PR return on investment once AVE is out of the value column

PR return on investment runs in four steps: name the evidence state, remove AVE from the value column, specify cost and counterfactual, calculate only when terms are supported. It is a ratio that needs a defensible incremental contribution in the numerator and the full economic cost for the same decision, population and period in the denominator.

What to take away

Name the evidence state

Create separate ledger rows for six states:

AMEC's Integrated Evaluation Framework introduction traces the distinction between inputs, activities, outputs, outcomes and impact. It is global industry methodology and does not prove a private organisation's return. Its practical value here is to stop a press release, placement or media mention being reported as revenue.

Remove AVE from the value column

AMEC's June 2025 Barcelona Principles 4.0 identifies advertising value equivalents as invalid and directs evaluation towards communication outcomes and impact. A space-cost equivalent does not observe a sale, contribution margin or causal effect. Multiplying it does not repair that gap.

If a buyer keeps a media-space comparison for a limited descriptive purpose, label the method and assumptions. Do not call the output value, impact, return, revenue or saving.

Specify cost and counterfactual

The denominator should cover agency service, tools, monitoring, distribution, research, production, paid placements, creators, travel, buyer labour, accessibility, rights, legal and privacy review, security, correction, crisis capacity and exit. Use finance-approved treatment for VAT and other taxes. Do not remove recoverable VAT until the actual organisation and records have been reviewed.

In England the buyer's VAT position follows UK rules. A VAT-registered business can usually recover VAT on agency and production invoices, leaving the net amount and any irrecoverable VAT in the denominator.

HM Revenue and Customs sets the UK standard rate at 20 per cent. A £40,000 net agency fee therefore carries £8,000 of VAT, once the records confirm how much is recoverable.

For UK cost inputs, the PRCA PR and Communications Census and the CIPR State of the Profession publish salary and budget figures. Pay data for public relations professionals sits in the ONS Annual Survey of Hours and Earnings, and the IPA Bellwether Report covers wider marketing budgets. Treat these as typical ranges, not a rate card.

Next define the alternative: no appointment, existing in-house work or another planned intervention. HM Treasury's Magenta Book, updated 15 May 2026, treats impact evaluation as a way to estimate causal effects and discusses comparison with what would otherwise have occurred. It is government guidance, so a qualified evaluator must adapt any design rather than cite it as validation.

Calculate only when terms are supported

Let the buyer define contribution consistently, commonly as revenue less the variable cost of fulfilling that revenue. Then use blank fields:

incremental contribution = estimated causal outcome difference x finance-approved contribution per eligible outcome

commercial ROI = (incremental contribution - full economic cost) / full economic cost

Worked example, illustrative only. A UK agency retainer of £4,000 a month over 12 months is £48,000. Tools and monitoring add £6,000, production £4,000 and buyer labour £2,000 over the same 12 months, so the full economic cost is £60,000.

Suppose the finance-approved contribution is £30 per eligible order and the estimated causal outcome difference is 3,000 orders from the UK eligible-buyer population in that 12-month period. Incremental contribution is 3,000 x £30 = £90,000. Commercial ROI is (£90,000 minus £60,000) divided by £60,000 = 0.5, or 50 per cent.

State currency, period, population, numerator, denominator, inclusions, exclusions and model version. Stop the calculation if cost is zero or incomplete, the counterfactual is not credible, outcome definitions change, concurrent activity cannot be separated, or contribution is not finance-owned.

The outcome window must also fit the mechanism. A short window may miss a delayed response; a long one may admit more unrelated influences. Pre-register the start, end and treatment of outcomes outside it.

If the organisation changes price, distribution, product availability or sales process during the study, document the change and ask the evaluator whether the comparison remains usable. Do not repair a broken design by widening the window after seeing the result.

Publish uncertainty, not polish

Report the observed outcome, attributed share and causal estimate in distinct columns. Test buyer-approved ranges for outcome difference, contribution, lag and total cost. Record missing data, spillover, selection, measurement changes and who reviewed the analysis.

A positive attributed ratio is not causal evidence and cannot clear a failed legal, privacy, security, accessibility, rights or advertising gate. The defensible conclusion may be "return not established". That is a usable result when it prevents an unsupported commercial claim.

Before you act

Common questions

Why is advertising value equivalent invalid for measuring PR return?

The Barcelona Principles 4.0 identifies advertising value equivalents as invalid and directs evaluation towards communication outcomes and impact. A space-cost equivalent does not observe a sale, contribution margin or causal effect, and multiplying it does not repair that gap.

What should the denominator of a PR return calculation include?

It should cover agency service, tools, monitoring, distribution, research, production, paid placements, creators, travel, buyer labour, accessibility, rights, legal and privacy review, security, correction, crisis capacity and exit. Use finance-approved treatment for VAT and other taxes, and do not remove recoverable VAT before review.

When should the return calculation be stopped?

Stop if cost is zero or incomplete, the counterfactual is not credible, outcome definitions change, concurrent activity cannot be separated, or contribution is not finance-owned. Do not repair a broken design by widening the outcome window after seeing the result.