Your catalog is your content.
Most of it was never written.

With the product data, every time. Google maintains a dedicated ecommerce documentation section and publishes its product requirements in full, so most of what determines whether your items can be shown is knowable rather than mysterious. Media spent against a broken catalog buys the wrong things efficiently. We audit the data first and tell you how much of your catalog is genuinely sellable before we recommend spending anything — the same sequencing our ecommerce SEO practice follows.

Usually a mismatch between the feed and the page, or a missing required attribute. The product data specification lists exactly what is required for each category, and the shopping ads policies govern what may be advertised. What most teams miss is that items can be quietly suppressed rather than formally disapproved, which produces no alert at all. That is why we compare the feed against the live page rather than only against validation — the approach on our Shopping page.

No, and attempting it is how catalog content projects die. Start with the products carrying your revenue and margin, which in most catalogs is a small fraction of the items. Supplier copy is identical across every retailer selling that product, so there is nothing for a search engine to prefer or an assistant to cite you on — but that only matters where the product can actually earn. Google's guidance on helpful, people-first content applies as much to a product page as an article. Our content team handles the writing.

Deliberately, because they multiply silently. A product in six sizes and four colours can generate two dozen near-identical URLs that compete with each other and dilute whatever authority the product had. Google's guidance on consolidating duplicate URLs covers the mechanics, and the right answer depends on whether shoppers genuinely search for the variant or only for the product. Getting it wrong is one of the most common and most expensive catalog defects. Detail on ecommerce SEO.

Because return on ad spend treats every dollar of revenue as equal and your accountant does not. A campaign optimized on ROAS grows whichever products convert most readily, which in most catalogs are not the ones you earn most on — so the metric improves while the sales mix gets worse. Where a partner cannot supply margin we say plainly that we are optimizing a proxy until they can, rather than pretending the proxy is the goal. The full approach is on our ecommerce PPC page, and the reconciliation runs through your systems, against the campaign types documented in Merchant Center's product category guidance.

Assistants increasingly summarize product options before a shopper reaches any store, and those summaries are assembled from structured product data — the same attributes that keep you eligible in shopping surfaces. So clean product data now pays twice, once in the auction and once in whether an assistant names your product. The work has not changed; the cost of skipping it has. We do not claim to control what any assistant says, because nobody does — the honest position is on our GEO page, and Google's product markup guidance is the foundation.

Your store, for the purpose of deciding anything. Ad platforms report conversions they believe they caused, using their own attribution windows and models; your commerce platform records orders that actually happened. The gap is normal and it is not a rounding error. What matters is that somebody reconciles them monthly and reports the difference rather than quoting whichever number flatters the work. We also net out returns, because revenue that came back is not revenue — the same standard applied across every paid engagement and grounded in measurable page experience on the store side.

It is usually where the record breaks first, because marketplaces are typically fed by a different integration than your own site. A price change or a discontinued variant reaches your store immediately and the marketplace on a schedule nobody remembers setting, so a shopper finds two different prices under your own brand. The answer is one authoritative product record with everything downstream reading from it, plus a reconciliation step on a schedule rather than when somebody notices. Google's shopping policies treat accuracy as a requirement, not a preference. Keeping the destinations in step is part of our record distribution discipline.

They carry obligations worth knowing about. The FTC's rule at 16 C.F.R. Part 465 addresses consumer reviews and testimonials, and the endorsement guides cover incentivized and affiliated reviews in detail. Google's review snippet guidance governs how they may be marked up. In practice the rules reward what a good store does anyway — collect reviews honestly, disclose incentives, do not suppress unfavorable ones selectively. Nothing here is legal advice and your counsel governs your position. Our approach to review programs is on the reputation page.

Data corrections can move quickly, because an item that was ineligible becoming eligible is a step change rather than a climb. Content and organic authority take longer and vary by category and competition. Paid can move immediately and will simply spend faster on a broken catalog, which is why sequencing matters. We will not quote a single timeline for a whole catalog — a product with a missing attribute and a product in a saturated category are different problems on different clocks. The A.R.C. Report tells us which is which, and Google's own ecommerce guidance is candid about the same point.

Yes, and it is a large part of who we do this for. Manufacturers with parts catalogs, franchise systems selling product alongside service, practices with retail lines and mid-market operators adding a commerce arm all face the same product data problem without a dedicated ecommerce team behind it. The catalog discipline is identical; what changes is that the store supports the business rather than being the business, which usually simplifies the margin question considerably. The same feed requirements apply whether the catalog is your whole business or part of it. Related: manufacturing and franchise systems.

Three things you can verify. We reconcile page, feed and markup item by item rather than checking the feed alone. We ask for margin by product line before structuring anything and tell you when we are optimizing a proxy instead. And we reconcile platform-reported conversions against orders in your own system monthly, net of returns, including when the gap is unflattering. We are also built as an AI-first agency rather than a traditional shop with AI added on. Start with the A.R.C. Report, or read the FTC's advertising guidance against whatever your current agency claims.