TapCub is live — analytics, insights and live chat, free on one platform
Glossary

Account Intelligence: Definition, Data and a Worked Example

Account intelligence merges what you know about a company, such as visits, chats and product events, into one account view and score. Data types, how it differs from ABM and a CDP, and a worked health-score example.

Sophia Bennett Reviewed by Oliver Reed Also called account-level intelligence Last updated Nov 20, 2026 Glossary

Quick answer

Account intelligence is the practice of merging what you know about a company, such as visits, chats, product events and CRM data, into one account-level view and score. The unit is the organization, not the lead, so you can see whether it is engaged and what to do next.

Formula: Example health (0 to 100) = 40 x active members / members + 30 x active days / 30 + recency (30 if seen within 7 days, 15 within 30 days, else 0)

What is account intelligence?

Account intelligence is the practice of gathering what you know about a company in one place so you can judge whether it fits, whether it is engaged and what to do next. The unit is the account (the organization), not the individual lead. Instead of asking "what did this person click?", you ask "how many people at this company are active, what are they doing and is it getting better or worse?".

One B2B data vendor defines it as the aggregation of first- and third-party data, turned into insights that help marketing and sales understand accounts and leads. The definition is useful but it comes from a company that sells such data, so read the second half of this page as a method you can run with your own data, with or without a vendor.

The term appears in three places, and it helps to keep them apart:

  • Sales and marketing teams use it to prioritise target accounts: who to call, who to nurture, who to leave alone.
  • Customer success teams use the same idea after the sale, to see whether a customer organization is still using the product.
  • Product teams use it to see usage by company rather than by individual, because in B2B products the buyer, the users and the admin are different people.

What they share is a merge: signals from visits, conversations and product events are joined by an account key, and the account gets a status. A person-level view cannot say whether a customer is still using the product. A visitor who never logs in tells you nothing about the organization, and a single enthusiastic user can hide a team that has gone quiet.

What data goes into account intelligence?

Vendors list many data types; they fall into two groups, first-party and third-party, as in the same vendor's explainer. The table adds product usage and conversations, which matter for teams that sell software.

GroupData typeExampleWhere it comes from
First-partyWebsite behaviourPages viewed by people from the accountYour own analytics
First-partyProduct usageEvents by members of the accountYour own product tracking
First-partyConversationsChats and support threadsYour chat or help desk
First-partyCRM and sales engagementDeal stage, emails, callsYour CRM
Third-partyFirmographicIndustry, size, locationA data provider
Third-partyTechnographicTools the company usesA data provider
Third-partyIntentTopics the company is researching elsewhereA data provider
Third-partyNews and hierarchyFunding, parent and subsidiary linksA data provider

You can start with first-party data alone. It is the part you have permission to hold, you know where it came from, and it answers the most useful question: is this account using us, and is that changing? Third-party data adds fit and timing, but it also adds cost, accuracy questions and privacy duties.

Account intelligence vs lead intelligence, ABM and a CDP

TermUnitQuestion it answersRelationship
Account intelligenceThe organizationWhat do we know about this company and what should we do?The insight layer
Lead intelligenceThe individualHow interested is this person?Feeds account intelligence
Account-based marketingTarget accountsHow do we run campaigns for a short list of companies?Uses account intelligence to choose and measure
Customer data platformProfilesWhere is customer data collected and unified?Often the plumbing underneath

A short way to remember it: a CDP stores and unifies the data, account intelligence reads it at company level, and account-based marketing acts on it.

Worked example: merging one account's signals

Sample data: Acme is a customer with five identified members. Over the last 30 days its members produced the following, each tagged with the account key company:acme.

SignalCountDetail
Website visits14Pricing page 4, docs 6, blog 4
Chats3Two product questions, one billing question
Product events41Mostly reports opened and exports
Members active3 of 5Two members have not appeared in 30 days
Active days9 of 30
Last seen4 days ago

One merged view. Instead of five people's histories, you read one timeline: three active members, a billing question and two visits to pricing in the last week suggest a renewal conversation is near. None of the individual signals shows this on its own.

One score. To compare accounts, combine the signals into a number. A simple health score (0 to 100) uses three parts: members active in 30 days (40 points), active days out of 30 (30 points), and recency (30 points if last seen within 7 days, 15 within 30 days, 0 after that). For Acme:

  • Members: 40 × 3 ÷ 5 = 24
  • Active days: 30 × 9 ÷ 30 = 9
  • Recency: seen 4 days ago = 30
  • Health = 24 + 9 + 30 = 63

Compare Beta, another account with eight members, one active member, 2 active days and last seen 20 days ago: 40 × 1 ÷ 8 = 5, plus 30 × 2 ÷ 30 = 2, plus 15 for recency, so 22. Acme is healthy; Beta is drifting even though somebody there still shows up.

