
Costs and pricing
Part of How to budget for copywriting services before you commission work
Check copywriting services return on investment step by step
How to test copywriting services return on investment, with labelled example figures, attributable assumptions and the compliance rules behind the claims.
What to take away
- Return on investment for copywriting is a contribution margin test, not a revenue test, so isolate what the work changed.
- Set the baseline before a single word is written, or the measurement collapses into opinion.
- Attribute with a model you can defend, and keep the working visible to whoever signed the invoice.
- Compliance rules on claims and personal data apply to the measurement itself, not just the copy.
- A worked example beats a percentage lifted from an unrelated case study.
Decide what the spend has to prove
Start with the decision, not the metric. If the next question is whether to renew a retainer, you need a contribution figure per period. If it is whether to move budget from paid search, you need an incremental comparison between the two channels.
Match the decision to the metric
Renew a retainer?
contribution figure per period
incremental channel comparison
Write the decision on one line, then name the period it covers. A quarterly review of a £4,500 retainer asks a different question from an annual review of a £54,000 programme.
Agree the definition of return now. For most teams that means incremental gross profit minus the copywriting fee, divided by the fee. Anything else is a proxy, and proxies need labelling.
Step-by-step method
- Record the baseline. Capture conversion rate, average order value and organic sessions for the four weeks before work starts. Note the date of each figure.
- Separate the fee from the surrounding cost. Label the copywriting fee, the internal review time and any tooling as distinct lines in pounds sterling, as the copywriting services cost guide in England sets out in more detail.
- Choose one attribution model and keep it for the full test. Last-click is simple but understates early-stage pages.
- Tag and track. Confirm that the rules letting search engines crawl and index new pages follow how Google interprets the robots.txt specification, so the traffic you measure is the traffic you earned.
- Compare like with like. Hold seasonality, price changes and campaign launches constant, or state plainly that they moved.
- Calculate contribution. Subtract the fee and the variable cost of fulfilling the extra sales from the incremental gross profit.
- Write the decision note. State the figure, the period, the model and the main caveat in under 200 words.
Work through a labelled example
Suppose a B2B supplier pays a £2,000 monthly copywriting fee for six months, so £12,000 in total. Over the same period, the reworked pages generate 40 additional qualified enquiries at a 25 per cent close rate and an average gross profit of £1,800 per customer. That is ten extra customers and £18,000 of gross profit.
Subtract the £12,000 fee and the contribution is £6,000, a 50 per cent return on the fee for the period. Every figure here is an illustrative example, not a benchmark.
If the same team also spent 30 internal hours reviewing drafts at a labelled internal rate of £45 an hour, that is £1,350 of hidden cost, which cuts the return to roughly 39 per cent. Show that line. The fuller ledger, including the overheads, is what the costs and budget guide for England walks through.
Keep the claims and the data clean
What you publish about results is advertising, and the Advertising Standards Authority can act when claims mislead. Its published sanctions guidance explains what happens when a business ignores a ruling, which is a good reason to keep every performance claim evidenced before it goes live.
Personal data used in attribution, from form fills to CRM records, sits under UK GDPR. The ICO's accountability and governance guidance sets out the records and controls expected of teams handling that data, so build the audit trail alongside the spreadsheet.
Common questions
How long before return on investment can be judged?
Most content tests need at least one full quarter, and often two, because indexing and ranking shifts lag publication. Judge the trend, not a single month.
Should I count organic traffic as the return?
Only as an interim proxy. Traffic is an input. Return needs a contribution figure, which means revenue or gross profit tied to a defined period.
What if the work only maintains existing rankings?
Treat maintenance as avoided loss. Estimate what the traffic would have cost to replace through paid channels, label it as an estimate, and compare it with the fee.
Do I need a separate model for each channel?
One model applied consistently beats several applied selectively. If you must change models mid-test, restart the measurement window and say so.



