Define · Add-on
Set data quality and trust measures
If people do not trust the data, they will not use it. Agree what good quality means and how you will show it.
About 2-4 daysEvergreenLast reviewed 5 November 2026
Who this applies to
- Skills it builds
- Data and measurement
- PM types
- Data
- Product types
- All
- Starting points
- All
- Ways of working
- All
- Experience
- All
- Size of organisation
- All
Steps, by experience
Beginner (under 1 year)
- Agree what quality means for your users: accuracy, completeness, freshness and consistency.
- Set measurable targets for each, such as 'updated by 8am daily'.
- Decide how you will check them automatically.
- Show users when data was last updated and any known issues.
- Agree who fixes problems and how fast.
Intermediate (1 to 4 years)
- Define quality dimensions and targets.
- Automate checks and show freshness.
- Agree ownership and fix times.
Experienced (5+ years)
- Publish data contracts between producers and consumers.
- Track trust through user feedback and usage.
- Report data quality as a product health measure.
Who can do this
Owner: Product manager
Can be delegated to
- Data Analyst or Data Scientist / ML Engineer: Define checks and targets
- Backend Engineer: Build automated quality checks
Reviewed by: Data Analyst, Tech Lead
Tool examples
- Looker
- Show freshness and quality status on each dashboard.
- Power BI
- Add a data quality page to the report.
Definition of done
- Quality dimensions and targets agreed
- Automated checks in place
- Freshness visible to users
Common pitfalls
- Users finding errors before you do
- No owner for fixes
- Quality defined only by engineers
If you lead the team
What good looks like: Data quality has agreed measures, owners and visible scores, so people trust it.
Ask your PM
- How do we measure data quality, and who owns fixing it?
- How do users know whether to trust a figure?
Warning signs
- People keep their own spreadsheets because they don't trust the data.
- Quality issues are fixed one complaint at a time.