"SEO in an AI-First Search Landscape: What eCommerce Brands Need to Reassess" - A Pulse Commentary

In the run up to the Pulse eCommerce Summit on the 13th - 14th May 2026, we are launching a series of commentary pieces on topics that will be a focus point at the conference, led by members of the Vervaunt team. Here is our commentary piece on SEO in an AI-first search landscape, with input from Barrett Ahern.

Over the past 18 months, the platforms where product discovery happens have expanded significantly. Google remains central, but ChatGPT, Perplexity, Gemini, and other AI-powered interfaces now handle a meaningful share of product research and purchase consideration.

The term "GEO" (generative engine optimisation), has emerged to describe optimising for these new surfaces, though the distinction from traditional SEO is arguably more marketing than substance. The fundamental discipline remains the same: making your brand and products discoverable, credible, and chosen. What has changed is where that discovery happens, how the algorithms work, and what outcomes you should be optimising for.

The shift matters for ecommerce brands because the mechanics of discovery have changed in ways that affect strategy. Users increasingly receive answers without clicking through to any website. AI platforms synthesise information from sources brands don't control, sometimes inaccurately, and users tend to accept those answers as authoritative.

Product data is being compared against competitors inside platforms where brands have no storefront presence. The traffic metrics that historically indicated SEO success are becoming less reliable as measures of actual business impact.

This article covers what has changed in practical terms, what we are seeing work for brands navigating these shifts, and what warrants reassessment if search, in its evolving forms, is part of how your customers find you.

From traffic acquisition to narrative control

Traditional SEO focused on ranking for relevant keywords and acquiring traffic. Success was measured in clicks, sessions, and conversion rates. Brands controlled the experience from the moment a visitor arrived on site.

That dynamic is shifting. When someone searches in ChatGPT or Perplexity for a product recommendation, they often receive an answer without visiting any brand's website. The platform synthesises information from across the web, presents a response, and the user either acts on it or clicks through to whichever source was cited most prominently. Visibility alone is no longer sufficient. Brands now need to ensure they are being accurately represented in the synthesis, that their products are being included in relevant comparisons, and that the information AI platforms present about them is correct.

The case of Tailwind CSS illustrates what this shift can mean in practice. Tailwind is a CSS framework widely used by developers, and its popularity is currently at an all-time high; it's the default styling system for tools like Lovable, and developer mindshare has never been stronger. Despite this, the company recently laid off 75% of its workforce, with revenue down 80% and traffic to documentation pages down 60%.

The explanation is instructive. Developers previously encountered a problem, searched Google for a Tailwind solution, browsed the documentation, and in the process discovered premium features worth paying for. Now they ask ChatGPT or Gemini, receive the code snippet they need, and never visit the Tailwind site. The brand is more visible than ever, constantly recommended, but that visibility is not converting to traffic, engagement, or revenue. Consumption happens inside the AI platform rather than on the brand's owned properties.

For ecommerce brands, the implications are similar. If an AI platform can answer "what's the best waterproof jacket under £200" by synthesising reviews, specifications, and comparisons, the brands cited in that answer gain consideration. Those not cited may never enter awareness. And if the information presented is incomplete or inaccurate, a brand loses to a version of itself that it did not create and cannot directly correct.

The strategic response is to treat AI platforms as surfaces where your brand narrative needs to be actively managed, not just indexed. This requires ensuring product data is structured for the comparisons users are asking for, monitoring what AI platforms say about your brand, and accepting that some portion of your search presence may not result in direct site traffic.

Zero-click behaviour and the measurement gap

Zero-click search behaviour predates AI. Google's featured snippets and knowledge panels started this pattern years ago, but AI overviews have accelerated it. Users are increasingly served information directly in the search interface rather than presented with links to explore.

The impact is visible in analytics. Brands report increased impressions in Google Search Console, indicating they appear in results, alongside reduced traffic to homepages and category pages. Users see the brand mentioned, may note it for future reference, but do not click through because they received what they needed from the results page itself.

This creates a measurement challenge. Traditional SEO success was measured by organic traffic volume and the revenue it generated. If impressions increase while traffic decreases, evaluating performance becomes less straightforward. If a brand is mentioned in AI overviews but users do not click through, the value of that exposure is real but difficult to quantify.

The tools available for measuring visibility in AI-generated responses are still developing. Ahrefs and SER Interactive have built monitoring capabilities, but the data is incomplete and methodologies are experimental. The platforms themselves do not provide the detailed analytics that Google Search Console offers for traditional search. Part of this is technical, AI platforms often do not expose referral data in ways that enable attribution. Part of it may be strategic, detailed user behaviour data may not align with platform interests.

For now, brands should track branded search volume and impressions as leading indicators of awareness, even if direct traffic is flat or declining. Establishing a baseline for visibility in AI-generated responses, which queries mention your brand, which mention competitors, which sources are cited, provides a foundation for understanding how presence is evolving, even without precise attribution.

