Aleyda Solis’s AI Search Checklist: What Changes in May 2026 for Your E-Commerce
Summarize this article with AI
An e-commerce client calls me. €43,000 ad budget. Zero AI appearances.
A client calls me on a Thursday afternoon. €43,000 in annual ad budget. 12,000 SKUs. Polished product pages. But one number is spoiling his life: zero clicks from an AI search engine.
Not one citation. Not one recommendation. Nothing in ChatGPT Search, Perplexity, or the new Bing.
€43,000 spent, while less organized competitors take the organic visibility spot whenever a buyer asks a question in natural language.
This client is not a one-off. I see the same thing every week. Aleyda Solis’s checklist, updated on May 27, 2026, lands at exactly the right time. It gives a clear structure to move from « we should optimize for AI » to « here are the pages to rework, the signals to activate, and the third parties to get talking. »
Aleyda delivers a 12-step roadmap. I focus on the 5 points that, for an e-commerce site, change everything immediately.
Here is an overview of the steps Aleyda Solis recommends to gain visibility in AI search engines. Follow this path step by step to turn your e-commerce site into a prime target for generative AI.
The 6 key steps of the AI Search checklist for e-commerce
From measuring your presence to reporting with full honesty
Measure your AI presence before touching a single line of code
Aleyda’s first lesson is a reflex few e-commerce sites have: measure your current presence in AI answers. Not clicks. Not referral traffic. Raw presence.
Before talking about content, entities or citations, you need to know which prompts your brand shows up on. Aleyda recommends using Bing Webmaster Tools, AI visibility tools, or sample prompts pulled from your analytics. The idea: map your AI « journeys » — the questions that trigger an answer where your site could exist.
At the €43,000 client, I sampled 127 prompts tied to his flagship catégories. Result: 4 appearances, all on outdated Bing versions. None on ChatGPT. None on Perplexity. 127 questions where the client was invisible.
This measurement step is the foundation. Without it, you optimize blind. The checklist even includes a « What good looks like » tab: AI engines should be able to cite the brand, not just the generic domain.
I keep this number for my client: 3.1% presence on targeted prompts. The goal set: 40% in 6 months, using the checklist’s levers.
Want to know how many AI prompts you dominate today?
Make commercial pages extractable and machine-readable
Aleyda stresses a point e-commerce sites overlook: your product pages must be « retrievable » and « extractable ». Crawlable, sure, but above all structured so AI can extract price, availability, rating, and shipping without distorting the information.
The May 2026 version emphasizes machine-readable commercial structured data. Product, Offer, ShippingDetails, AggregateRating, MerchantReturnPolicy. More than just adding them: nest them, keep them current, and make them self-sufficient.
At my client, we audited 1,200 product pages. 68% had incomplete markup. The price was missing in 22% of Products. Real-time availability was never filled in. Result: AI engines picked the pages of a competitor who pushed a complete Merchant Center feed and markup matching the Google schema.
After fixing this, the correct extraction rate (verified through monthly test prompts) rose from 31% to 88%.
The gain goes beyond technical. Once real availability was surfaced, AI engines stopped recommending out-of-stock products. The bounce rate on pages coming from AI dropped by 19 points.
And that is a competitive edge advertising cannot buy.
Build decision-support content, not just a catalog
The checklist makes one point clear: AI engines favor pages that help you decide, over pages that merely inform. Aleyda names this step « Build decision-support and comparison content ».
For an e-commerce site, that means: comparison guides, criteria tables, structured questions and answers about use cases, constraints, and alternatives.
I applied this logic on an outdoor gear site. We created 14 comparison pages between products in the same range, listing 8 differentiating attributes. Each page answered queries like « X vs Y for a winter hike ».
In 90 days, these 14 pages captured 1,730 clicks from AI answers. 1,730 clicks that did not exist before.
Aleyda’s tip: these pages include decision validations (when to choose A over B), and an extractable format (lists, tables, side-by-side comparisons). AI has no time to waste on vague walls of text.
Do your category pages answer « Which one should I choose? » or only « What is it? »
Earn the citation and the click by playing as a team
Aleyda distinguishes two levels: being cited, then being clicked. Since May 2026, the « earn the citation and the click » lever rests on citable assets that compound.
