eCommerce SEO Services

By not quoting per page. We have worked on catalogs ranging from about 900 products to over 60,000, roughly 93,000 products in total, and the per-page model breaks somewhere in the low thousands. Most catalog content already exists, it is just held by your manufacturers rather than by you: specifications, materials, dimensions, certifications, applications. We build an intake process that collects it in a format which uploads directly, so the constraint becomes how fast suppliers respond rather than how many hours we bill. On one account that turned a writing project nobody could afford into a collection process that has been running continuously since.

You keep it. The intake process is the deliverable as much as the pages are: a structured sheet your suppliers fill in, and a route that turns what they return into published pages in bulk, formatted consistently. We have built this on more than one catalog, at different sizes and on different platforms, and in each case it kept running after our involvement changed. That consistency is worth as much as the content itself, because a catalog where every page follows the same structure is readable by search engines, quotable by assistants and comparable by buyers. A catalog written by hand over five years is none of those things.

Category pages, almost always. They are usually the biggest block of wasted surface area on the site, a heading and a product grid with nothing a search engine or an assistant can use, and they carry the head terms. On one account 71 of them were rebuilt in the first pass. Product pages come next, prioritised by demand rather than by position in the catalog.

Not on its own, and we know because we argued that it would and were wrong. On a live account a fully enriched page with over a thousand impressions of demand was converting at zero. The enrichment had worked on visibility and done nothing for conversion, because the page still showed no reviews, was case-only in minimum order quantity, and stated no price advantage. Being found and being chosen are different problems. We will tell you which one you have.

Often the site is not really a storefront. On one client we reconciled analytics against the full order ledger and found the referral channel converting at 65.7%, which is not shopper behaviour, it is existing customers reordering from a supplier they already use. That single finding changed every recommendation on the account, because the campaigns that looked strongest were mostly defensive. If we do not know what your site actually does, we are optimising in the dark.

The ones our clients run: BigCommerce, WooCommerce, and Commercebuild, where we have seven and about ten years of platform-specific experience respectively. The platform matters far less than most people expect. Catalog problems are remarkably consistent across all of them, because they come from how product data is managed rather than from the software it sits in.

Yes, and one of our longest engagements started exactly there. A hardware business came to us in 2016 with a showcase website that sold nothing, and four months later it was a B2B and B2C storefront integrated with the ERP the company already ran on. The integration mattered more than the launch: product data coming from the system of record rather than from a spreadsheet is what made eight subsequent years of organic growth possible. Launching a store on top of unmanaged product data just moves the problem.

Not necessarily, and one of our longest engagements ran that way deliberately. The business had no digital marketing budget at all, so organic search had to carry both lead generation and online sales. Paid came years later, only after the organic numbers proved which products and categories actually converted, and it started with shopping campaigns at the bottom of the funnel. That sequence is not a compromise forced by poverty, it is the cheaper way round. Pointing a paid budget at catalog pages before you know which ones convert is the most expensive form of market research available.

No, and the longest engagement we have run ended precisely because it did not need us. We helped that client hire a digital marketing specialist and a product information and content manager, trained both, and stepped back once they could run the program themselves. On another account we recruited the client’s own digital project manager, whose job was explicitly to reduce their dependence on us. Catalog work is well suited to this, because most of it becomes a repeatable process rather than a permanent skill shortage. If we are still indispensable after three years, something has gone wrong.

Then you have two sources of truth and a decision to make quickly, because the cost of deferring it compounds. On one account the client acquired a competitor and we consolidated both catalogs into a single product information system, which then fed both websites. Over seventy thousand products, one place where a specification is correct. The alternative, maintaining product data separately per site, fails quietly: the two stores drift, the same product is described differently in two places, and eventually nobody trusts either.

Almost never, and we will usually argue against it. Replatforming is expensive, it puts hard-won organic traffic at risk, and most of the problems we are called in for turn out not to be platform problems. If your catalog is invisible on one platform, the same catalog will usually be invisible on the next one.

Batches, baselines and a control group of pages we deliberately did not touch. On one account a 23-page batch produced a 507% rise in non-brand impressions. On the same account we validated the conversion tracking end to end against the client’s invoices, matching 99 of 108 orders to within 0.15% of the billed total. Without that validation the rest of the reporting would be decoration.

It is the same work, read by a different consumer. Assistants recommending products need structured, parseable, attributable product data, which is exactly what a properly built catalog has. The measurement differs completely though, so we report it separately. See our answer engine and generative engine pages for that half.

That is the normal case in distribution and it is the reason this is a supply problem rather than a writing one. The manufacturers hold the content and generally want their products described accurately. Our job is to make supplying it effortless for them and usable for you, which is a templated process rather than a negotiation.