Product data as the foundation

Product data is the clearest priority for brands preparing for AI-driven discovery. Clean, structured, retrievable product data determines whether your products surface in AI recommendations and whether they are represented accurately when they do.

The integration between Shopify and ChatGPT demonstrates why this matters. When users ask ChatGPT for product recommendations, the system can pull directly from Shopify product catalogues. Brands with well-structured data, clear attributes, accurate pricing, complete specifications, are more likely to appear in responses. Brands with incomplete or inconsistent data risk being excluded or misrepresented.

The baseline requirement is having product attributes properly structured in catalogue feeds: colour, size, material, price. The additional requirement, which fewer brands have addressed, is structuring data for comparison intent. When someone asks "what's the best running shoe for flat feet under £150," the AI needs to compare products across multiple dimensions. If product descriptions focus on brand storytelling but lack the functional attributes that enable comparison, arch support level, cushioning type, recommended use case, the product is disadvantaged against competitors whose data supports direct evaluation.

There is also a technical dimension. AI platforms are less sophisticated than Google at parsing JavaScript, accordions, tabs, and interactive elements. If a size guide is hidden behind a JavaScript toggle or embedded in a PDF, the AI may not access it. When this happens, a retailer or comparison site that scraped the data and presented it in clean HTML can become the authoritative source. The original brand created the information but lost control of it because the format was not accessible to the AI.

We have observed this with size charts specifically. A brand has detailed sizing information locked in a third-party fit tool that renders via JavaScript. A comparison site copies that data into a simple HTML table. When users ask about sizing, the comparison site is cited as the source rather than the brand itself.

The practical response is to present important product information in clean, crawlable HTML, even if a more sophisticated interactive version also exists. Product descriptions should be written as answers to comparison questions, not only as brand copy. Functional attributes that enable side-by-side evaluation should be explicit. Testing what major AI platforms actually say about your products, and identifying the gaps between intended communication and actual synthesis, should be part of ongoing governance.

"We’re seeing structured data and offsite signals play a much bigger role in AI led search. Marking up your content using structured data can greatly aid the understanding of AI engines, whilst their increasing reliance on external sources to generate comparisons and recommendations has also made growing a brand's off site citations critical." - Chris Shelbourn, Technical SEO Director, IDHL

Managing brand reputation on surfaces you do not control

Brand reputation now requires management on surfaces that brands do not own and cannot directly edit. Research indicates that AI platforms are incorrect or hallucinating approximately 40% of the time when queried about basic brand information; opening hours, contact details, phone numbers. If an AI platform tells a user your store closes at 6pm when it closes at 9pm, you have potentially lost a visit, and there is no mechanism to submit a correction.

The sources AI platforms draw from are weighted toward specific sites. Wikipedia features prominently. Reddit threads carry significant authority. Review sites, comparison articles, and forum discussions all contribute to the synthesis. A brand's own website is one input among many, and not always the primary one.

This means brand reputation management now extends to monitoring and, where possible, influencing external sources. Wikipedia accuracy, Reddit sentiment, Google Business Profile completeness, these were always relevant for SEO, but they carry more weight now because AI platforms often treat them as more authoritative than brand-owned content.

There is no equivalent to Google's URL deprecation request tool for AI platforms. If ChatGPT states something incorrect about your brand, the only recourse is to reinforce correct information elsewhere; update Wikipedia, ensure accurate coverage in indexed sources, maintain consistency across every touchpoint you control, and wait for training updates to incorporate the corrections.

This is not a problem with a clean solution, it requires ongoing governance and monitoring. Brands that treat their external presence as something to manage proactively will be better positioned as AI platforms become more central to how customers discover and evaluate options.

Authority signals and algorithmic maturity

AI platform algorithms are sophisticated enough to be useful but have not developed the spam and quality protections that Google built over two decades. This creates both opportunity and risk.

One study demonstrated the current state clearly. Researchers identified articles referenced in AI responses that ranked around position 60 in citation frequency. They changed nothing except the publication date, updating it from 2020 to 2025. On average, those articles jumped to position 5 in citation frequency. The platforms weight freshness heavily but lack robust mechanisms to verify whether a "fresh" article has been substantively updated or merely re-dated.

Author profiles, publication dates, and other credibility signals that Google evaluates with nuance are treated more simply by AI platforms. Relatively straightforward optimisations, ensuring content appears current, establishing clear authorship, demonstrating expertise, can have disproportionate impact. The corresponding risk is that these signals can be gamed, and users are unlikely to verify what they are told.

The combination of user trust and algorithmic immaturity creates reputational exposure. Users assume AI outputs are accurate and do not click through to verify. They accept recommendations and act on them. A brand can be misrepresented, and the user will not discover the error because they will not visit the source to check.

