Good ROAS. Should you raise the budget?

Start with what goes into the numerator.

ROAS compares attributed sales value with advertising spend. Ask what value reaches that calculation: purchases, in which currency, over what period and with which attribution? Google supports different conversion types, so a column label is not a substitute for checking its contents.

Then consider campaign roles. People already searching for your name and people discovering you for the first time start from different places. Reading everything together can obscure which part of the result you are actually hoping to expand.

Put products and costs beside the ads.

Two retailers can report the same ROAS while operating very different businesses. Product costs, delivery subsidies, returns and other relevant costs change what remains. The products and orders producing the value matter more than an account average alone.

Avoid turning this into a decorative calculation. Compare real orders from the period with the report using consistent definitions. Where numbers differ, understand why before using the gap to justify more spending.

Check whether more orders can be fulfilled.

Is the promoted item still in stock? Can the team deliver? Does the offer still apply? If the result came from a temporarily available selection, a larger budget for the same ad may send more people towards a choice that is no longer there.

This is also a conversation between marketing and operations. The agency needs to hear about stock changes, timing and bottlenecks. In a useful working relationship, that information arrives before the budget decision rather than after the first complaints.

What can you afford to pay for an order?

Take order revenue on a consistent tax basis and subtract variable costs: product cost, fees, packaging, your share of delivery and average return costs. What remains is contribution before advertising. It still needs to cover fixed costs, the agency fee and the profit you want to retain.

Spend all that contribution on ads and nothing remains for the rest. Breaking even on the variable cost of acquisition is not automatically a healthy business target. The calculator in the website version of this edition makes the distinction using your own figures.

Separate products with very different cost structures. A store-wide average can hide ads pushing the category that leaves the least money behind. A bigger basket is not necessarily a better basket.

The next pound of spend can behave differently.

Your current result includes the customers found at your current spending level. Raising the budget does not buy that result a second time. You may reach people who need longer to decide, different placements or more expensive auctions.

Look at the additional contribution left after the additional advertising spend. Don’t rely only on the account average remaining attractive. Record changes to price, stock, promotions and attribution too, or you may credit the budget change for something caused elsewhere.

If the first order loses money and the plan relies on repeat purchases, check actual customer cohorts. When do they return, what do they buy and what margin remains? Until you have that evidence, making the money back on a later order is an assumption funded by today’s cash.

Profitable orders can still drain cash.

The supplier needs paying before you collect from the customer. Returns settle after the ad report has counted the sale. Growth increases the money needed between those moments. The advertising account cannot show your stock financing on its own.

Bring the payment calendar into the budget conversation: when you pay for inventory, when you can deliver and when you collect. Growth you can finance for a few days may be difficult to sustain for a month. That is a capital and timing problem even when an individual order has a healthy margin.

Increase with a question, not an assumption.

Agree what to observe after the change and a sensible review period. Follow orders, costs and availability. Do not assume that the previous result will scale in direct proportion to the budget.

In a food project from our portfolio, Google looked stronger than the overall month. We separated brand traffic, examined the decline on Meta and reallocated the Google budget while keeping the configured total unchanged. The useful question was where existing money deserved to go.

Our answer to “should we increase?” starts with three things on the table: what produced the result, what remains for the business and what operations can support. Once those are clear, the next allocation becomes a much better conversation.