For an e-commerce site, that means three actions.
1. Proprietary data AI engines want to cite (product tests, real return rates, customer surveys).
2. Pages that signal their citation potential (« according to study X by brand Y », « based on 4,200 verified reviews »).
3. A mesh of external sources pointing to these assets.
The client case speaks for itself. We published an internal report on the most-returned running accessories. 3,200 rows of data. We marked up the source, made the table public, and reached out to 7 specialized media outlets.
4 weeks later, Bing’s AI Search cited this table in 11 prompts. The snowball effect brought 2,100 organic sessions in 30 days.
Aleyda explains: citations are not decreed, they are built over time. But a data-driven asset, exposed correctly, speeds up the process.
This is the opposite of « let’s publish a blog post and wait ».
Localize by market, not just by language
The mistake I see often: translate the site and think the job is done. Since May 2026, the SEO checklist includes a point many forget: adapt signals by market, not just by language.
Aleyda says: make sure that hreflang tags, Google Business profiles, marketplace listings, local partners, local-currency pricing, and shipping options are aligned for each market.
An e-commerce site present in 4 European countries with the same platform asked me why it ranked well in France but not in Germany. The content was translated. But the prices stayed in euros without German VAT, the delivery times were not adapted, and the LocalBusiness markup pointed to a French address.
We fixed the signals by market. 6 weeks later, AI citations in German jumped by 580%.
My takeaway: AI Search reads more than the language. It reads the market context. Logistics proximity, currency, payment method, local return policy. So many signals that decide how relevant the answer is.
If you only translate the pages, you give the local market half the expected signals. The other half is provided by the local competitor.
Report with full honesty
The last point I keep for e-commerce sites is reporting. Aleyda devotes a full step to it: « Report without overclaiming« .
She warns against over-attributing AI conversions. A click arrives via ChatGPT Search; that does not mean the conversion funnel is attributable to AI optimization alone. Often, the same user combines AI, classic organic, and direct visits.
The checklist advises a 3-layer tracking: AI presence, readiness (pages and signals), and business impact. For impact, it suggests tracking assisted conversions, beyond last click alone.
At the outdoor client, we set up dedicated UTM tracking plus a data-driven attribution model (Data-Driven in GA4). Result: 34% of conversions from users who had an AI touchpoint did not have it as the direct conversion source. They were attributed elsewhere. Without this model, we would have understated the impact by a third.
The client report keeps it real. It shows three curves: AI visibility, traffic, and assisted incremental revenue. Clear, verifiable.
This is what I call reporting that fits in 4 slides, instead of 15 pages of jargon.
Have you already compared your current attribution to what Aleyda flags as overclaiming bias?
Your AI Search diagnostic in 45 minutes
I show you which AI prompts overlook your catalog, which pages are extractable, and which competitor captures your customers while skipping the ads. A live audit, not a 30-page PDF.
Book a diagnostic call — 45 minFrequently asked questions
What changes in Aleyda Solis’s checklist in May 2026?
This update shows how to localize by market, integrate machine-readable commercial structured data, create decision-support content, and report while keeping AI impact grounded in reality.
Why do my e-commerce pages stay absent from AI answers?
I see two reasons. Either the product data is missing from the structured tags (price, stock, shipping absent). Or the site lacks comparison and decision-support content. AI engines love that kind of content; when they find none, they move on.
How do I measure my visibility in AI search engines?
Aleyda advises: sample business prompts, use Bing Webmaster Tools and AI visibility tools, then cross-reference with your AI referral traffic data.
Should I adapt local SEO for AI Search?
Yes, and beyond the language level. This checklist focuses above all on market signals: currency, local delivery times, consistent LocalBusiness markup, and return policies by country.
Does going machine-readable require advanced technical skills?
It mainly takes completing your Product, Offer, AggregateRating and MerchantReturnPolicy schemas. Most e-commerce CMS platforms can do it without a developer, though an initial audit is needed.
How long until the first AI Search results show up?
I observe first shifts in citations within 4 to 6 weeks, and measurable traffic impact within 3 months, after fixing the checklist’s key signals.
