AI-integrated web design & SEO
Industries
Shopping ads have no ad copy. The title, the image and the price are the creative, and they come out of your product feed — which is why feed work usually moves more revenue than campaign work does.
What you can expect
Product feed treated as the primary creative asset
Faceted navigation and index bloat fixed at the source
ROAS targets set against margin, not revenue
Service focus
In Shopping campaigns you do not write an advert. Google builds it from your Merchant Center feed, which means the product title, the main image, the price, the availability status and the attributes are your creative — and most stores are running whatever their platform exported by default. That is where the fastest gains usually sit. Titles that lead with the brand and the product type rather than an internal SKU description. Correct product identifiers, because missing GTINs limit where an item can serve. Colour, size, material, age group and gender populated, because those attributes decide which queries you can match. Images that meet the requirements without watermarks or promotional overlays. Prices and stock that match the landing page, because mismatches cause disapprovals that silently remove products from the auction. A store can spend months tuning bids on a feed where a third of the catalogue is ineligible and never see the cause in the campaign reports.
The technical SEO problem is different from every other sector because it compounds. A brochure site has thirty pages; a catalogue has thousands, and every structural mistake is multiplied by the product count. Faceted navigation is the usual culprit. Filters for size, colour, price band and brand combine into an effectively unlimited set of URLs, each one crawlable, most of them near-duplicates of a page that already exists. Crawlers spend their time on parameter combinations nobody searches for, real product pages get crawled less often, and the index fills with thin variations that dilute the pages you actually want ranking. The fix is a decision, not a plugin: work out which facet combinations have genuine search demand and deserve a real indexable page, then make sure the rest are consistently handled — canonicalised or blocked, but not both at once, since a URL blocked in robots.txt can never be read to discover the canonical you set on it. Add discontinued products, variant duplication, paginated series and manufacturer descriptions copied across the industry, and you have the four issues that account for most lost ecommerce organic traffic.
The economics are where good campaigns still lose money. A revenue-based ROAS target ignores the two numbers that decide whether a sale was worth making: gross margin and returns. At a 20% gross margin you need a 5x return on ad spend simply to stand still, and that is before you have paid for picking, packing, delivery or the proportion that comes back — and in clothing and footwear the return rate is high enough to change the answer entirely. Different products in the same account therefore need different targets, which is what custom labels in the feed are for: margin bands, seasonality, stock depth, clearance. The same discipline applies to channel roles. Shopping captures people who already know the product, Search covers category and problem-led intent plus defending your own brand name against resellers, email and retention carry the repeat purchase, and Performance Max blends all of it while giving you far less visibility into what it bought. Retail seasonality then concentrates the risk: a large share of the year's trading arrives between Black Friday and Christmas, so a feed or tracking problem discovered in November is not a technical issue, it is a lost quarter.
Common challenges
01
Default titles built from internal SKU names, missing identifiers, empty attributes, stale stock status and a tail of silent disapprovals. Nothing in the campaign interface tells you this is the problem, so budget gets moved between campaigns that were never the constraint.
02
Filter and sort parameters generate near-limitless URLs, crawlers spend their budget on combinations nobody searches, and product pages that matter get visited rarely. On a large catalogue this is usually a bigger organic problem than anything on the page itself.
03
A single account-wide target applied to products with wildly different margins guarantees you overspend on the thin-margin lines and underspend on the profitable ones. Then returns come out of the profit and never out of the reported ROAS.
Service breakdown
Title structure, identifiers, attribute completeness, image compliance, price and availability accuracy, and a clear-out of the disapprovals that have been quietly removing products from the auction. Then custom labels so margin, season and stock depth can drive bidding rather than a single blanket target. This is where we start on almost every retail account, because it changes what the campaigns are able to do.
Crawl and index control for faceted navigation, canonical strategy for variants, a policy for discontinued and out-of-stock products, pagination handling, and internal linking that pushes authority to the categories worth ranking rather than spreading it across parameter URLs.
Separating campaigns by the job they do, protecting brand terms, and keeping enough visibility into search terms and product performance to make decisions rather than accept the black box.
Category pages written for the way people search rather than a copied manufacturer paragraph, structured data that supports rich results, and product content that answers the questions your returns data says people got wrong.
Server-side or enhanced conversions where consent and tracking loss are distorting the data, revenue values passed correctly, and new-customer versus repeat-purchase reporting so acquisition is not being credited with sales you already had.
Email and post-purchase flows, because the second order costs a fraction of the first and is the number that makes acquisition affordable.
We will review your site, your search visibility and your competitors, then send back the specific fixes worth doing first.
Why us
Without gross margin and return rate by category, any ROAS target is a guess. If you cannot supply them yet, that is the first piece of work, and we will say so.
Bid adjustments on a broken feed and content work on a site that cannot be crawled properly are effort spent where it cannot compound. We fix the constraint first even when it is less visible in a monthly report.
Nothing structural gets changed in the run-up to peak trading. Migrations, replatforming and crawl-directive changes belong in January to September, and we will push back on a December deployment.
Explore
Related services
Browse more
FAQ
There is no ad copy in Shopping. The advert is assembled from the feed, so the levers are the product title, the image, the price and the attributes. If a competitor's identical product is beating you, it is usually because their title matches the query more precisely, their image is cleaner, their price is visibly lower, or your item is limited by a missing identifier or an unnoticed disapproval. Start in Merchant Center diagnostics rather than in the campaign.
Some of them. A filter combination with genuine search demand — a colour and product type people actually type — deserves a real page with its own title, description and stable URL. The long tail of price bands, sort orders and multi-select combinations does not, and letting it into the index costs you crawl efficiency and dilutes the pages you care about. The rule that saves the most trouble: pick one control per URL pattern. Do not block a URL in robots.txt and also rely on a canonical or noindex on it, because a blocked page cannot be read.
It often performs well and it costs you visibility. The common failure is that it absorbs brand searches and existing customers, then reports them as acquisition, which makes the account look better than the business is doing. If you use it, keep brand traffic separated where you can, watch new-customer share rather than blended ROAS, and make sure feed segmentation lets you steer it toward the products you actually want to sell. For stores with a small catalogue or thin margins, a structured Shopping and Search setup is usually easier to control.
Whatever break-even is for that product, plus the profit you need. Work it from gross margin: at a 20% margin you need roughly 5x just to cover the cost of goods and the ad spend, before delivery, payment fees and returns. At a 60% margin the same 5x is comfortably profitable. That is why one account-wide target is nearly always wrong — it overspends on your worst products and starves your best. Segment by margin band in the feed, and deduct returns before you judge any category, particularly in clothing and footwear.
Free consultation
Tell us what you want to grow and we will review the best next steps for your website, search visibility, and enquiries.