AdAce AdsReview brand and nonbrand budgets without confusing demand capture with growth
A low account CPA can hide an acquisition problem. People who already know your business may be converting cheaply while unfamiliar prospects cost much more. Before reallocating money, distinguish capturing existing demand from creating an opportunity with someone new. This review is a proposed budgeting method, not a claim that branded advertising is unnecessary.
Separate two jobs inside one account
Build a classification from the business, not from campaign names. Brand includes your trading names, common variants and product names that genuinely identify your business. Nonbrand describes a category, problem or service without that identifier. Keep competitor searches and ambiguous expressions in separate buckets. A campaign called Prospecting can still contain brand demand; an expression that resembles a company name might be an ordinary product term.
Use a consistent business result: qualified enquiries, completed purchases or another defined outcome. Match the currency, attribution basis and reporting maturity before comparing costs. Existing customers should be marked separately when reliable records permit it. A branded query does not prove someone is an existing customer, and a nonbrand query does not prove someone has never heard of you. Preserve an unknown category rather than manufacturing a precise acquisition split.
In an educational example, brand spends $600 for 30 qualified enquiries and nonbrand spends $2,400 for 40. Their CPAs are $20 and $60; the blended CPA is $3,000 / 70, or about $42.86. If the account target is $45, the blend looks acceptable, but it does not establish that the next $600 should go to brand. Brand may already cover most available demand, and those conversions may partly reflect earlier activity elsewhere.
Write down the intended role of each bucket. Brand may protect a clear customer journey or communicate a current offer. Nonbrand may discover profitable new demand. Compare their observed outcomes with those roles, then ask whether an extra unit of spend has a plausible useful destination. Google documents delayed conversion reporting, so recent apparent differences need a maturity check before they become a budget conclusion.
Three lines a budget memo needs
- A capture line reports branded demand, its cost and the known customer mix. It stays useful without pretending that every attributed sale was caused by the ad.
- An acquisition line reports nonbrand outcomes and downstream quality. It makes a more expensive lead visible when that lead is also more valuable or more likely to buy.
- A decision line states what additional spend is intended to achieve, which evidence supports it and which uncertain assumption could reverse the recommendation.
Build a reproducible brand review
- Define the brand dictionary with someone who knows the business. Include ambiguities, ownership changes and a date so later analysts can reproduce the grouping.
- Classify the available search evidence and report unclassified spend. Do not assume the visible search-term rows explain every click in the account.
- Calculate spend divided by compatible outcomes for each bucket. Attach the result definition and the conversion maturity cutoff to the comparison.
- Reconcile qualified outcomes or purchases with aggregate business records. Separate customer status from search classification instead of treating one as a substitute for the other.
- Review demand limits, lead handling capacity and contribution after fulfilment. A cheap bucket with little remaining demand cannot absorb unlimited budget usefully.
- Propose one bounded investigation or test, with an owner and review date. Record the blended result too, so a channel-level improvement cannot conceal a worse business total.
What the split cannot prove
- The educational amounts above demonstrate arithmetic only. They are neither a client case nor a forecast of what a budget change will produce.
- Attributed brand conversions do not measure incrementality. A causal answer needs an appropriate comparison design and attention to other channels and seasonality.
- Missing query or customer-history data should reduce confidence. If using an AI assistant, request read-only classification and a written plan, with no propose_change calls or other writes.
Sources and further reading
Try it on your own accounts
Create a workspace, connect Google or Meta in a couple of clicks and see your accounts clearly. Changes follow your approvals or the policy you configure.
Create your workspace