Costs
Part of Customer analytics: what to keep and what to drop
Customer analytics tools: a buyer scorecard
Customer analytics tools need more than feature checks. Use a buyer scorecard to test identity, portability, cost, and operating burden before purchase.
What to take away
- A customer analytics stack contains different systems with different evidence boundaries.
- Identity and permission tests matter more than the number of connectors.
- Require reproducible exports and deletion tests before purchase.
Customer analytics tools include source systems, warehouses, customer-data platforms, product analytics, web analytics, survey systems, notebooks, experiment services, business intelligence, and activation platforms. No single category supplies a complete customer truth.
On a marketing analytics and measurement site, this stack is the evidence layer beneath attribution and metric analysis. Attribution models, channel comparisons, and conversion metrics depend on the identity rules, event definitions, cohort logic, and export controls described here.
If those inputs are weak, a clean attribution report or dashboard can still be wrong. The buyer scorecard below therefore treats identity, cost, and exit as measurement questions, not only procurement questions. Readers responsible for attribution, campaign measurement, or marketing analysis can use the same tests to judge whether a customer analytics tool will support defensible marketing answers.
Use a buyer scorecard
Buyer scorecard
- Identity
- Keep source IDs, confidence, conflicts, and history
- Events
- Preserve time, keys, value, quality, and revisions
- Cohorts
- Control entry, eligibility, age, and maturity
- Segments
- Version rules, overlap, refresh, exit, and suppression
- Privacy
- Apply purpose, consent, rights, retention, and deletion
- Analysis
- Separate description, prediction, and causal evidence
- Activation
- Trace audience, delivery, and outcome
- Exit
- Export data, rules, models, logs, and deletion proof
The named examples below show how public documentation can inform a buyer scorecard for identity, cost, and exit; they are not a like-for-like price comparison.
Tool comparison
Identity rule
- Google Analytics 4
- User ID, device ID, and Google signals; reporting identity can be blended, observed, or device-based.
- Mixpanel
- distinct_id and $device_id; identify() merges anonymous and identified profiles.
- Amplitude
- User ID, Device ID, and Amplitude ID; identify() merges users.
- Segment
- external ID, email, and anonymous ID; Profile Unification merges profiles.
- Microsoft Customer Insights
- Mapping, matching, and merging in data unification.
Cost evidence
- Google Analytics 4
- Free within standard limits; Analytics 360 is commonly quoted at about $150,000 per year.
- Mixpanel
- Public pricing lists a free tier and paid plans scaled by monthly events.
- Amplitude
- Public pricing lists a free tier and paid plans scaled by monthly tracked users or events.
- Segment
- Free tier 1,000 monthly tracked users; Team plan from $120 per month for 10,000 monthly tracked users.
- Microsoft Customer Insights
- List price $1,500 per month for 100,000 profiles.
Exit rule
- Google Analytics 4
- BigQuery export for raw events; standard retention is 2 or 14 months, 360 up to 50 months.
- Mixpanel
- Export API for events and profiles; deletion API for profiles.
- Amplitude
- Export API for events and users; deletion API for users.
- Segment
- Warehouse destinations and Webhooks; Profile deletion API.
- Microsoft Customer Insights
- Export profiles to Azure Data Lake; deletion via API.
Google documents funnel exploration as a way to visualize and analyze user steps in Google Analytics. Its open or closed funnel settings and product user scope must be understood before the result is called a customer journey.
Test identity unification directly
Microsoft's data-unification documentation describes mapping, matching, and merging data in Customer Insights. It documents a product workflow, not proof that a buyer's identifiers are correct. False merges and false splits head the list of customer analytics mistakes a pilot should trigger on purpose.
Identity Unification Test Steps
- Load duplicate, conflicting, stale, missing records
- Test shared email and household edge cases
- Change identity rule and compare profile history
- Revoke permission and verify downstream suppression
- Restate returned order and affected cohort
- Export complete decision and correction record
Price licensing, usage, identity services, storage, computing, implementation, consent management, security, analysts, experiments, support, and migration. Ask who pays when a connector changes or an audience must be recalled.
Use a bounded paid pilot with representative defects. A demonstration built on clean sample profiles cannot show how the system behaves when real identities conflict or customer rights must propagate.
Score the pilot with evidence, not presentation quality. Require source reconciliation, change history, access logs, representative query timing, user training, error handling, and a complete export. Have the future operator rerun the acceptance pack without vendor help before signing a long contract. That pack works best as a customer analytics checklist with a named reviewer against each line.
Test the operating burden
Give each tool the same source sample, user roles, required output, failure case, and export task. Measure setup time, recurring labor, specialist help, correction work, shutdown effort; a feature earns credit only when an intended user completes the task safely.
Save the tested plan and date: a current product page doesn't promise the same function or price in 2027.
The GOV.UK technology selection guidance recommends choices that can change over time, preserve data control, address security risk, and include ownership cost. Those public-service rules make useful buying questions, but they are not private-sector mandates.
The GOV.UK open standards guidance connects interoperability with reuse and reduced supplier dependence. Test whether data, definitions, and decision records can move in documented formats. The page governs UK public-service work, so treat it as a portability prompt.
Keep the evidence record beside the decision so a reviewer can reproduce the reasoning. The pillar on customer analytics sets out what the stack is being bought to answer.
Common questions
Is a CDP the same as customer analytics?
No. A CDP can unify and activate data; analysis also needs governed questions, methods, outcomes, economics, and review.
Should tools predict churn automatically?
A prediction still needs a defined outcome, validation, error review, permitted use, treatment test, and accountable owner.
What must be portable?
Export source mappings, identities where permitted, event definitions, segments, models, quality states, audiences, decisions, corrections, and deletion records.
