Introduction
If you’re honest, does your backlog health inspire confidence—or dread? Many teams plan with good intentions but get blindsided by stale items, half-ready stories, and a growing queue that never seems to shrink. In this guide, we’ll make backlog health concrete using four measurable signals—item age, readiness ratio, arrival vs. completion rate, and reorder churn—and give you a no-nonsense weekly inspection checklist you can start using today.
Why this matters: flow-based discipline is still rare. In the latest State of Agile, only 18% of respondents said they use flow metrics at an executive level, and just 20% are very satisfied with Agile in their companies—evidence that better measurement and backlog stewardship are overdue. 2288549.fs1.hubspotusercontent-na1.net
What “Healthy Backlog” Actually Means
A healthy backlog is small, current, outcome-focused, and ready near the top. Practically, that means:
- New work enters at a sustainable pace.
- The “next up” slice is truly ready (meets your DoR).
- Items don’t linger or keep jumping around in priority.
- You can forecast with the throughput you actually achieve.
The four signals below quantify those behaviors so you can inspect and adapt.
Signal 1 — Item Age: Are We Letting Work Go Stale?
Definition (two views):
- In-progress item age: time since work started on a ticket. This is the Kanban “aging WIP” lens. If items age without finishing, feedback is delayed and risk rises. prokanban.orgTeamhood
- Backlog item age (not yet started): time since creation or last meaningful update. Excessive age near the top signals staleness and unclear value.
How to measure (weekly):
- Create an “Aging” report for In Progress items; highlight anything above your 85th percentile cycle time as at risk.
- For the top 30–50 Product Backlog Items (PBIs), flag items older than 90 days (or untouched in 30+ days).
Target heuristics:
- ≤10% of In Progress items exceeding your 85th percentile cycle time.
- ≤15% of the top 50 backlog items older than 90 days (or untouched in 30+).
Why it matters: Aging items correlate with hidden blockers, over-commitment, and invisible WIP—all flow killers. prokanban.org
Internal resource for deeper practice: see Master Product Backlog Refinement: 5 Steps for Effective Results on my blog.
Read the guide
Signal 2 — Readiness Ratio: Is “Next Up” Truly Ready?
Definition: The readiness ratio is the percentage of items in your “next-up” slice (e.g., the top 10–15 PBIs) that meet your Definition of Ready (DoR)—clear acceptance criteria, small enough, dependencies identified, etc. AtlassianScrum.org
How to measure (weekly):
- Pick a fixed inspection window (e.g., top 15 PBIs).
- Count how many meet DoR.
- Readiness ratio = Ready PBIs ÷ Window size.
Target heuristics:
- ≥80% ready in the “next-up” window.
- 0 critical external dependencies unresolved for the next Sprint.
Why it matters: Low readiness creates sprint churn, rework, and missed goals. DoR is a simple pre-flight that protects flow. Atlassian
Internal resource for product-side enablement: Product Owner Program (coaching for robust DoR & prioritization).
Explore the program
Signal 3 — Arrival vs. Completion Rate: Is the Queue Growing?
Definition: Compare how many work items arrive (enter your workflow) to how many are completed (depart) over the same period. If arrival > completion for long, your backlog and WIP will grow; if completion ≥ arrival, you’re stabilizing. (This is the heart of flow and the logic behind cumulative-flow analysis.) 2288549.fs1.hubspotusercontent-na1.net
How to measure (weekly):
- Track arrivals (tickets moved from “Backlog” into “In Progress”) and completions (Done) in the last 7–14 days.
- Visualize a simple control chart or CFD; check that arrival and completion trend lines are roughly parallel across weeks.
Target heuristics:
- Completion ≥ Arrival on a rolling 4–6-week horizon.
- If arrivals spike, explicitly reduce intake the following week.
Why it matters: Persistent mismatch = growing queues, longer cycle times, declining predictability. (The State of Agile also highlights cycle/wait times and bottlenecks as top DevOps wishes—precisely what stable flow addresses.) 2288549.fs1.hubspotusercontent-na1.net
Signal 4 — Reorder Churn: Does Priority Whiplash Your Team?
Definition: Reorder churn is the percentage of backlog items whose rank changes materially (e.g., ±10+ positions) within a week or sprint. It’s a volatility lens on your Product Owner’s ordering.
How to measure (weekly):
- Snapshot the backlog rank each Friday.
- Next Friday, compute % of items that moved by ≥10 positions (or jumped in/out of the top-N window).
Target heuristics:
- ≤10–15% reorder churn week-over-week in the top-N (say top 30).
- Bigger moves are fine for brand-new discoveries—just keep them visible and bounded.
Why it matters: High churn breaks flow, invalidates forecasts, and forces unplanned work mid-sprint (aka “churn” inside the iteration), a known anti-pattern. IBM
Signal 5 — Backlog Size & Staleness: Are We Hoarding Ideas?