As a decision aid for this example only, not an industry standard: review any account below 40 this week, and look at accounts between 40 and 70 monthly. The thresholds should come from your own history: take accounts that churned and see what their score was 60 days earlier.

Two cautions about the score: it is a summary of behaviour, not a prediction, and it changes with the weights you choose, so keep the formula visible to the people who use it.

How to build account intelligence in five steps

  1. Pick the account key. Decide what an account is for your business: a company, a team, a store. Use a stable id, not a name.
  2. Identify people before you group them. An account is made of known people. Anonymous visitors cannot be assigned to an account by behaviour alone.
  3. Attach the key everywhere. Put the same account id on product events, chats and server events, so they merge.
  4. Define health with a small number of parts. Three to four signals you can explain are better than twenty you cannot. Write the formula down.
  5. Review it on a schedule and test it. Weekly for target accounts. Every quarter, check which accounts that churned had a low score beforehand; if none did, change the formula.

Privacy and what account intelligence cannot do

Account-level numbers about a company are not personal data, but the pieces that build them often are. Under the GDPR, personal data is any information relating to an identified or identifiable natural person, including by reference to an online identifier. Member lists, individual event histories and chat transcripts all qualify. Treat the member-level data accordingly: collect what you need, set a retention period, and be able to delete one person's data.

Account intelligence also has limits:

  • It cannot tell you who anonymous visitors are. If nobody is signed in, there is no account.
  • It is only as good as the account key. If the same company appears under three ids, you get three half-views.
  • It does not predict. A low score is a reason to look, not proof of risk.
  • Third-party data goes stale. Firmographic and intent data are snapshots; check how old they are.

Common mistakes

  • Scoring accounts on page views alone. Reading is cheap; use people active, days active and recency.
  • Counting members without checking identity. Shared logins make one person look like a team.
  • Mixing accounts and leads in one number. Keep the unit consistent.
  • Setting thresholds once and never revisiting them. Test the score against real outcomes.
  • Acting on a score without reading the timeline. The merged timeline is the explanation; the number is only the sort order.

Account intelligence in TapCub

TapCub's version is built from first-party data and gives you an organisation view; it does not supply third-party firmographic or intent data, so traits such as company name or plan come from what you pass in.

  1. Identify the user. After login, call tapcub.login, so the person exists in All-platform analytics › Users › People with a timeline of behaviour.
  2. Attach the account. Call tapcub.group('company', 'acme', { name: 'Acme' }). Later events in that session stay on the account; server events can carry the same account field. One person can belong to several accounts; anonymous visitors are never members.
  3. Open the account list. All-platform analytics › Users › Accounts shows the number of accounts, accounts active in 30 days and members. Open a row for 30-day activity, top events, members, traits and hierarchy; you can set a parent account, for example a store under a region.
  4. Read the health score. Sort by Last active, Members or Health. Health is 0 to 100 and is computed exactly as in the example above (40 for the share of members active in 30 days, 30 for the share of active days, 30 for last seen within 7 days or 15 within 30 days); it is blank if the account has no last-seen time.
  5. Ask questions at account level. Switch the funnel or retention identity to account when the question is about the organisation. In that view, rows without an account key are not counted, and you cannot open person-level details or add segments to the analysis.

Chats from identified visitors show who they are, so the same person's conversations and behaviour sit side by side. For setup details see analyze accounts with groups.

Sources

Was this explanation clear?

We use your answers to fix and expand this page.

Sophia Bennett

Product Adoption Consultant

Explores customer onboarding, live conversations, and self-service, turning common usage questions into practical guidance and improvements.

FAQ

Questions people ask

Short answers to the follow-up questions that come up most.

Still have a question?

Chat with the team behind TapCub — we usually reply within a few hours on business days.

What is account intelligence?

Account intelligence is company-level insight built by merging first-party signals (visits, chats, product events, CRM) and sometimes third-party data (industry, size, intent). It tells you whether an organization fits, is engaged and what to do next.

What is the difference between account intelligence and lead intelligence?

Lead intelligence scores one person, such as how many emails they opened. Account intelligence combines every identified person at one company, so a quiet champion and an active colleague show up as one account with one status.

What data do you need for account intelligence?

You can start with first-party data: website visits by identified people, product events, chats and CRM stage, joined by one account id. Third-party data such as firmographics or intent adds fit and timing but also cost and privacy duties.

How is an account health score calculated?

One simple version gives 40 points for the share of members active in 30 days, 30 points for the share of active days, and 30 points if the account was seen within 7 days (15 within 30). A 5-member account with 3 active members, 9 active days and a visit 4 days ago scores 63.

See this number on your own site

Add one script tag for web and apps, and the first reports fill in within minutes.

See your data clearly. Find your growth.

Every click, backed by data. Install one line of code and see your first numbers in a minute.

No credit card · Free plans for analytics and chat · Cookieless analytics