Account scoring: how fit and intent decide who to reach first
How to score accounts on ICP fit and intent strength, why rule-based lead scores fall short, and how AdsGTM tailors scoring and priority tiers to your offer.
Dag Holmen · Updated · 7 min read
Account scoring ranks every company in your market on two things: how well it fits your ideal customer profile, and how strongly it’s showing buying intent. Fit tells you whether a company should ever buy from you. Intent tells you whether it’s looking now, and together they decide who sales calls first and who marketing warms up.
Most scoring goes wrong by mixing the two into one number, or by scoring only the people who already clicked your emails. A good account score keeps fit and intent separate, looks at the whole buying group and covers every company in your TAM, including the ones that have never heard of you.
Account scoring vs lead scoring
Lead scoring gives points to individual people, usually for who they are (title, company size) and what they’ve done with your marketing (opens, clicks, form fills, page views). It lives in your marketing automation tool and produces MQLs.
Account scoring rolls signals up to the company. It asks whether this business fits, and whether the business as a whole is moving toward a purchase.
The case for scoring accounts is simple: companies buy, not individuals. A Gartner survey published in May 2025 puts B2B buying groups at between five and 16 people, across as many as four functions. A lead score looks at one of them.
It also misreads the signals. Forrester makes the point that individual lead scoring misses something important: three people in the same buying group each taking a single action is worth more than one person taking several. Activity spread across an account is a stronger sign of a real project than one keen individual.
| Lead scoring | Account scoring | |
|---|---|---|
| Unit | One person | One company and its buying group |
| Usual inputs | Title, engagement with your emails, forms and site | ICP fit, plus intent across the web and across people |
| Who it can see | People who already engage with you | Every company in your TAM |
| Output | MQLs | Tiers and priorities for sales and marketing |
You still need contact-level detail. Once you know which accounts matter, you need to know which people to reach there. The score points to the account, and the contacts give reps someone to talk to.
What is rule-based scoring?
Rule-based scoring adds and subtracts points by fixed rules. A VP title might earn 15 points, an email click 5, a demo request 30, and a personal Gmail address minus 20. Cross a threshold and the lead is passed to sales.
It’s easy to set up and easy to explain, which is why most teams start there. It has three common weaknesses:
- The weights are guesses. Someone decided an email click is worth five points. Few teams go back to test it against closed deals.
- It rewards activity, not buying. Students, job seekers and competitors can rack up scores fast.
- It only sees people who already found you. A company researching your category on other sites, but never yours, scores zero.
Lead grading is a related idea some teams use to fix the first problem: a letter grade for fit (A to D) kept separate from a points score for engagement. That separation is the right instinct. Account scoring takes it further by replacing “engagement with us” with intent across the market.
The fit score
Fit is the part of the score that rarely changes, and it’s what people usually mean by ICP scoring. It answers “if this company were buying, would we want them?”
Typical fit criteria:
- Firmographics: industry, employee count, revenue band, region and funding stage.
- Technographics: tools the company uses that make you a better or worse fit, such as a CRM you integrate with.
- Buying group coverage: whether the company has the roles you sell to, and how many of them you can reach.
An example ICP might be B2B SaaS, 50 to 1,000 employees, North America and the EU, Series A to C, using HubSpot, selling to heads of marketing, demand gen, RevOps and CMOs. Every company in the TAM gets scored on how closely it matches.
Build fit criteria from your won deals rather than your hopes. The ideal customer profile guide covers how, and firmographic data explains each field.
The intent score: strength, volume and type
Intent is the part that changes week to week. In AdsGTM builds, it’s scored on three things.
Strength. How clearly a signal points to buying. Time on a competitor’s pricing page says more than a read of a general blog post.
Volume. How many related signals show up around the same topic. One article on “ABM” is curiosity. Research across ABM platforms, ad targeting and intent data over two weeks, from several people at the same company, is a project.
Type. Informational intent is learning about a problem: “what is account-based marketing”. Commercial intent is shopping for a solution: comparisons, pricing, alternatives and competitor engagement. Combinations of the two at one account often mean a company has moved from learning to evaluating.
| Signal pattern at an account | What it usually means | What to do |
|---|---|---|
| A single informational signal | Early curiosity | Keep in Marketing-ready and keep warming |
| Several informational signals on related topics | Getting to grips with the problem | Teaching content and ads, then watch on the next refresh |
| Commercial signals: comparisons, pricing, competitors | Evaluating vendors | Sales-ready: reach out |
| Informational and commercial signals across several people | An active buying project | Top priority for sales |
Hiring news and funding rounds aren’t in this table on purpose. They describe a company’s situation, and they can belong in fit. They don’t show anyone researching what you sell. The intent data guide explains why.
