Small budgets fail for one reason: they get spread so thin that nothing gathers enough data to optimise. Below is the split I start with, and why.

The 60 / 25 / 15 starting point

  • 60%: high-intent search. People already looking for what you sell. Lowest risk, fastest read on whether the offer converts.
  • 25%: retargeting. Cheap, and it protects the traffic the other 75% paid for.
  • 15%: cold social. Treated as research, not revenue. It tells you which message earns attention.

Give each line item enough to learn

A campaign needs roughly thirty conversions a month before the numbers mean anything. If a split leaves a campaign below that, merge it into another one. Fewer campaigns, more signal.

Two campaigns that learn beat six that guess.

Review on a fixed rhythm

  1. Weekly: spend pacing and obvious waste, such as search terms, placements, frequency.
  2. Fortnightly: creative and ad copy rotation.
  3. Monthly: the split itself, moved by no more than 10 points at a time.

The reporting side of this lives in the five metrics I report every month.