Strategy & Budgeting for Paid Media

Guide · Strategy & Budgeting

Set budgets that move the needle, not just the spreadsheet

How to ground your paid media budget in real unit economics, separate testing spend from scaling spend, and decide where your money genuinely earns its place.

Start with the maths that actually matters

Most paid media budgets are set the wrong way round: a number is agreed first, then the strategy is bent to fit it. The stronger approach is to work back from your unit economics. If you know your average order value, gross margin, and the customer lifetime value you can reasonably expect, you can derive a target cost per acquisition that keeps each sale profitable rather than hopeful.

The one number to anchor everything: your maximum allowable cost per acquisition (mCPA). It is the ceiling that keeps growth profitable, and it is the figure every campaign decision should be measured against.

1

Testing budget

A ring-fenced amount you are willing to spend to learn. Judged on insight and signal quality, not immediate return. Treat it as research, and protect it from short-term performance pressure.

2

Scaling budget

Spend directed only at combinations of audience, creative and channel that have already proven they clear your mCPA. This is where efficiency is expected and defended.

3

Always-on budget

The baseline that captures existing demand — brand search, retargeting, and high-intent audiences. Steady, predictable, and rarely the place to experiment.

4

Reserve

A deliberate 10–15% held back for moments that deserve it: a breakout creative, a seasonal spike, or a competitor stumble worth capitalising on.

Split testing spend from scaling spend

The single most common budgeting mistake is judging every pound by the same yardstick. Testing spend and scaling spend answer different questions and should be reported separately. When they are blended, promising tests get killed early for missing an efficiency target they were never meant to hit, and tired winners keep running because the average still looks acceptable.

  • Report test and scale performance in separate lines, never a blended average.
  • Give each test a clear decision date and a minimum data threshold before you judge it.
  • Promote a test to scaling only once it has cleared your mCPA on meaningful volume.
  • Retire scaling spend the moment efficiency drifts past your ceiling for two reporting periods.

Decide where money genuinely moves the needle

Not every channel deserves an equal share. Map your spend against where incremental revenue actually comes from, not where the last click happened to land. Retargeting and brand search often look like heroes in a last-click report while contributing little genuinely new demand. Prospecting and upper-funnel work frequently do the heavy lifting but get under-credited. Budgeting well means paying for incrementality, not for attribution artefacts.

Put the strategy to work

Browse the tools we actually recommend for planning, testing and scaling paid media — chosen because they earn their place in a real budget.

Explore the tool stack →