Combining fit and intent into priority tiers
Keep the two scores separate and read them together. A simple grid does most of the work.
| High intent | Low intent | |
|---|---|---|
| High fit | Call now. These are your Sales-ready accounts. | Warm up. Your best Marketing-ready accounts. |
| Low fit | Check first. Either noise or a sign your ICP is too narrow. | Leave out or keep to low-cost channels. |
From there, tiers turn the grid into something your CRM can sort. An AdsGTM list scores and tiers every company and gives every contact a priority, so a rep can filter to priority A Sales-ready contacts and start working, while marketing builds audiences from Tier A Marketing-ready accounts.
The low-fit, high-intent box deserves a look each quarter. If a segment you excluded keeps showing strong intent, and you win deals there, your ICP might be wrong.
How to build an account scoring model
If you’re building your own, keep it small and test it against reality.
- Pull your history. Export closed-won and closed-lost deals from the last 12 to 24 months.
- Find what separates winners. Compare won and lost accounts on size, industry, region, tech and who was involved.
- Pick five to eight fit criteria. More than that and nobody can explain the score.
- List your intent topics. Include your category, the problems you solve and your main competitors. Mark each as informational or commercial.
- Set tier thresholds. Decide what fit and intent levels make an account Tier A, B or C.
- Back-test it. Would the model have ranked your past wins near the top? If not, adjust the weights.
- Review it every quarter. Markets shift, and so does your offer.
Does AI lead scoring help?
Machine-learning scoring trains on your CRM history to find the patterns behind won deals. It can work well when you have lots of closed deals and clean data.
Many companies with 10 to 200 staff have neither. With a few dozen wins a year, a model has little to learn from, and it can’t see the companies you’ve never talked to. For most of these teams, a clear rules-based fit score plus good intent data gets further than a black box, and it’s easier to trust when a rep asks why an account is Tier A.
How AdsGTM tailors the score to your offer
Scoring only works if it reflects what you sell and who you want. AdsGTM tailors the model to your ICP, your offer and your priorities, agreed on the 30-minute kickoff call.
That means the intent topics come from your category, the problems you solve and the competitors you meet in deals, not a generic topic list. It also means the fit criteria match the customers you want more of, and your priorities count: if one segment matters most this year, the scoring reflects that. Three revisions are included, so you can adjust the model once you see the list.
The result lands in your CRM with fields for intent topics, list and priority on every contact. The Sales-ready list goes to your reps, and the rest becomes your Marketing-ready TAM.
Frequently asked questions
- How is a lead score calculated?
- Most lead scores add points for fit attributes, like title and company size, and for engagement, like email clicks and form fills, then subtract points for poor fit. When the total passes a threshold, the lead goes to sales.
- What is a scoring model in B2B marketing?
- A scoring model is the set of rules or weights that turns data about a company or person into a priority. Good models keep fit and intent separate and are tested against past won deals.
- What is lead grading?
- Lead grading gives a letter grade for how well a lead fits your ideal customer, kept separate from a points score for engagement. It helps sales tell a good-fit, quiet lead from an active, poor-fit one.
- What is the difference between a fit score and an intent score?
- A fit score measures how closely a company matches your ICP and rarely changes. An intent score measures how actively it is researching what you sell right now, and it can change from week to week.
- How often should account scores be updated?
- Fit criteria can be reviewed quarterly. Intent goes stale within weeks, so the intent side needs fresh data, and a delivered list should be refreshed when you need a new read on who is in-market.
Sources
Related guides
- Sales-ready TAM: in-market accounts, and how they differ from MQLsWhat makes a contact Sales-ready, how intent-based readiness compares with MQLs, SQLs and BANT, and why these accounts should get your reps’ first hours.
- Intent data and buying signals: what they are and where they come fromWhat intent data is, the buying signals that actually predict purchases (web research, competitor engagement, search), and the generic signals that don’t.
- How to define your ideal customer profile (ICP)A practical way to define your B2B ideal customer profile from your best customers and won deals, with an ICP template and how it differs from a buyer persona.
