A Google Ads account can show a healthy ROAS and still make the business poorer. That is the uncomfortable reality behind Google Ads profitability for eCommerce brands. Revenue is easy to report. Profit depends on what remains after product cost, fulfilment, transaction fees, discounts, returns, VAT treatment and advertising have taken their share.
For an established retailer, the question is not whether Google Ads can generate more sales. It usually can. The question is whether the next £10,000 of spend produces contribution profit, strains operations, or merely buys revenue that would have arrived anyway. Scaling before answering that question is how apparently successful accounts become expensive problems.
Start with the numbers Google Ads cannot see
Google Ads sees conversion value and media cost. It does not automatically understand your landed product cost, pick-and-pack charges, returns rate or whether a sale came through a discount that destroyed its margin. Platform ROAS is therefore a useful signal, not a profitability report.
The commercial starting point is your breakeven ROAS or, for many brands, your target cost of sale. Both tell you the maximum you can afford to pay to acquire a sale before paid media ceases to contribute profit.
If a £100 order leaves £30 in contribution before advertising, your maximum profitable cost of sale is 30%. Your breakeven ROAS is 3.33x. Spend £30 to generate £100 in revenue and you have covered advertising, but generated no profit to fund overheads or growth. A sensible target must sit above that threshold.
That target will not be universal. A high-repeat-purchase skincare brand may choose to acquire a first order at a lower margin because it can evidence profitable retention. A furniture retailer with low purchase frequency cannot make the same assumption. Nor should a business apply one target across a catalogue where gross margins range from 20% to 70%.
Work from contribution margin by product or product group, not a blended figure that hides weak economics. Include the costs that genuinely vary with the order: cost of goods, delivery subsidy, packing, payment processing, expected returns and promotional discounts. Then decide how much contribution needs to remain after ad spend.
Why blended profitability matters more than campaign ROAS
A brand can improve a campaign’s reported ROAS simply by concentrating spend on branded searches and existing customers. The account looks efficient, but incremental growth may be limited. Equally, prospecting campaigns can appear weaker in-platform while introducing customers who later buy again through email, direct traffic or branded search.
This is why Google Ads profitability needs two views: platform performance for day-to-day optimisation, and blended business performance for investment decisions.
At platform level, assess revenue, cost of sale, product-level margins, search terms, conversion rate and the quality of traffic. At business level, compare total paid spend against total revenue, contribution profit, new-customer volume and repeat purchase behaviour over a meaningful period. Neither view is sufficient on its own.
Attribution will never be perfect. Google can over-credit itself, particularly when demand is already strong, while last-click reporting can under-credit upper-funnel activity. The right response is not to dismiss the data. It is to avoid treating any single dashboard as absolute truth.
Use consistent reporting windows, monitor changes in total revenue and margin as spend increases, and test carefully. If you increase non-brand spend by 20% and total profitable revenue barely moves, the account is likely harvesting demand rather than creating enough new demand. If total contribution rises faster than spend, there is a stronger case for further investment.
Find waste before asking for more budget
Most eCommerce accounts do not need a bigger budget first. They need a clearer view of where budget is leaking. Wasted spend is rarely one dramatic error. More often it is a collection of small inefficiencies: irrelevant queries, low-margin products, weak feeds, indiscriminate bidding and campaigns built around convenience rather than commercial control.
Search terms still deserve attention
Broad match and automated bidding can perform exceptionally well when conversion data, product economics and campaign structure are sound. They can also expand into expensive, low-intent searches when left unchecked. Review search term themes against actual margin, not just clicks and conversion count.
Negative keywords have a role, but they are not a substitute for strategy. The more useful question is whether the campaign is being asked to chase the right type of demand. A profitable search term for a premium product may be completely unsuitable for an entry-level range, even when both technically convert.
Product segmentation protects margin
Treating every SKU identically is one of the fastest ways to blur profitability. Best sellers, high-margin products, clearance stock, low-stock items and poor converters should not automatically receive the same bid logic or budget priority.
Segment campaigns and product groups where there is a genuine commercial reason to do so. High-margin, proven products may justify aggressive investment. Products with narrow margins may need a stricter cost-of-sale target, a different landing page, or exclusion from paid activity altogether. Do not build a complicated structure simply because it looks sophisticated. Build one that allows decisions to be made and controlled.
Feed quality is a profit lever
For Shopping and Performance Max, the product feed is not admin work. It is a sales asset. Product titles, categories, attributes, pricing, availability, images and promotional annotations affect which searches your products enter and how persuasive they are once shown.
Poor feed quality often forces the account to pay for less relevant traffic. A precise title that includes the attributes buyers use to search can improve relevance without increasing bids. Correct product categorisation and clean custom labels make margin-based segmentation possible. For retailers with hundreds or thousands of products, this work can have more impact than another round of superficial campaign tweaks.
Scale only when the marginal return holds
A campaign that delivers a 6x ROAS at £2,000 per month may not deliver 6x at £20,000. As spend increases, Google moves beyond the cheapest and most obvious conversions. That is normal. The issue is whether marginal returns remain above your target, not whether the original headline metric is preserved forever.
Scale in controlled increments and allow enough time for data to settle. The right pace depends on conversion volume, seasonality, stock position and how volatile demand is. A brand with thousands of weekly transactions can test budget increases faster than one generating twenty orders a week.
Monitor the additional spend and additional revenue, rather than admiring the account average. If an extra £3,000 in spend creates £15,000 in revenue at a profitable marginal cost of sale, continue. If it creates £6,000 at a loss, pull back and diagnose the constraint. It may be audience saturation, weak creative, a poor feed, an uncompetitive offer or simply a market with limited search volume.
Performance Max deserves the same commercial discipline. It can expand reach quickly, but it should not become a black box that absorbs budget because reporting is less granular. Use feed segmentation, asset quality, exclusions where appropriate, and clear profitability targets. Assess whether its growth is incremental, especially when branded search demand is material.
Make measurement fit the way you trade
Profit-first advertising requires accurate conversion values. Sending the same static value for every purchase is inadequate when order values vary. Sending gross revenue without accounting for cancellations, refunds or discounts can be worse.
Where possible, pass transaction-level revenue accurately, reconcile platform data with your eCommerce platform, and review refunds after an appropriate lag. If your margins vary materially, use custom labels or product reporting to bring those differences into optimisation decisions. The goal is not perfect data – that is rarely available – but data good enough to stop profitable products subsidising unprofitable ones.
Also separate new and returning customers where your technology permits it. A returning customer buying a product they already intended to purchase has a different value from a first-time buyer who may become a loyal customer. If lifetime value is being used to justify lower first-order profitability, it must be based on real cohort data, not optimism.
Accountability is the real growth mechanism
Profitable Google Ads management is not about chasing a fashionable campaign type or reporting the biggest revenue number. It is a repeated commercial process: establish the margin thresholds, structure campaigns around product economics, eliminate waste, improve the feed, test expansion and judge success against contribution.
That process requires honest conversations. Sometimes the best decision is to reduce spend while the offer, website conversion rate or stock availability is fixed. Sometimes a lower ROAS is the right decision because it produces more profitable new customers. And sometimes the account is performing well, but the business cannot yet fulfil the demand without damaging customer experience.
The strongest next step is simple: take your real margin data, calculate the cost of sale you can afford, and compare it with what each meaningful product group is actually delivering. Growth becomes far easier to fund when every additional pound of ad spend has a job to do.