Designing a performance dashboard at Kestrel Physiotherapy
Kestrel Physiotherapy runs 24 private clinics. Over two years appointments rose 9% to 412,000 and revenue rose 7% to £18.6m, but EBITDA margin fell from 18.2% to 13.4%. Clinician turnover is 28% against a sector norm of 19%; the proportion of new patients rebooking within six weeks fell from 71% to 58%; net promoter score fell from 42 to 17; complaints about waiting times rose 60%. The board pack contains only revenue, appointment volume and EBITDA, shown by clinic. Clinic managers' bonuses are based 70% on appointment volume and 30% on revenue, and marketing has been running discount offers to fill empty slots. Senior clinicians say they have no time for administration or mentoring, and two regional managers have left. The CEO has asked for a dashboard that tells us what is really happening before it hits the numbers.
1. Critique Kestrel's current performance measurement, using the data to explain how the incentive structure has produced the outcomes described.
Read the worked answer
What is measured
The board sees only lagging outcome measures - revenue, volume, EBITDA - and sees them only in aggregate by clinic. There is nothing on customers, quality, workforce or process, so the pack reports the result of past decisions and gives no early warning.
What the numbers imply
Revenue rose 7% while appointments rose 9%, so average revenue per appointment fell by about 1.8% (£45.98 to £45.15). That is consistent with discount-led marketing filling capacity rather than genuine demand growth. EBITDA margin fell 4.8 points, which suggests the extra volume was bought with discounts and delivered with more expensive, less productive labour as turnover rose to 28%.
How the incentive did the damage
With 70% of the bonus on appointment volume, managers are rewarded for filling slots whatever the price and whatever the strain on clinicians. Discounting, tight scheduling and pressure to rebook quickly follow logically. The results are visible: waiting-time complaints up 60%, rebooking down 13 points, NPS down 25 points, turnover 9 points above the sector. Rebooking is the critical leading indicator - a 13-point fall on 412,000 appointments points to a future revenue decline that the board pack cannot show. The two departing regional managers suggest the pressure is also affecting management capacity. The system is measuring activity, not value.
2. Design a balanced scorecard for Kestrel, giving one lagging and one leading indicator for each perspective with a brief rationale. Explain how you would reduce gaming of the new measures.
Read the worked answer
| Perspective | Lagging indicator | Leading indicator | Rationale |
|---|---|---|---|
| Financial | EBITDA margin per clinic | Revenue per available clinical hour | Margin shows the outcome; revenue per hour shows whether discounting is eroding yield before margin falls |
| Customer | Net promoter score | Rebooking rate within six weeks; wait time to first appointment | Rebooking and waits predict next quarter's revenue and NPS |
| Internal process | Complaints per 1,000 appointments | Percentage of appointments starting on time; clinician utilisation (capped at 85%) | Access and reliability drive complaints; capping utilisation protects quality |
| Learning and growth | Voluntary clinician turnover | Training and mentoring hours completed; internal promotion rate | Turnover is the outcome; development and progression predict retention |
Reducing gaming
Define rebooking as a patient-initiated appointment booked within 14 days, not one keyed in by staff, and audit a sample monthly. Set targets at clinic level rather than individual clinician level to discourage cherry-picking easy cases. Cap any single measure at 30% of the bonus and add quality gates, so volume cannot be earned by breaching waiting-time or complaint thresholds. Use an independently administered patient survey, publish data definitions, and review them quarterly with clinicians. Finally, pilot the dashboard for one quarter and compare it with the old pack before linking it to pay, so measurement error does not become a pay dispute.