Account scoring ranks companies, not individuals, by how worth pursuing they are. It combines fit, how closely a company matches the ideal customer profile, with timing, signs that it may be ready to buy, into a number or tier that tells sales and marketing where to spend effort first.
How a score is built
Most models have two parts. Fit is stable: industry, size, location and the technology a company runs. Timing moves: a funding round, a burst of hiring in one function, a new tool, a rise in interest on a topic.
Keep the two apart and show both. A grid of fit against timing is easier to act on than one blended number, because a good fit with no timing is an account to watch, while a lively account that does not fit is rarely worth the effort. Give timing signals a decay, so that last week counts for more than last quarter.
Testing a score
A score is a hypothesis until it meets outcomes. Group past accounts by score band at first contact and compare win rate, cycle length and deal size between the bands. If the top band does not win more often than the middle, the model predicts nothing: at Acme Robotics (illustrative), a top band that wins 24 percent of the time against 22 percent for the middle band is too close to act on. Repeat the check when the product, the market or a data source changes, and keep the reasons visible: a rep who cannot see why an account ranks high will ignore it.
Scoring with Fokals data
Intent Scores supplies timing: a weekly 0 to 100 score per company and topic, built from the dated signals of the trailing 90 days, with a surge flag and the five strongest signals as evidence. Company Signals holds every signal with its kind, topics and weight. Fit comes from Technology Stack, Website Profile and Hiring Activity, joined on the company ID.
Each score is evidence-backed and names the dated signals behind it, so a rep or a model reviewer can open the source and test the score against your own outcomes. The guide to prioritising accounts with intent scores shows a weekly routine, and the intent dataset lists the topics.
Related terms
Account scoring uses an ideal customer profile for fit and an intent surge for timing. Its person-level counterpart is lead scoring. The inputs are described under intent data.
Frequently asked questions
What is the difference between account scoring and lead scoring?
Account scoring rates a company, using attributes of the organisation (industry, size, technology) and signals about it (hiring, announcements, intent). Lead scoring rates a person's record, using their role and behaviour such as email replies and page visits. They work well together: the account score says whether a company is worth pursuing, and the lead score says which contact there is active now.
How often should account scores be refreshed?
As often as the signals move and as often as the team acts on them. Weekly scores suit a weekly planning meeting, and daily signals such as a new job posting or announcement can justify a daily refresh for the accounts a team is working. A score refreshed less often than it is used hides recent changes, and one refreshed more often than its sources change adds nothing.
What data do you need for account scoring?
You need past accounts with outcomes (won, lost, churned) and their dates, so the score can be tested. You need attributes for each company that you can repeat, such as industry, size, location and technology in use. And you need dated signals, such as hiring, announcements and intent scores, held with the time each was observed, so that you can score an account as it stood on the day you first contacted it.