KPI Progress and Recovery: Target Attainment, Forecasts, and Feasible Actions
Shusaku Yosa
When a KPI looks likely to miss its target, checking it more often is not enough. Distinguish whether performance is behind today’s plan, whether the forecast misses the final target, and whether the remaining capacity can close the gap. These questions require different numbers. A single attainment percentage cannot answer all three.
Separate the target, plan, actual, and forecast
A KPI tracks an important aspect of performance or service. Some KPIs describe intermediate activity, but meeting them does not guarantee the final business outcome. State the assumed connection between the measure and the objective, and revisit that connection when conditions change.
Number | Meaning | Question |
|---|---|---|
Final target | Desired value by a deadline | Where do we want to finish? |
Plan to date | Expected value at this point | Are we following the planned profile? |
Actual | Recorded value on consistent definitions | What has happened? |
Forecast | Expected finish under stated assumptions | Where might we finish? |
Include the unit, population, reporting period, direction of improvement, and data-finalization date. An incomplete daily count should not be compared with a finalized weekly figure as if both were equally complete.
Use an attainment formula only where it fits
Actual divided by target, multiplied by 100, can describe progress for a positive target where higher is better, such as a cumulative count. It becomes misleading if readers interpret a value above 100% as good for a measure that should fall. Division is undefined when the target is zero.
Measure | Example | Clearer reporting |
|---|---|---|
Acquisitions | Target 120; actual 45 | 37.5% of the final count target |
CPA | Target 10,000; actual 12,000 | 2,000, or 20%, above the cost target |
Critical defects | Target zero; actual one | One incident; no ratio to zero |
Purchase rate | Different weekly populations | Report numerator, denominator, and rate |
Do not simply average weekly percentages when the denominators differ. Aggregate the underlying counts, and keep the conversion definition and cost scope consistent.
Plan for working days, seasonality, and delays
Half the calendar month passing does not imply that half the outcome should have arrived. Working days, scheduled promotions, end-of-period bookings, and the time from action to outcome may produce an uneven plan. Compare actual performance with the planned cumulative value. A time-elapsed ratio is only a useful shortcut when the plan is reasonably uniform.
Review frequency should reflect data updates, usable sample size, and the delay before action can have an effect. Daily changes to a low-volume process may follow noise and make learning harder.
Calculate the shortfall and remaining work separately
Consider a hypothetical target of 120 acquisitions over 20 working days. At day 10, the plan is 50 and the actual is 45. Final-target attainment is 37.5%; attainment against the plan to date is 90%; the plan variance is minus five. These are different descriptions.
Calculation | Result | Assumption or meaning |
|---|---|---|
45 / 10 × 20 | Forecast 90 | Current daily pace continues |
120 − 90 | Forecast shortfall 30 | Additional outcome above that forecast |
120 − 45 | Remaining requirement 75 | Total still needed before the deadline |
75 / 10 | 7.5 per remaining day | Required future pace |
45 + 6 × 10 | Capacity-based finish 105 | Remaining capacity is six per day |
The current pace is 4.5 per day. Ask how it could reach 7.5, and whether capacity can exceed six. “Work harder” does not resolve the arithmetic. The linear forecast is a scenario, not a promised future.
Diagnose the constraint before choosing an intervention
First rule out collection or aggregation errors. Then examine the relevant components: reachable volume, response, conversion, price, or delivery capacity. A shortage of suitable prospects requires a different response from adequate reach but an unsuitable offer. Compare customer mix and selling conditions before attributing a change to the latest action.
The GOV.UK performance-metrics guidance connects measures with purpose, hypotheses, baselines, and context. Use that approach to document the change you intend to test. Avoid launching several unrelated interventions when you will not be able to distinguish their effects.
Decide what is feasible and when to reconsider
- Assign an owner, budget, start date, and earliest meaningful outcome date.
- Monitor quality, complaints, and additional cost alongside the target count.
- Define conditions for stopping or revising the intervention.
- If time or capacity is insufficient, reconsider scope, resources, or the target rather than hiding the constraint.
Revisit the customer and value assumptions through an STP worksheet, define customer interventions with a 1to1 execution plan, and assess what can be automated through MA tools. Keep the next decision grounded in the remaining work, capacity, and cost.