Data Quality Score — know how much to trust your sprint forecast
Your forecast is reading from your data. If your data is broken, the forecast is too. Avium scores the gap.
Garbage in, garbage out — quantified
Every team's Jira data has gaps. Unassigned tickets, missing estimates, tickets with no sprint, tickets that should have been closed but were forgotten. Most teams have learned to ignore the gaps; the analytics ignore them too and confidently report numbers they shouldn't trust.
The insidious version: a velocity chart that looks great because half the team's work is in 'unestimated' tickets that don't count. Or a sprint forecast that says you're on track because the four blocked tickets have no due date.
How teams try to keep Jira clean today
The familiar treadmill:
- 'Jira hygiene' meetings — well-intentioned, attended once, ignored thereafter.
- Periodic audits where the scrum master combs the backlog manually. Catches 60% of issues; misses the systemic ones.
- Required-fields rules in Jira itself — works at creation time, but tickets created in bulk or imported from other systems bypass them.
- Trusting the analytics anyway because there's no other choice. The dashboard shows a number; you quote it; nobody investigates the foundation it sits on.
How Avium Signals computes data quality
Avium runs through every open ticket and counts gaps in the fields that matter for analytics. The score is a weighted aggregate, 0-100, where 100 is clean.
- Missing assignee — heaviest weight (1.0). An unassigned ticket has no person to attribute capacity to; corrupts both sprint risk and over-allocation signals.
- Unestimated — weight 0.8. Tickets without story points or planned hours can't be counted in velocity or capacity.
- Missing story points (Jira-imported tickets only) — weight 0.6. Separate from unestimated to flag the 'we use points except sometimes' inconsistency.
- No sprint — weight 0.5. A ticket that exists outside any sprint isn't being planned; in a points-based shop that distorts velocity reads.
- Overdue — weight 1.0. Tickets whose due date passed are either done-and-not-closed (data lie) or genuinely late (operational issue). Either way, surface.
- Score = 100 minus the sum of (issue weight × percent of open tickets affected). Clamped to [0, 100].
Who reads this Signal
See your data quality score
Free tier shows the score and the top gaps the first time you load data. The Avium Intelligence layer on Business turns the gap list into a prioritized cleanup queue.
Start free — no credit card →