A £20,000 monthly ad budget can look healthy in a platform dashboard while quietly destroying margin. Revenue is rising, clicks are cheap and conversion volume is up, but returns, discounting, shipping costs and poor product mix tell a different story. That is the central challenge of paid search for online retailers: generating sales is not the same as generating profitable growth.
For established eCommerce brands, paid search should be treated as a commercial system, not a traffic source. The job is to put the right products in front of buyers with intent, control what is wasted, and scale only where the numbers support it. Everything else is noise.
Paid search for online retailers starts with economics
Before changing a campaign, establish what a viable sale looks like. A target return on ad spend is not universal. A 4x ROAS may be excellent for a high-margin own-brand product with repeat purchase potential, yet unacceptable for a low-margin product with expensive fulfilment and a high return rate.
Start with contribution margin, not a headline revenue target. Account for cost of goods, VAT where relevant, fulfilment, payment fees, delivery subsidies, returns and promotional discounts. Then decide how much of the remaining revenue can be spent acquiring a customer. This gives you a breakeven cost of sale or minimum viable ROAS.
That figure should guide campaign decisions, but it should not become a blunt instrument. Some products deserve investment below their immediate target if they introduce high-value customers or support a proven repeat-purchase model. Others should be capped or excluded even if they produce sales, because every order loses money after costs.
The stronger your commercial inputs, the better your media decisions. If an agency or in-house team cannot explain which products can scale profitably and why, they are managing platform activity rather than managing growth.
Product data determines who sees your ads
For retail brands, the product feed is often the biggest missed opportunity in Google Shopping and Performance Max. It is not admin. It is the information Google uses to decide when, where and against whom your products appear.
Weak titles, generic descriptions, incomplete attributes and inaccurate pricing restrict visibility before bidding even enters the conversation. A well-structured feed makes products easier to match with high-intent searches, while clearer segmentation exposes which categories, brands, price points and margins are carrying the account.
A practical feed strategy usually means improving titles around real search behaviour, supplying accurate GTINs and product attributes, maintaining clean images, and ensuring availability and price are always correct. Apparel, for example, needs reliable size, colour, gender and material information. Homeware may depend more heavily on dimensions, finish, style and brand. The right fields depend on the category, but missing data is rarely harmless.
Feed optimisation also makes budget control possible. If all products sit in one undifferentiated campaign, your best sellers can consume spend while profitable but lower-volume ranges receive no meaningful exposure. Separating products by commercial priority creates room to bid with intent.
Build campaigns around decisions, not platform defaults
Google’s automation can be highly effective, but it needs clean inputs, enough conversion data and firm commercial guardrails. Letting a single Performance Max campaign absorb the catalogue and budget is convenient. It is not automatically strategic.
Campaign structure should answer clear questions. Which product groups can tolerate higher acquisition costs? Which ranges need seasonal support? Where is brand demand protecting profitable revenue, and where are generic searches creating incremental customer acquisition? Which products are consistently wasting budget?
Search campaigns remain valuable because they offer control over intent. Brand search often converts efficiently, but it can overstate performance if it simply captures customers already looking for you. Non-brand search can introduce new demand, yet it will usually be more expensive and requires disciplined search-term management. Both matter, but they should not be judged as if they do the same job.
Shopping and Performance Max are typically the engines for product discovery and scale. Their strength is reach across Google inventory and the ability to use product-level signals. Their weakness is reduced transparency compared with traditional search. The answer is not to reject automation. It is to build a structure that gives it profitable products, accurate data and a sensible budget allocation.
For some retailers, this means separate campaigns for hero products, high-margin ranges, clearance stock and seasonal collections. For others, a simpler structure performs better because conversion volume is limited. More campaigns do not create more control if they leave each campaign underfunded and data-starved.
Stop funding products that cannot carry their spend
The fastest route to a better account is often waste reduction, not a new campaign launch. Retail catalogues contain products that are out of stock too often, attract irrelevant queries, compete in overcrowded markets or simply lack enough margin for paid acquisition.
Look beyond account-level ROAS. Review performance at product, category and search-term level over a meaningful period. A single week can mislead, particularly for higher-consideration products or brands with uneven demand. But a recurring pattern of spend without profitable sales is a decision waiting to be made.
Actions may include excluding low-margin products from prospecting campaigns, reducing bids on weak categories, adding negative keywords, separating a strong product range from a poor one, or prioritising stock that can actually support demand. If products have different economics, treating them identically guarantees inefficient spend.
This is also where stock management and paid media must work together. There is little value in pushing a product that is nearly sold out, has a long delivery promise or is likely to be returned. The media team needs visibility of stock levels, promotions, merchandising priorities and supply constraints. Paid search cannot be profitable in isolation from operations.
Measure profit signals, not just platform revenue
Platform reporting is useful, but it is not the company accounts. Google Ads can report attributed conversion value while overlooking cancelled orders, returns, new versus existing customer mix and the impact of discounts. That does not make the data useless. It means it needs context.
At minimum, track accurate purchase revenue, transaction counts and spend. Better still, feed margin-aware data into your reporting and compare platform performance against your eCommerce platform, analytics and finance figures. Expect small differences in attribution. Focus on whether the direction of travel and commercial outcomes align.
It is also worth separating metrics by purpose. ROAS and cost of sale help control efficiency. New customer acquisition shows whether spend is expanding the customer base. Contribution margin reveals whether growth is worth having. Conversion rate can flag issues with landing pages, pricing, delivery or product-market fit. None should be used alone.
A retailer selling premium furniture may accept a longer decision cycle and higher cost per acquisition than a beauty brand selling replenishable products. A promotional period may justify temporarily lower efficiency to clear stock. The right target changes with your margins, stock position and customer lifetime value. Good management makes these trade-offs explicit rather than hiding behind an average account ROAS.
Scale only after performance is repeatable
Scaling is not increasing budgets because the previous seven days looked good. It is proving that a campaign can absorb more spend without a disproportionate drop in return.
Increase budgets progressively, then watch how spend distributes across products and queries. When a campaign scales, Google may move beyond your highest-intent audience into less efficient auctions. That is normal. The issue is whether the resulting marginal sales still meet your profitability threshold.
Keep a close eye on impression share, product-level spend, conversion lag and stock availability as budgets rise. Expand only when the account has room to find additional profitable demand. If branded search is already saturated, adding budget there will not create meaningful new growth. The opportunity may sit in stronger Shopping coverage, better feed quality, new category campaigns or a more compelling offer.
Creative and landing pages matter too, particularly where Performance Max and paid social support demand generation. But no advert can compensate for unclear delivery terms, weak product pages or prices that are uncompetitive for the market. Paid media amplifies what is already present. It exposes operational weaknesses quickly.
Demand accountability from your paid media partner
A retailer with meaningful ad spend does not need vague monthly commentary about impressions. It needs clear answers: where has spend gone, what has improved, what is being tested, what is being stopped and what needs to happen next?
That requires transparent access to ad accounts and data, realistic targets, and a willingness to challenge unprofitable activity. No-contract arrangements can reinforce that accountability, but only if the work itself is commercially rigorous. Retention should come from performance and clarity, not contractual friction.
Oxedent approaches eCommerce PPC with that standard in mind: channel specialists focused on profitable revenue, feed quality, campaign control and the removal of wasted spend. For brands past the early validation stage, that focus is more valuable than a broad agency menu.
The next productive move is not necessarily spending more. Put your margin targets, product priorities and current campaign data in the same room, then identify the spend that is earning its place. That is where profitable scale begins.