Splitting budget between performance and brand

Picking a ratio first is the mistake. A ratio is the answer to a question about your market, your category's buying cycle and how much demand already exists for what you sell. Those vary enormously, which is why borrowed percentages produce confident nonsense.

Start instead with three questions.

Is there demand you are not buying? If people are searching for your category and your competitors are winning those auctions while you sit at a fraction of available impressions, you have unbought demand. Buy it. Brand investment that creates more demand you then fail to capture is money spent twice.

What happens when you add budget to capture? This is the real test. Increase spend on your best-performing capture channel. If cost per acquisition holds roughly steady, you had headroom. If cost climbs quickly, you have exhausted the demand that exists and you are now paying a premium to compete harder for the same finite pool. That is the signal to fund creation.

How long is your buying cycle, and can your cash survive it? Demand creation pays back over a horizon longer than a performance click. A business that cannot fund a longer payback period should not start with brand, regardless of what the theory says.

The fourth row catches more teams than any other. A conversion problem presents as a budget argument, and reallocating money never fixes it. Check the funnel before the split.

1. Establish the acquisition ceiling. Derive it from contribution margin and payback tolerance. Everything downstream references this number.
2. Fund capture to the point of diminishing return. Raise budget on capture channels in steps, holding targets constant, and record the cost per acquisition at each step. Stop when the next increment pushes cost past the ceiling. This is the only part of the exercise with a clean answer.
3. Record the capture spend level. That figure is your capture allocation. It is not a percentage you chose; it is where the market stopped paying.
4. Set a creation floor. The remaining budget available for creation has a minimum below which it does nothing. Reach spread too thinly across too many channels produces no memory. Pick one or two channels and one message.
5. Hold the creation allocation for a full buying cycle. Reviewing brand spend monthly against last-click returns guarantees you cancel it. Agree the review horizon before it starts, in writing.
6. Re-run step two each quarter. Successful creation raises the demand ceiling, which means capture has headroom again. The split should move. That is the system working.

The output is a split derived from observation. It will not be a round number, which is a good sign.

Les Binet and Peter Field's analysis of IPA Effectiveness Awards data popularised a roughly 60/40 brand-to-activation split as a long-run average across the cases studied. The finding is genuine and the direction is well supported: activation alone tends to deliver short-term response without building the base that makes future activation cheaper.

Two cautions. It is an average across many categories, weighted toward mature brands in broad consumer markets, and it describes long-run effectiveness rather than what any single business should do next quarter. A young company with unbought demand and a short runway is not the subject of that dataset. Cite it as evidence that brand investment has a return, not as your allocation.

Brand spend dies in budget reviews because it has no defence that uses the same vocabulary as performance spend. Build one in advance.

• Name the job. Not "awareness." Something like: increase the number of people who can name us unprompted when they next need this category. A stated job can be assessed.
• Name the horizon. "We will not judge this on monthly cost per acquisition. We will assess it at two quarters on these indicators."
• Name the indicators before you start. Branded search volume, direct and organic session trend, share of total conversions arriving without a paid click, unaided recall if you have the budget to survey, and the trend in capture efficiency over time.
• Name the kill condition. What result would make you stop. A brand program with no kill condition is not a plan.

• Geographic holdouts. Pick matched regions, withhold creation spend in some, run it in others, and compare total sales and branded search. This is the most practical incrementality test available to most businesses, and it works without any cookie.
• Branded search as the leading indicator. Imperfect and influenceable by competitors, but it moves before revenue does and it is free to track.
• Baseline versus incremental. Over a long enough window, look at what sells with no paid click attached. A rising baseline is what successful creation looks like.
• Capture efficiency over time. If creation is working, your capture channels should get cheaper or scale further at the same cost. That second-order effect is often the strongest evidence.
• Platform attribution is the wrong instrument. A last-click report will systematically undervalue anything that worked weeks before the purchase. Do not hand your brand program to a tool that cannot see it.

• Both budgets judged on one metric, so the longer-horizon work always loses.
• Brand search campaigns counted as brand building. They are capture, and cheap capture, which is why they flatter the whole account.
• Creation spread across five channels at a level too low to register anywhere.
• A creative message that is correct and forgettable. Reach without distinctiveness is a cost without an asset.
• Reallocating in response to a single bad month, resetting the clock on both sides.
• No baseline captured before the creation program started, making any later assessment arguable.

The split is a planning decision, so it belongs in the diagnostic phase, not in a channel review. We run the diminishing-return test on capture, establish the baseline indicators, and agree the horizon and kill conditions before any creation budget moves. That work sits in digital strategy and advisory, and the capture side of it is executed through performance marketing. Agree what better means before the work starts.