This argues for regular monitoring of what AI platforms say about your brand and products, tracking which sources are cited when your brand is mentioned, and treating this as an ongoing responsibility rather than a one-time optimisation.

The connection between discovery, loyalty, and transaction costs

If discoverability becomes more standardised, if brands can appear in AI recommendations provided their data is structured correctly, differentiation shifts to other factors. Service quality, customer experience, and loyalty programme strength become more important when initial discovery is less of a competitive advantage.

There is also a direct financial consideration. The ChatGPT commerce integration carries a 4% fee to OpenAI and 2% to Shopify on transactions processed through the platform, in addition to existing payment processing costs. For brands operating on typical ecommerce margins, this represents a meaningful difference between AI-originated transactions and direct site visits.

The strategic response is to treat AI-driven discovery as an entry point while building systems that bring customers back to owned channels for repeat purchases. Loyalty programmes, differentiated service, post-purchase engagement that builds relationships rather than simply pushing the next transaction, these become more valuable when the first touchpoint occurs on a platform the brand does not control.

There is a reinforcing dynamic here. Customers with positive experiences are more likely to write reviews, post on Reddit, and recommend the brand in forums, sources that feed into AI training data. Brands that deliver strong service create positive signals that shape how AI platforms describe them. Brands that disappoint create the opposite effect.

The intersection of paid search and organic visibility

The relationship between paid and organic search is evolving as AI overviews and platform-native commerce become more prominent.

Branded impressions are likely to increase as AI platforms mention more brands in responses. Competition for branded terms, in traditional paid search and emerging AI advertising formats, may intensify as competitors seek to intercept the awareness that AI recommendations create.

There are also emerging opportunities in unexpected keyword categories. Targeting highly specific identifiers; product SKUs, UPC codes, model numbers, has shown early results. These terms have minimal competition in traditional paid search but are frequently referenced in AI-generated recommendations. Users who see a specific product recommended often search for that exact identifier; being the top paid result for that search captures high-intent traffic efficiently.

More broadly, the boundary between organic and paid visibility is less distinct than it was. AI overviews in Google blend organic information with shopping ads. ChatGPT is integrating sponsored content. Brands will likely need to coordinate organic and paid strategies more closely, treating AI-driven discovery as a channel requiring both content investment and media spend.

What to reassess before investing further

If search visibility across traditional and AI-driven platforms is part of how customers find you, these areas warrant evaluation.

On product data: 

• Audit whether your catalogue is structured for comparison queries, not only attribute completeness. The question is whether the attributes that matter for purchase decisions are explicit and comparable across your range.
• Review how important product information is rendered technically. Size guides, specifications, FAQ content, is it in clean HTML that any system can parse, or is it dependent on JavaScript, PDFs, or third-party widgets that may not be accessible to AI platforms?
• Test what ChatGPT, Perplexity, and Google's AI overviews say about your products. Search for your brand, product categories, and comparison queries you would expect to appear in. Identify gaps between what you intend to communicate and what these systems actually synthesise.

On brand reputation: 

• Review external sources likely to influence how AI platforms describe your brand. Wikipedia, Reddit, Google Business Profile, major review sites, assess whether they are accurate, current, and consistent with what you want potential customers to hear.
• Establish a monitoring process for what AI platforms say about your brand. Regular manual checks can surface problems. The process needs to be ongoing rather than one-time.
• Accept that direct correction of AI misinformation is not currently possible. The approach is to reinforce correct information across sources that AI platforms trust.

• Track branded impressions and search volume as leading indicators, even if traffic metrics are flat or declining.
• Establish a baseline for visibility in AI-generated responses. Note which queries mention your brand, which mention competitors, and which sources are cited.
• Accept that precise attribution is not currently possible. Focus on directional indicators and qualitative assessment while measurement tools mature.

• Treat AI-driven search as an evolution of SEO rather than a separate discipline. The fundamentals of clean data, credible content, and consistent external presence apply across platforms.
• Recognise the connection between post-purchase experience and future discoverability. Positive customer experiences generate the reviews, Reddit posts, and recommendations that shape AI training data. Investment in loyalty and service is also investment in how AI platforms will describe your brand.
• Model the economics of transaction location. Discovery through AI platforms may be unavoidable, but there is a cost difference between customers who transact through those platforms and those who return to your site directly. Build systems that convert AI-driven discovery into owned-channel relationships.

Many of the themes explored here - from managing international risk and rethinking market entry economics, to building operational resilience and local relevance - will be unpacked in far more depth at the Pulse eCommerce Summit on the 13th and 14th May 2026. Across two days, we’ll bring together senior eCommerce leaders, operators and specialists to share real-world experiences, practical frameworks and honest lessons from scaling brands internationally in a far more complex global landscape. If international growth is on your roadmap for 2026 and beyond, register now to secure your place.