Costs and pricing
Part of Finance content budgeting in England: separate cost views
How attributed and incremental finance creator ROI stay separate in measurement
Finance creator ROI method: define contribution and total economic cost, separate attributed from incremental outcomes, and choose an evidence design before release.
Finance creators return on investment should use incremental contribution and total economic cost. In England, the FCA's financial promotion rules and HMRC's VAT guidance are the public anchors for claims and costs.
The measurement framework was checked on 5 September 2026. It gives no personal financial advice or expected return. The example is fictional and exists only to show how definitions change the answer.
What to take away
The core formula is:
(attributed contribution - total economic cost) / total economic cost
For the incremental result, use incremental contribution in the numerator. With one set of illustrative inputs, the attributed ratio was 20% while the incremental ratio was -64%. That contrast shows why the causal definition can reverse the result.
- Report attributed and incremental results in separate columns.
- Total economic cost includes external cash, internal time and unrecoverable VAT, not just invoices paid.
- A last-click report supports attributed outcomes only and cannot show what would have happened without the content.
- Without a valid counterfactual, publish attribution and cost rather than a precise incremental ROI percentage.
Define the numerator
Start with recorded transactions, then subtract refunds and the variable costs assigned to fulfil those transactions under a documented method. The remainder is the contribution used here. It is an internal measurement definition, not a tax or statutory-accounting rule.
Three Outcome Levels
Observed
- Definition
- Business recorded event
- Evidence
- Transaction log
- Causal claim
- No
Attributed
- Definition
- Tracking rule assigned event
- Evidence
- Last-click report
- Causal claim
- No
Incremental
- Definition
- Would not happen without content
- Evidence
- Counterfactual estimate
- Causal claim
- Yes
Next divide outcomes into three levels:
- observedthe business recorded the event;
- attributeda tracking rule assigned the event to content;
- incrementalthe evidence estimates that the event would not have happened without the content.
A last-click report supports the second level only. It does not describe people who would have bought anyway, other channels or concurrent changes.
HM Treasury's 2026 Magenta Book says impact evaluation examines what effects an intervention had, for whom and why. It explains experimental and quasi-experimental counterfactual methods. Where a strong counterfactual is unavailable, theory-based approaches can examine contribution, but they need a reasoned causal chain and evidence about alternative explanations.
Define the denominator
Total economic cost includes external production, distribution, licences, review, corrections and allocated staff or owner time. Record cash separately. A subscription paid in advance may create current cash movement but support several periods. Existing staff time consumes capacity even when no new invoice appears.
Allocate owner time and shared staff capacity by the hours actually consumed. Use time records or a documented estimate. If one person splits across projects, allocate a proportionate share and record the basis. Do not treat salary as the hourly cost without adding employer costs.
Include unrecoverable VAT in cost; HMRC's reclaim guidance sets out the conditions and evidence for reclaiming qualifying input VAT.
Work the illustrative numbers
The table below is a sensitivity table. Change one input at a time and recalculate the formulas above.
| Input | What determines it | Your figure |
|---|---|---|
| External cash cost before VAT | Invoices, licences, distribution | |
| Internal hours | Time records for owner and shared staff | |
| Hourly cost | Role, employer costs, opportunity cost | |
| Attributed conversions | Tracking rule | |
| Contribution per conversion | Revenue less refunds and variable fulfilment | |
| Incremental conversions | Holdout or evaluation estimate |
Use the formulas above. The contrast is directional: an attributed ratio can be positive while the incremental ratio is negative if the valid incremental conversion count is much lower. The 20% and -64% figures are illustrative only and predict nothing about a real creator campaign.
Choose an evidence design before release
Where feasible, define a comparable holdout, staggered release or other qualified design before the campaign. Measure contamination, sample size, time period and concurrent activity. HM Treasury's quality guidance for impact evaluation says outcome monitoring alone cannot show whether an intervention caused the change and stresses the counterfactual.
If a causal design is disproportionate, publish a contribution account: expected path, observed intermediate evidence, rival explanations and uncertainty. Do not attach a precise incremental percentage that the method cannot support.
Finally, keep performance claims out of promotional copy unless the evidence and legal route have been reviewed. FCA social-media guidance expects relevant financial promotions to be fair, clear and not misleading. An internal dashboard label does not make a returns claim suitable for consumers.
Before you act
- Define contribution after refunds and variable fulfilment costs.
- Separate observed, attributed and incremental outcome levels.
- Include internal time and unrecoverable VAT in total economic cost.
- Choose a holdout or staggered release design before launch.
- Publish a contribution account if a causal design is disproportionate.
- Review performance claims against FCA social media guidance before promotion.
Common questions
Why does total economic cost include internal staff time?
Existing staff time consumes capacity even when no new invoice appears. Total economic cost includes external production, distribution, licences, review, corrections and allocated staff or owner time. Record cash separately because a subscription paid in advance may support several periods.
What should you do if a causal design is disproportionate?
Publish a contribution account with expected path, observed intermediate evidence, rival explanations and uncertainty. Do not attach a precise incremental percentage that the method cannot support. HM Treasury guidance says outcome monitoring alone cannot show whether an intervention caused the change.
