how the buying group interacts with you

Engagement is the account's relationship with your brand right now. Email replies, webinar attendance, multiple contacts from the same company active in a 30-day window, a sales call that happened. Roll these up: three people from one account each opening two emails is a stronger signal than one person opening six.

Recency matters more than volume here. An account that engaged hard six months ago and went dark is colder than one with modest but recent activity. Decay older engagement points so a stale account does not keep an inflated score.

Build the point model: a worked example

Here is a starting framework. The weights below are illustrative, your data should reset them, but the shape holds: fit and intent carry most of the weight, engagement breaks ties.

Signal	Example criterion	Points (illustrative)
Fit	Target industry	+20
Fit	Employee count 200 to 2000	+15
Fit	Runs a tool we integrate with	+10
Fit	Recent funding or leadership hire	+10
Intent	First-party: pricing page, 2+ visits	+25
Intent	Third-party topic surge	+15
Engagement	3+ contacts active in 30 days	+15
Engagement	Webinar or demo attendance	+10
Negative	Employee count under 20	-20
Negative	Student or personal email domain	-15
Negative	No activity in 90 days (decay)	-10

Negative scoring earns its place. Without it, every account drifts upward over time and your tiers lose meaning. Subtract points for disqualifiers (too small, wrong region, a competitor domain) and decay stale signals on a schedule.

Two rules keep the model honest. Cap each category so a single screaming-loud signal cannot dominate: an account with maxed-out intent but zero fit should not outrank a perfect-fit account showing early interest. And keep version one simple. A model with eight criteria you understand beats a 40-variable model nobody can explain to a skeptical rep.

From scores to tiers your team will act on

A number from 0 to 100 means nothing to a seller on a Tuesday. Translate scores into tiers tied to plays.

The flow most teams converge on:

Tier 1 (top 5 to 10% of the list): high fit and live intent. These get full one-to-one ABM: a named account owner, custom outreach, executive involvement, and dedicated LinkedIn Ads targeting aimed at the specific buying group.

Tier 2 (next 20 to 30%): strong fit, early or moderate intent. Run one-to-few plays grouped by industry or use case. Personalized at the segment level, not the individual.

Tier 3 (the rest of the qualified list): good fit, low intent. Keep them in programmatic nurture and ads until a signal lifts them. No rep time yet.

Set the cutoffs by capacity, not by round numbers. If each rep can genuinely work 15 accounts well, your Tier 1 across the team is the number of reps times 15. A threshold that floods sales with 200 Tier 1 accounts is a threshold that gets ignored.

Common mistakes that quietly break the model

Scoring people instead of accounts. If your system still routes on a single lead crossing a threshold, you are running lead scoring with an ABM label. Roll signals to the account.

No feedback loop. The model is a hypothesis. Compare scores against outcomes every quarter: do high-scoring accounts actually close faster and bigger? If your Tier 1 win rate looks like Tier 3, your weights are wrong. Closed-loop reporting makes this visible, and our piece on closed-loop reporting covers the plumbing.

Letting marketing own it alone. Sales has to trust the tiers or they will work their gut instead. Build the model with reps in the room, then revisit it together. This is the same sales and marketing alignment problem that sinks most ABM programs.

Confusing a high score with readiness to buy. A maxed-out score means "prioritize attention here," not "send a contract." Pair the score with sales judgment on timing and authority.

FAQ

How is account scoring different from lead scoring?

Lead scoring ranks individual people; account scoring ranks the whole company and its buying group. For B2B deals with multiple decision-makers, the account is the unit that matters, since one person's behavior rarely reflects the group's intent.

Do I need expensive intent data to start?

No. Your first-party signals (website visits, email engagement, demo requests) plus firmographic fit will carry a usable version one. Third-party intent adds reach into accounts not yet engaging with you, but it is an enhancement, not a prerequisite. Add it once the basic model proves itself.

How many tiers should I use?

Three is the common sweet spot: one-to-one, one-to-few, and programmatic. Two tiers can feel too blunt for a large list, and four or more tend to blur together and create routing arguments. Start with three and split only if a tier gets unmanageably large.

How often should I recalibrate the model?

Quarterly is a reasonable default. Review whether high-scoring accounts actually closed, retire criteria that no longer predict anything, and adjust weights against fresh closed-won data. Recalibrate sooner if you change your ICP or launch in a new market.

What is a good Tier 1 list size?

Match it to capacity. Multiply the number of reps by the accounts each can genuinely work (often 10 to 25 in true one-to-one ABM) and let that set your Tier 1 ceiling. Sizing by capacity keeps the list actionable instead of aspirational.

Can I run account scoring in a spreadsheet?

Yes, to start. A spreadsheet with fit, intent, and engagement columns and a weighted total is enough to prove the model and earn buy-in. Move it into your CRM or an ABM platform once the logic is stable and you need it updating automatically.

A short checklist before you ship version one

• Pull 18 to 24 months of closed-won data and define fit from your best, not your loudest, customers.
• Use all three signal inputs: fit, intent, engagement. Weight first-party intent above third-party.
• Add negative scoring and decay so scores stay meaningful over time.
• Cap each category so no single signal dominates.
• Translate scores into three tiers, each tied to a specific play and an owner.
• Size Tier 1 by sales capacity, then schedule a quarterly recalibration against real outcomes.

The first model you build will be wrong in places, and that is fine, because a rough ranking your team trusts beats a perfect one nobody uses. Ship it, watch which tiers close, and tighten from there.

If you want a second set of eyes before you commit, we can run a 30-minute review of your target account list and scoring logic, and point out the weights most likely to mislead your reps. Get in touch and bring your closed-won export.