Methods and worked examples

How Impact calculates

Four worked examples. They are illustrative, and they double as release checks: each figure below is computed by the same engine the product uses and compared with independently written expected values. A mismatch stops the release.

Illustrative examples Corrected 5 October 2026: an earlier version showed £3.00 for the equal-allocation cost per visitor in example 1. The correct figure is £1.00.

The core formulas

  • Cost per result at any stage = investment ÷ results at that stage.
  • Blended view = total investment ÷ total results across all audiences.
  • Allocated view = each audience's allocated investment ÷ that audience's results.
  • Conversion = later-stage results ÷ earlier-stage results (same period unless a matched cohort is stated).
  • Estimated value = value basis × the share you attribute to marketing (an assumption, labelled as one).
  • Estimated return = (estimated value − investment) ÷ investment. Attributed value ÷ spend is shown separately; finance teams define ROMI in different ways, so we name the calculation instead.

Money is shown to two decimal places below £1,000 and to whole pounds above. Percentages to one decimal place. Calculations keep full precision; rounding happens only on screen.

Example 1 · Blended and equal allocation

A university invests £750,000 in a year. Its research, study and employer pages each receive 250,000 visitors (750,000 in total). Each audience also gains 25,000 newsletter subscribers.

MeasureCalculationResult
Blended cost per visitor£750,000 ÷ 750,000£1.00
Equal allocation, per audience£750,000 ÷ 3£250,000
Each audience's cost per visitor, equal allocation£250,000 ÷ 250,000£1.00
Blended cost per subscriber£750,000 ÷ 75,000£10.00
Each audience's cost per subscriber, equal allocation£250,000 ÷ 25,000£10.00

Example 2 · Unequal allocation

The same £750,000 is allocated 50% to students, 20% to academic and 30% to employers. Allocation audit: Fixed percentages: students 50.0%, academic 20.0%, employers 30.0%.

AudienceAllocatedVisitorsCalculationCost per visitor
Students£375,000400,000£375,000 ÷ 400,000£0.94
Academic£150,000150,000£150,000 ÷ 150,000£1.00
Employers£225,000200,000£225,000 ÷ 200,000£1.13

Example 3 · The student funnel

Students are allocated £375,000.

StageResultsCalculationCost per result
Enquiries12,000£375,000 ÷ 12,000£31.25
Applications6,000£375,000 ÷ 6,000£62.50
Enrolments900£375,000 ÷ 900£416.67

Enquiries to enrolments = 900 ÷ 12,000 = 7.5% (same period, not a matched cohort).

Example 4 · Estimated return and its sensitivity

Value basis: lifetime net tuition per student = £15,000 × 3 years × 0.90 retention = £40,500. Total = 900 × £40,500 = £36,450,000. How much of this to credit to marketing is an assumption, so we show a range.

Share attributed to marketingEstimated valueEstimated returnAttributed value ÷ spend
5%£1,822,500386.0%4.86×
10%£3,645,000872.0%9.72×
15%£5,467,5001,358.0%14.58×
What this does and does not show. These are estimates under stated assumptions, measured against the £375,000 student allocation. They do not show that marketing caused the enrolments. Multi-touch attribution assigns credit using CRM touchpoints; only an experiment or other defensible comparison can estimate incremental impact. Lifetime tuition is not a cohort cash-flow model, and net tuition is not contribution after delivery costs.

Try it with your own figures