Account engagement score: how to measure ABM by company
An account engagement score ranks companies by fit, intent and activity so you can measure ABM per account. What goes into it, how to weight it, how to use it.
An account engagement score is a single number per company that combines how well it fits your ideal customer profile, which pages it reads, and how much it interacts with your ads and website. It lets you measure ABM per account: which target companies are warming up, which are cold, and which sales should call now.
Lead scoring rates people. Account scoring rates companies, which is how B2B buying actually works: several people from one business research, compare and decide together. This article explains what goes into a good score, how to weight it, and how to use it without fooling yourself. For the reporting side, see company engagement reporting.
Why score accounts instead of leads
In most B2B deals, the people who research are not the same people who fill in forms. One person reads the blog, another checks pricing, a third looks at security documentation. Lead scoring sees three weak signals from three anonymous visitors, or none at all. Account scoring adds them up under one company.
| Lead score | Account engagement score | |
|---|---|---|
| Unit | One person | One company |
| Needs a form fill | Usually | No |
| Sees the buying group | No | Yes, combined |
| Fits ABM reporting | Poorly | Directly |
| Main risk | Ignores anonymous research | Over-weighting one busy visitor |
The three ingredients of an account score
Most account scores, whatever they are called, are built from three groups of signals.
1. Fit
How closely the company matches your ICP: industry, size, region, and whether it is on your target account list. Fit changes slowly. A company that is a poor fit should never reach the top of the list, however active it is.
2. Intent
Which pages the company reads, and what they suggest about buying stage. A visit to pricing, comparison or integration pages means more than a visit to a careers page. Intent signals can also come from campaign context: a click on an ad written for "alternatives to X" says something about where the account is.
3. Engagement
How much and how recently the company interacts: number of visits, return visits over several days, number of different people (where visible), ad clicks, and time on key pages. Recency matters. A burst of activity this week outweighs the same activity three months ago.
How to weight the score
There is no universal formula. The weights should come from your own deals. A practical way to set them:
- Take accounts that became opportunities in the last few quarters.
- Look at what they did in the weeks before: which pages, how many visits, from which channels.
- Compare with accounts that engaged but never became opportunities.
- Give more weight to the signals that separated the two groups.
Until you have that history, a simple starting model works. The example below is illustrative, not a recommendation for every business.
| Signal | Example points |
|---|---|
| On target account list, tier 1 | High fixed points for fit |
| Matches ICP industry and size | Medium fixed points for fit |
| Visited pricing or comparison page | High intent points |
| Visited product or integration page | Medium intent points |
| Visited blog only | Low intent points |
| Returned on a second day | Engagement bonus |
| Clicked an ad | Engagement bonus |
| No activity for 30 days | Points decay |
Cap the score, for example at 100, so one very busy visitor can't push an account off the scale. Apply decay so old activity fades.
From score to stage
A number on its own doesn't tell sales what to do. Pair it with a buying stage, inferred from the pages the account reads.
| Stage | Typical pages | What to do |
|---|---|---|
| Research | Blog posts, guides, glossary | Keep showing ads; offer useful content |
| Comparison | Comparison pages, alternatives, case pages, integrations | Show comparison-specific ads; sales prepares |
| Decision | Pricing, security, trial or demo pages | Sales follows up with context |
In IntentBid, every identified company gets a score from 0 to 100 built from fit, intent pages and engagement, plus a stage of research, comparison or decision based on the pages read. Accounts scoring 80 or more are flagged as hot so sales can see them first.
Where the signals come from
An account score is only as good as the data under it.
- Company identification. Most of the activity comes from website visits that never fill in a form. A pixel that identifies the company behind each visit from business networks provides the raw signals. Visits from home broadband, mobile data or VPNs often can't be matched to a company by anyone, so treat the score as a lower bound. See company identification.
- Ad platforms. Clicks carry campaign parameters back to your site. Microsoft Ads also reports impressions, clicks and spend by company, which shows exposure before any visit. See impressions per account.
- Your CRM. Open deals, closed customers and disqualified accounts, so the score doesn't keep surfacing companies you already know about.
- AI answers. Visits that arrive from ChatGPT, Perplexity or other AI assistants show buyers who shortlisted with AI first. AI visibility tracking shows which questions mention you.
How to use the score to measure ABM
The score becomes an ABM metric when you look at it across the target list, not just for one company.
- Share of target accounts engaged. How many companies on the list have any score above zero this month?
- Movement. How many accounts moved up a stage, or crossed your "hot" threshold, in the period?
- Score by tier. Tier 1 accounts should, on average, score higher than tier 3. If they don't, tier 1 work isn't landing.
- Score by channel. Which campaigns brought the activity that raised scores?
- Score to pipeline. Of the accounts that crossed the threshold, how many became opportunities?
ABM reporting shows these views by list, tier and period.
Common mistakes with account scores
- Counting page views without weighting. Ten blog visits aren't worth one pricing visit.
- No decay. Accounts stay "hot" months after they went quiet.
- Scoring everyone the same way. Fit should cap how high a poor-fit company can go.
- Ignoring internal traffic. Exclude your own office, agencies and partners, or they will top the list.
- Treating the score as truth. It is a ranking aid. Sales judgement and conversations still decide.
FAQ
What is a good account engagement score?
It depends on your model. What matters is the threshold that predicts opportunities in your own data. Start with a threshold, check how many accounts above it became opportunities, and adjust.
How is an account engagement score different from intent data?
Intent data usually means signals from outside your website, such as research activity on other sites. An account engagement score is built mostly from first-party signals: your website, your ads and your CRM. Many teams use both, with first-party engagement weighted higher because it is about you specifically.
Can I calculate an account score without a form fill?
Yes. That is the main reason to score accounts. Company identification links anonymous visits to a company, so the score can rise before anyone fills in a form.
How often should account scores update?
Daily or faster for engagement and intent, so sales sees activity while it is fresh. Fit changes less often, so it can be refreshed when the target account list or ICP changes.
Should sales see the account score?
Yes, with the context behind it: which pages, which campaigns, which stage. A score without context is easy to ignore.
See an engagement score for every company that visits your site. Start a 14-day free trial, no card needed.