The state of organic search for retail commerce, as of 2026, is considerably less hospitable than the state that produced the dominant SEO playbooks of the previous decade. The Google search results page now devotes the great majority of its visible real estate, on the great majority of queries, to surfaces that do not produce clicks on any organic result: the AI-generated answer at the top of the page, the shopping carousel, the local pack, the featured snippet, the People Also Ask block, the video pane, and the assorted vertical-specific surfaces that the search engine continues to add. The proportion of search queries that produce no click on any organic result has, on the data I have available from the retail clients I work with, exceeded sixty per cent and continues to rise.
Easy SEO is the technical SEO toolkit I built to address what remains of the organic-search opportunity in this environment. The toolkit is, considered as a piece of software, less ambitious than the SEO platforms that dominate the procurement conversations; it is also, considered as a guide to the work that actually moves organic traffic in 2026, considerably more focused on the small set of interventions that produce measurable results and considerably less concerned with the larger set of interventions that the established platforms continue to charge for despite their having been overtaken by the changes in the search environment. This post is an account of what the toolkit does, what it deliberately does not do, and how the work of organic SEO has changed in the post-zero-click era.
What still works, in 2026
The interventions that continue to produce measurable organic-traffic improvements, on the data I have collected across approximately twenty client deployments, fall into a small number of categories.
The first is structured data — the schema.org markup that signals to the search engine the type and content of the page. The category has expanded considerably in importance over the previous three years, principally because the AI-generated answer at the top of the search results page draws disproportionately on structured-data sources when constructing its responses. A retailer whose product pages emit complete, accurate Product schema is considerably more likely to be cited in the AI answer than a retailer whose pages emit only the basic markup; the citation, in turn, is one of the few mechanisms by which an AI-answer query produces a click-through to the source. Easy SEO’s structured-data validator audits each page on the site against the schema types relevant to its content and reports, in a small dashboard, the gaps that should be filled.
The second is technical performance — the Core Web Vitals scores and the broader set of technical signals that the search engine uses as ranking inputs. The category remains as important as it has been for some years; the difference is that the threshold above which the technical signals contribute meaningfully to ranking has, in my observation, risen. A site that was technically competent in 2022 is, on the same metrics in 2026, often technically below the threshold of contribution; the search engine’s standards have moved upward in step with the broader improvement in web performance, and sites that have not moved with them no longer accrue the ranking benefit they previously did. (For a more detailed account of the technical performance work that produces the improvements, see A Core Web Vitals diary.)
The third is content depth — the substantive, expert-authored content that the search engine increasingly favours over the thinner content that dominated the previous decade’s SEO output. The change is connected to the search engine’s introduction of the so-called helpful-content system, which evaluates content on a set of qualitative dimensions that proxy for the content’s actual usefulness to the searcher. The system is imperfect, and its outputs vary in ways that are sometimes difficult to interpret; on aggregate, however, the direction of travel is clear, and content that would have ranked competently in 2020 frequently does not rank at all in 2026 unless it has been substantially deepened.
What no longer works
The interventions that have, in my observation, ceased to produce measurable organic-traffic improvements are worth listing explicitly because they continue to be sold by the larger SEO platforms as if they remained productive.
Keyword density manipulation, in any of its forms, no longer produces measurable improvement and frequently produces an active regression as the search engine’s content-quality systems penalise the resulting prose for its unnatural composition. The tools that continue to recommend specific keyword densities for specific page types are, in 2026, recommending an SEO practice that has been obsolete for some years.
Link-building, considered as a category of activity, has not ceased to be a ranking factor; the specific tactics that were sold as link-building under the previous era’s playbooks have, however, been substantially neutralised by the search engine’s efforts to identify and devalue manufactured link profiles. The link-building services that remain effective are those that produce links from sources the search engine treats as authoritative independent of any link-acquisition campaign; the services that produce links from networks of sites whose authority is itself manufactured produce no measurable improvement and occasionally produce a measurable penalty.
The dense thin-content production that characterised the previous era — the long-tail content farms producing one short article per long-tail keyword — has been, in operational terms, deprecated by the helpful-content system. The output of these farms is no longer rewarded with organic traffic, and the operational cost of producing the content is no longer recoverable. The retailers who have been advised by their SEO suppliers to maintain thin-content production through 2024 and 2025 are, in 2026, almost without exception in the position of having spent considerable resources on content that is producing no organic value.
The toolkit, in operational terms
Easy SEO assembles the interventions that continue to produce measurable improvement into a small set of audits and ongoing checks. The structured-data audit examines every product, category and editorial page on the site against the appropriate schema types and reports the gaps. The technical-performance audit examines the Core Web Vitals scores against the current thresholds and identifies the specific files, scripts and styles producing the largest contributions to the regression. The content-depth audit examines the editorial pages against a small set of qualitative criteria — the presence of expert authorship indicators, the depth of treatment relative to the query the page targets, the freshness of the content relative to the volatility of the topic — and identifies the pages most likely to benefit from deepening.
The audits are, individually, modest in their ambition; the combined effect of acting on the recommendations they produce has, on the deployments I have measured, consistently produced organic-traffic improvements in the range of fifteen to forty per cent against the baseline. The principal predictor of where in the range a deployment lands is the starting point: sites that were broadly competent on the previous era’s metrics typically achieve the lower end of the range, while sites that had neglected the technical and structured-data work for some years typically achieve the higher end.
What the toolkit does not address
The toolkit does not, in its current form, address the question of how to operate within the search-engine environment in which the AI answer increasingly substitutes for the click-through. The question is, in some sense, the question of the era; the answer that is emerging from the practitioner literature is approximately that retailers should optimise for citation in the AI answer rather than for clicks on the organic result, and that the optimisation requires the structured data and content-depth work that the toolkit already produces. The optimisation does not, however, recover the click-through that the AI answer is consuming; the retailer who is cited in the AI answer is in a better position than one who is not, but is in a worse position than the same retailer would have been in the pre-AI search environment.
The implication, which the toolkit does not directly address, is that organic search is, in 2026, a smaller share of total retail traffic acquisition than it was in 2020, and that the retailer’s traffic-acquisition strategy should be diversified accordingly. The diversification is a strategic question rather than a technical one, and is therefore beyond the scope of the toolkit; the data the toolkit produces is, however, useful in informing the strategic decision.
(For a related view on the broader procurement narrative around AI integration in retail commerce, see AI search is overhyped: auto-complete is underbuilt.)
An advisory close
The technical SEO work that produces measurable improvement in 2026 is a smaller and more focused set of interventions than the work that was sold under the SEO label in the previous decade. The retailer’s first task is to distinguish the work that continues to be productive from the work that has been deprecated by the changes in the search environment; the distinction is, in the procurement conversations I have, frequently obscured by suppliers whose business models depend on continuing to sell the deprecated work as if it remained productive.
It is recommended that any retailer commissioning SEO work request, from any supplier under consideration, an explicit account of which interventions the supplier expects to produce measurable improvement and which are included for completeness. The explicit account is, in my experience, decisive for distinguishing the suppliers whose work will recover its cost from those whose work will not.