Definition: The working set (the portion you might deliver within ~3 months) should be small and current. Unchecked growth is a red flag. Atlassian
How to measure (monthly + quick weekly glance):
- Size of top-N working set (e.g., next 2–3 releases).
- % of items in that set older than 90 days or never touched.
Target heuristics:
- Working set fits your real throughput (e.g., 2–3 months of work).
- ≤15% of working set is stale (age >90d or no update >30d).
The Weekly Backlog Health Inspection (15–20 minutes)
Who: Product Owner (+ Scrum Master, Tech Lead).
When: Same time every week; time-box to 20 minutes.
Inputs: Throughput last 2–4 weeks, aging report, DoR checklist, arrivals vs. completions, reorder churn snapshot.
Checklist
- Throughput check
- 2–4 week avg throughput = ____ items/week.
- Capacity next sprint adjusted? Yes/No.
- Readiness ratio (top 15)
- Ready = ____ / 15 (≥80% target).
- Gaps: AC missing / too big / dependency / unclear value.
- Arrival vs. completion
- Arrivals last week = ____ ; Completions = ____ .
- If arrivals > completions for 2+ weeks, cut intake and finish aging items.
- Item age & staleness
- In-progress items over 85th percentile: ____ (target ≤10%).
- Top-50 backlog: ____ items >90d (target ≤15%). Remove or refresh.
- Reorder churn
- % of items moving ≥10 positions (top-30): ____% (target ≤10–15%).
- Document reasons for any big moves (new evidence, strategy shift).
- Outcome alignment
- Top-10 trace to product goal/user impact? Yes/No. Remove “orphans.”
- Small batch pledge
- Split anything > 1 sprint of effort. Avoid mega-stories.
Tip: Keep the checklist inside your backlog tool (custom fields or saved views) so the numbers fill themselves.
Putting It Together: Healthy Backlog Scorecard (example)
| Metric | Today | Target | Status |
|---|---|---|---|
| Readiness ratio (top 15) | 73% | ≥80% | 🔶 Improve |
| Arrivals vs. Completions (4-wk) | 42 vs. 45 | C ≥ A | ✅ OK |
| Reorder churn (top 30) | 22% | ≤15% | 🔴 High |
| In-progress above 85th pct. | 3/28 (11%) | ≤10% | 🔶 Close |
| Stale in top 50 (>90d) | 5 (10%) | ≤15% | ✅ OK |
How to Repair Each Signal (Playbook)
- Low readiness ratio → Run a DoR tune-up workshop; add example-based AC, slice to “one-sprint fit,” eliminate external dependencies before commitment. Atlassian
- Aging items → Set explicit WIP limits; swarm on oldest in column; make blockers visible daily. Teamhood
- Arrival > completion → Freeze new starts for a week; finish started work; renegotiate scope intake until lines stabilize (CFD logic). 2288549.fs1.hubspotusercontent-na1.net
- High reorder churn → Establish a cadence (weekly ordering), agree a “top-N protected” policy during sprints, and require evidence (data, experiment, customer signal) for big jumps. IBM
- Backlog bloat → Introduce a “Use-By Date” for ideas; archive or test via small experiments; keep the working set to what your throughput can credibly deliver in ~3 months. Atlassian
FAQ
Q1: Is a Definition of Ready “anti-Scrum”?
No. Scrum doesn’t mandate DoR; it’s an optional practice. Many teams use DoR as a lightweight quality gate for smoother flow and clearer commitments. Atlassian
Q2: What window should I use for “next-up” readiness?
Pick a constant inspection size (e.g., 10–15 PBIs). Big enough to plan, small enough to be specific.
Q3: Points or items for arrival/completion?
Use items to sense flow stability; use points only if your team estimates consistently.
Q4: How much churn is acceptable?
Some churn is healthy (learning!). Sustained >15% reorder churn in the top-N usually indicates reactive prioritization that harms predictability. IBM
Q5: Do we need a CFD to use arrival vs. completion?
No. A simple weekly count works. CFDs just visualize it nicely (parallel bands = healthy flow). 2288549.fs1.hubspotusercontent-na1.net
Conclusion (and Next Step)
Backlog health isn’t a fuzzy feeling—it’s a handful of simple, visible behaviors you can measure every week. Track item age, keep a high readiness ratio, balance arrival vs. completion, and dampen reorder churn. The weekly inspection keeps you honest and aligned with outcomes. If you want help setting up your scorecard and dashboards, let’s talk—this is a fast win for predictability and stakeholder trust.
Explore related posts and services on my site:
• Product Backlog Refinement: 5 Steps → igorlakic.com/blog/mastering-product-backlog-refinement-5-steps-for-effective-results
• Scrum Program (team enablement) → igorlakic.com/agile-coaching/scrum
