E-commerce SEO traffic strategy: why optimizing category pages pays far more than publishing blog posts

In many e-commerce teams, SEO strategy comes down to one sentence: "we need to publish content". A blog is launched, two posts go out every month, organic traffic rises. Six months later, revenue from SEO has barely moved.

This is not a copywriting failure. It is a targeting error. The traffic that buys does not land on your blog posts, it lands on your category and facet pages. This guide explains why, with figures and sources, and what it changes in your priorities.

Two kinds of traffic, two different jobs

Talking about "SEO traffic" as a single block is the first source of confusion. A visit is not worth the same depending on the question the user asked Google.

	Content traffic	Transactional traffic
Typical queries	"how to fold a fitted sheet", "what size duvet for a 160 bed"	"160x200 fitted sheet", "240x220 winter duvet"
Landing page	Blog post, guide, advice	Category page, facet page, product page
What the user wants	An answer	A product
What the visit delivers	Brand awareness, audience, links, AI citations	Orders
Exposure to AI Overview	High	Low

Both have their uses. But they are not steered with the same metrics, and they do not deserve the same budget. Content traffic is measured in audience and visibility. Transactional traffic is measured in revenue. Adding both together in a single "organic sessions" chart hides what matters.

The journey of a blog reader: three successive losses

For an informational search to turn into an order, the user has to clear three hurdles that a visitor to a category page does not face.

1. Earning the click from Google. On a question, Google increasingly answers for itself: featured snippets, related questions and, since summer 2026 in France, AI Overview. Some users leave without clicking.
2. Clicking again, from the blog post to a conversion page. A blog post is not a sales page. To buy, the reader has to follow a link to a category or a product. Most do not: they already got their answer.
3. Switching from an informational intent to a buying intent. Those who click arrive "cold". They were not looking to buy when they searched, and they convert less well than a visitor who landed directly on the category page.

These losses compound. The effective conversion rate of a visit to a blog post reads:

blog post conversion = internal click-through rate × category conversion × intent factor

The quantified model: 25 orders versus 2.5

The diagram above relies on deliberately conservative assumptions. Here they are, with what underpins them.

Assumption	Retained value	What it is based on
Conversion of an SEO visit to a category page	2.5%	Industry benchmarks put average e-commerce conversion around 2 to 3%.
Share of readers who click through to a category page	15%	There is no reliable public benchmark. 15% is an assumption favorable to blog posts: many e-commerce blogs do less.
Conversion of these "cold" visitors	1.7%	That is one third less than a visitor who landed directly on the category page. This is very conservative: in the case published by Grow & Convert, bottom-of-funnel content converts 25 times better than informational content (4.78% versus 0.19%).

Result: for 1,000 SEO visits, the category page produces 25 orders and the blog post 2.5. A factor of 10, before even counting the clicks lost inside Google.

Change the assumptions with your own figures, and the order of magnitude barely moves. For the blog post to catch up with the category page, all of its readers would have to click through to a product page and buy as often as a visitor who came to buy. That never happens.

Measure it on your own site

The calculation takes twenty minutes to check in Google Analytics 4:

1. Build an exploration on sessions from the Organic Search channel, over twelve months.
2. Classify landing pages by type using regular expressions on the URL: categories, facets, products, content.
3. Compare the session conversion rate and revenue per session for each group.
4. For blog posts, look at the share of sessions that then view a category or product page: that is your real internal click-through rate.

If your blog accounts for a large share of your organic sessions, your overall SEO conversion rate is mechanically dragged down. Segmenting it is often enough to change your trade-offs.

Since July 2026, AI Overview has widened the gap

Google rolled out AI Overview and AI Mode in France on July 22, 2026, two years after the United States. French sites are therefore now discovering a phenomenon that US studies have documented since 2024. Their conclusions converge.

AI Overview appears mostly on informational queries. Seer Interactive observes it on 36% of informational queries, 8% of commercial queries and only 5% of transactional queries.

When it appears, clicks collapse. Ahrefs measures a 58% drop in the click-through rate of the first position on the keywords concerned, across 300,000 keywords. Pew Research, which tracked the actual browsing of 900 US users, finds that they click on a classic result in 8% of cases when an AI summary is present, versus 15% without. And only 1% click on a link cited in the summary.

