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

Data Intelligence: Definition, Process and How to Test It

Data intelligence turns data into conclusions people can trust and act on. Two definitions reconciled, how it differs from analytics and BI, and a worked example that closes the loop.

Xu An Reviewed by Lucas Hayes Also called data intelligence platform Last updated Nov 19, 2026 Glossary

Quick answer

Data intelligence is the practice of turning data into conclusions people can trust and act on: understanding what data exists, where it came from and its quality, then using it to make a decision and checking the result.

What is data intelligence?

Data intelligence is the practice of turning data into conclusions that people can trust and act on. The term has no single agreed definition, and published definitions emphasise different halves of it:

  • The understanding half. IBM describes data intelligence as combining data and metadata management with tools such as AI and machine learning, to help an organisation "understand how enterprise data is produced and used". Its core functions are metadata management, data lineage, data governance, data quality and data integration. Here the question is: what data do we have, where did it come from, and can we trust it?
  • The decision half. Other vendors describe the same word as the process of collecting, preparing and analysing data, generating insight and then acting and learning from the result. Here the question is: what should we do, and did it work?

This page uses both: data is only "intelligent" when you can trust it and it reaches a decision. A dashboard nobody trusts, or a trusted number nobody acts on, is not data intelligence.

Data intelligence compared with analytics and BI

Question it answersTypical outputWho uses it
Data analyticsWhat happened, and why?Charts, models, a findingAnalysts
Business intelligenceHow are we doing against targets?Dashboards, scheduled reportsManagers
Data intelligenceCan we trust this, and what should we do about it?Trusted, documented metrics tied to an actionEveryone who decides with data
Product intelligenceHow do people use our product, and what should we change?Behaviour insight for product teamsProduct, growth

IBM describes analytics and business intelligence as "ways of using" data, with data intelligence supporting both by helping people find and understand the right datasets. The difference is a matter of scope, not a ranking: analytics and BI are consumers of data intelligence.

The five steps of a data intelligence loop

  1. Collect and integrate. Bring data from sources into one place with consistent definitions.
  2. Clean and describe. Fix duplicates and gaps; record what each field means, who owns it and where it came from (metadata and lineage).
  3. Analyse. Look for patterns and causes with queries, funnels, segments or models.
  4. Turn the result into a recommendation. A finding is not an insight until it names a change: "move the shipping cost to the cart page".
  5. Act and measure. Make the change, then check whether the metric moved. The measurement feeds back into step 1.

Steps 1 and 2 are the "trust" half, and 4 and 5 are the "decision" half. Teams often stop at step 3.

A worked example: closing the loop

Sample data: an online shop sees 10,000 visitors in a two-week period.

Funnel stepUsersRate from previous step
Visited10,000
Added to cart4,2004,200 ÷ 10,000 = 42%
Started checkout1,2601,260 ÷ 4,200 = 30%
Paid630630 ÷ 1,260 = 50%

Insight. The largest relative loss is between cart and checkout: 70% of carts do not start checkout. Before acting, check the data: are the checkout-start events arriving on every device and browser? If 3% of mobile sessions never send the event, part of the "loss" is measurement. That is the trust half.

Action. Show the shipping cost on the cart page. Result after another two weeks: 36% of carts start checkout, with the same 50% payment rate.

MetricBeforeAfterCalculation
Carts4,2004,200
Checkout starts1,2601,5124,200 × 36%
Paid orders6307561,512 × 50%
Orders per 10,000 visitors630756
Change+126 orders, +20%126 ÷ 630

The loop closed because the final row is a number the team decided in advance to look at. If the visitor count or season also changed, this before-and-after comparison cannot separate the effects; a holdout or experiment would.

What data intelligence needs underneath

NeedWhat it meansWhat goes wrong without it
Shared definitionsOne written meaning for each metric, such as "active user"Two teams report two numbers for the same thing
Quality checksRequired fields present, volumes in rangeDecisions rest on missing or duplicated events
LineageWhere a number came from and what changed itNobody can explain why a figure moved
Ownership and accessA named owner per dataset, with rules; see data governanceSensitive data spreads, or useful data is locked away
A route to actionA person or system whose job is to act on the findingInsights pile up in a slide deck

A five-question test for your own team

Pick one metric you report every week and answer these in order. Each "no" is where the loop is open.

  1. Is its definition written down, with an owner?
  2. Can you say where the number comes from and what would change it?
  3. Is there a check that would tell you if the underlying events stopped arriving?
  4. Is a named person expected to act when it moves beyond a set range?
  5. After the last action, did anyone compare before and after?

Four or five "yes" answers means the metric is data intelligence in the sense used here. Two or fewer means it is reporting, which is useful but is only one half of the loop.

Common mistakes

  • Buying tools before defining questions. A catalog or an AI layer does not tell you which decision the data should support.
  • Treating AI answers as facts. An answer is only as good as the data and definitions behind it.
  • Stopping at the dashboard. If no one is accountable for an action, the loop is open.
  • Skipping the quality check. A confident insight built on missing events is the most expensive kind.
  • Calling it data intelligence when it is reporting. Reports answer "what happened"; the term implies trust and a decision.
  • Measuring nothing afterwards. Without a before-and-after number, you cannot tell a good decision from a lucky one.

Data intelligence in TapCub

Take the shop example above. TapCub covers the first four steps for behaviour data, and it does not take the action for you.

  1. Define what you collect. In All-platform analytics › Data › Tracking plan, record each intended event with its properties, platform, owner and status, so "checkout started" has one written meaning. See data insights for how insights are built on top.
  2. Check that it arrives. Turn on the real-time validation (last 24 hours) to see arrivals per event and how complete the required properties are. New events land in All-platform analytics › Data › Event inbox until someone claims them, and Property dictionary documents each property.
  3. Find the drop. In All-platform analytics › Behavior › Behavior analysis, build the funnel and read where users leave.
  4. Ask in plain language. All-platform analytics › Operations › AI Assistant answers a question with a chart, from queries run on your project's data, and its change attribution breaks down why a number moved by channel, source domain, country, device or landing page.
  5. Watch the result, and act yourself. A monitor in All-platform analytics › Operations › Monitors can notify you when a metric crosses a threshold. TapCub only notifies. It does not change your shipping page or run an experiment; the action and the decision stay with your team.

Sources

Was this explanation clear?

We use your answers to fix and expand this page.

Xu An

Customer Experience Consultant

Writes about live chat, visitor conversations, and self-service, sharing practical ways to make customer communication clearer and more effective.

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 data intelligence?

It is the practice of making data trustworthy and useful for decisions: knowing what data you have, where it came from and how good it is, then analysing it, acting on the result and measuring the outcome.

What is the difference between data intelligence and data analytics?

Analytics finds out what happened and why. Data intelligence is wider: it also covers understanding and trusting the data and carrying the finding through to an action and a measured result.

Is data intelligence the same as business intelligence?

No. Business intelligence reports performance against targets, usually in dashboards. Data intelligence supports it by making sure the data is understood and governed, and by tying findings to decisions.

What are the parts of data intelligence?

IBM lists metadata management, data lineage, data governance, data quality and data integration. In practice add an owner for acting on findings and a before-and-after measurement.

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