5.3QHSE indicators & dashboards
The essentials in 30 seconds
- Results + means: combine lagging (accidents, scrap) and leading (audits, reports) indicators — never one without the other.
- 5 to 8, not 30: each with an owner, a dated target and a 12-month trend.
- A dashboard that decides: each gap = comment + action + deadline, arbitrated in a monthly review.
Threaded mini-case — Hakim steers by traffic lights
Hakim no longer waves "zero accidents" alone: he tracks FR and SR with reported near-misses, scrap, closed actions and audits done — 7 indicators, each with target and owner. In the monthly review, each red light leaves with a dated action. Sequel to cases 5.1-5.2 — Hakim moves from findings to steering by figures.
Sequel to cases 5.1-5.2 — Hakim moves from findings to steering by figures.
1. Measure what matters: the pyramid
Outcomes (FR/SR accidents, scrap %, complaints) + means (% actions closed, audits done, training): steering on outcomes alone is driving by rear-view mirror. 5-8 indicators, neither 2 nor 30: each has an owner and a dated target.
Hakim keeps his board to 7 lines: enough to see, few enough to act.
2. FR/SR and the pyramid: use the right words
FR = lost-time accidents relative to hours worked, SR = days lost relative to hours: the French classics. Zero accidents must be completed by reported near-misses (Bird pyramid): a company reporting only zeros isn't looking.
Hakim celebrates the month's 6 reported near-misses: they prevent the real ones.
3. The dashboard that decides
One page: traffic lights (green/red vs target), trends (12 months, not one point), comment + action per gap. Presented at a monthly review that decides: a dashboard nobody comments on is decoration.
Each of Hakim's reds leaves with owner and deadline; the next review checks closure.
🔎 Concrete example — A one-page monthly dashboard
Hakim's 7 traffic lights.
FR 12 (target <15 OK) | SR 0.8 (target <1 OK) | near-misses 6 (target >4 OK)
Scrap 2.1% (target <2 KO -> set line 2, Karim, 15/04)
Closed actions 85% (target 90 KO -> follow-ups)
-> 2 decisions, not 20 readings.Result: each red has a dated action and an owner.
Lesson: reported near-misses are good news.
Section vocabulary
- QHSE indicator
- Quantified data measuring QHSE performance to make it objective, track and steer.
Simple example: The workshop's monthly FR tracked by Hakim with its target.
What it is not: a feeling ("things look better") with no figure or target.
Reflex: does each indicator have owner + dated target?
- Lagging
- Measures past outcomes, after the fact: accidents, scrap, complaints.
Simple example: FR, SR, scrap rate and complaints of the month.
What it is not: the number of training courses (that's leading).
Reflex: where do we stand? (rear-view mirror).
- Leading
- Measures prevention actions, before outcomes: forward-looking.
Simple example: Audits done, reported near-misses, closed actions.
What it is not: the number of reported accidents alone.
Reflex: what are we doing so tomorrow is better?
- Dashboard
- A concise, visual page: traffic lights, trends, gaps — to decide.
Simple example: Hakim's single page presented at the monthly review.
What it is not: 30 detailed tables nobody comments on.
Reflex: does it fit on one commented page?
- Actionable
- Serves to decide and act; beware perverse effects (under-reporting).
Simple example: Each of Hakim's reds = action + owner + deadline.
What it is not: waving zero without listening to weak signals.
Reflex: does this indicator drive action or cheating?
In practice — Steering by indicators
- Choose relevant indicators in limited number, combining lagging (outcomes: accidents, frequency/severity rates, nonconformities, waste) and leading (prevention: training, audits, reports).
- Ensure their reliability (clear definition, correct data, consistent measurement) and track trends + gaps against targets.
- Build a concise, visual dashboard (charts), sent to the right recipients, at the right frequency.
- Use indicators to decide and act (steer improvement); beware perverse effects (e.g. don't incentivise under-reporting).
Ready-to-say script — present the red without drama (1 min)
"Scrap at 2.1% for a target of 2: we are 0.1 point red, rising for two months. Likely cause: line 2 setting. I propose: Karim sets by the 15th, we check the next batch, review next month. Who approves?"
Say in review: fact, gap, cause, action, deadline — then note the decision.
Copy-paste template — indicator line
INDICATOR (lagging / leading): _______________
DEFINITION + DATA SOURCE: _______________
DATED TARGET + OWNER: _______________
VALUE + 12-MONTH TREND: _______________
GAP -> ACTION + OWNER + DEADLINE: _______________Small business / tight budget
Hakim keeps 5 lines in a spreadsheet: FR, near-misses, scrap, closed actions, audits — a 30-minute review.
Algeria SME
Hakim completes FR/SR with shop-floor reports: without feedback, the local zero means nothing.
Excel / Power BI link
One chart per indicator (rolling 12 months) beats ten tables of raw figures.
Perverse effect
If accidents fall but reports drop to zero: audit reporting, don't celebrate performance.
✓ Best practices
- Keep to 5-8 indicators, each owned and targeted.
- Read 12-month trends, not isolated points.
- Hold a review that decides (action + deadline per gap).
✗ Common pitfalls
- Stacking 30 unreadable indicators.
- Waving zero with no near-misses (blindness).
- Leaving the board as never-commented decoration.
Key takeaways
- Lagging = past outcomes; leading = prevention; combine both.
- Dashboards: concise, visual, trends and gaps against targets.
- Relevant, reliable, limited, actionable.
- Perverse effects: measuring only reported figures can push under-reporting.
48h challenge — apply tomorrow
Choose 5 indicators (at least 2 leading) with target + owner, plot their 3-month trend on one page.
- I distinguish lagging (where we stand) from leading (what we do).
- I can read FR/SR with near-misses.
- My one-page board is ready for review.
Frequently asked questions
Why combine lagging and leading indicators?
Lagging indicators tell where you stand (consequences, after the fact), leading ones tell what you do to improve (before, forward-looking). Steering on outcomes alone is driving by rear-view mirror: you see the wall after. Both give the full picture.
Should you aim for zero accidents?
Yes as a course, never as sole proof: a zero with no reported near-misses or prevention actions often signals you are not looking (Bird pyramid). A good zero comes with living reports, audits and closed actions.
Zero accidents posted, zero near-misses reported. Reading?