Metric definitions
Specify the calculation, population, period, units, and owner. Record exclusions and changes in meaning over time.
Data analytics and business reporting
Turn existing product and operational data into a report your team can use for a defined decision. ACW reviews sources, aligns metric definitions, and investigates changes with the population, time period, and data limitations made explicit.
Contact UsResolve the gaps behind conflicting numbers
Data from separate tools may describe different users, time windows, or stages of a process. Combining it without checking those differences can make a dashboard look consistent while answering the wrong question.
Start with the measure that affects the decision, the records behind it, and the comparison your team needs to make. Identify missing or unreliable inputs before treating the report as a baseline.
Structure the analysis around four tasks
What your team receives
A source and metric inventory for the agreed question.
Data preparation and transformation work within scope.
A report or dashboard with documented calculations.
Findings, limitations, and recommended follow-up checks.
Update and handover instructions where recurring reporting is included.
From source review to a usable report
Define the question
Agree the decision, reporting audience, and comparison period. Confirm access rights and which fields may be processed or retained.
Check the data
Inspect representative records, coverage, identifiers, and reconciliation gaps. Resolve or label issues before analysis.
Build the analysis
Prepare the selected dataset and calculations. Validate key totals against available source records.
Review the result
Walk through findings with the team. Identify actions supported by the evidence and questions needing another source or a test.
Verify the report before relying on it
Acceptance should cover metric definitions, key totals, filters, date handling, and any refresh process included. A reader should be able to understand where a number came from and which records it excludes. An unexplained discrepancy remains visible until resolved.
This is an example of a possible engagement, not an ACW client result.
Illustrative analysis: enquiry progress
The team compares enquiries received with records reaching an agreed qualification stage. The analysis checks identifiers, duplicate entries, and date definitions before reporting conversion. Missing records are documented; the observed change becomes a question for investigation, not automatic proof of better sales performance.
Questions before analysis
Can you combine several sources?
Yes, if access, identifiers, definitions, and data quality support the join. Some sources require preparation or cannot be compared directly.
Does this include tracking implementation?
This service focuses on existing data. New event collection and instrumentation can be defined through Product Analytics.
Will the report explain why a metric changed?
It can investigate plausible explanations. Establishing cause may require additional research or a suitably designed experiment.
Show us the number your team cannot explain
Share the decision, source systems, and the comparisons that currently produce conflicting answers.