In other words, the first of the three losses, the one that plays out inside Google, is worsening precisely for informational content. Category-type queries, meanwhile, remain largely sheltered. We checked whether this holds in France, two months after launch: across 599 e-commerce queries, AI Overview appears on 70% of informational queries, 26% of commercial ones and 9% of transactional ones, and only 7% of long transactional queries, typical of facet pages. Our study details the results sector by sector.

An honest nuance is in order. Semrush observes that AI Overview is growing fast on commercial comparison-type queries (+71% between November 2025 and April 2026), such as best robot lawn mower. It is, however, slightly receding on transactional queries. Comparison buying guides, often presented as the miracle solution for e-commerce content, are therefore increasingly exposed as well.

Category pages concentrate the volume that sells

The usual argument in favor of blog posts is volume: "people ask questions, we need to answer them". On retail sites, it is often the opposite. Let us compare pairs of queries about the same product, in France.

Category-type query	Monthly volume	CPC	Related question	Monthly volume	CPC
160x200 fitted sheet	14,800	0.83	how to fold a fitted sheet	1,600	0.11
cordless stick vacuum	22,200	0.52	what power for a vacuum cleaner	1,000	0.31
compact stroller	8,100	0.35	how to choose a stroller	110	0.20
food for neutered cats	9,900	2.08	how much food per day for a cat	480	1.10
women's white sneakers	27,100	0.45	how to clean white sneakers	1,000	0.16
long dress	33,100	0.36	what shoes to wear with a long dress	390	0.20
bathroom paint	9,900	0.35	which paint for a bathroom	590	0.20
robot lawn mower	90,500	0.97	how does a robot lawn mower work	210	0.86

Average monthly search volumes and average Google Ads CPC in dollars, France, October 2026. Source: DataForSEO.

Two lessons. First, the category-type query almost always has more volume than the question. Second, its CPC is higher, often double to seven times as much. CPC is the price advertisers are willing to pay for a click: the advertising market, which measures its sales down to the cent, assigns more value to these visits.

This potential is spread across thousands of combinations: every size, color, material, brand or use case forms a precise query carried by a buying intent. That is exactly what your facets cover. A study by seoClarity across some thirty large US e-commerce sites already found in 2020 that category pages generated 413% more traffic than product pages.

Should you stop blogging? No, but change its role

An e-commerce blog is not useless. It does things your category pages do not:

• build the brand with people who are not buying yet;
• attract links from other sites, which internal linking then redistributes to category pages;
• feed remarketing audiences and the newsletter;
• get your site cited in AI answers, which sustains brand awareness even without a click.

What needs to change is the place it gets in your priorities, and the way it serves the business:

• every blog post links back to the relevant category pages through in-text links, not just a banner at the bottom of the page. HubSpot found that these contextual links generated between 47 and 93% of the leads from its articles;
• topics are chosen for their proximity to purchase, not just for their volume;
• its success is measured in audience and contribution, not direct revenue, so it is not artificially set against category pages.

Where to put SEO effort first

If your category and facet pages bring in the most, they are the ones that deserve the first euro of SEO budget. In order:

1. The tags on your existing category pages. Title tag, meta description and H1 decide your position and your click-through rate. On most catalogs, they are generated automatically by the CMS and identical from one page to the next. It is the fastest lever: see our method for category page title tags.
2. The copy on your category pages. Neither missing, nor an 800-word wall written for robots: short, useful copy specific to each page. See what really works for the SEO copy of a category page.
3. The facets that match real demand. Not every filter combination deserves to be indexed. Those that match searched-for queries do, with their own tags and their own copy.
4. Content, in support. Once the pages that sell are optimized, blog posts amplify their visibility.

The bottleneck is rarely strategy. It is volume: manually optimizing the tags and copy of 3,000 categories and facets represents months of work. That is precisely what we automated with unblind.

The test that settles it

For every page you are about to create or optimize, ask yourself one question:

Would I be willing to pay for advertising to get people to this page?

For a category page like "160x200 percale fitted sheet", the answer is yes, and advertisers confirm it by paying for that click. For a blog post on how to fold a fitted sheet, the answer is probably no. That is not a reason to never write it. It is a reason not to start with it.