Table of content
- Getting found now means two different search engines, not one
- Technical SEO basics, in brief
- Product and application pages: the pages that actually sell
- Structured data: what still earns a rich result, and what doesn’t
- AI answer engines: being cited, not just ranked
- Measuring visibility: mature for Google, still immature for AI
- Traditional search vs. AI search, side by side
- Putting this into practice
- Common mistakes
- Key takeaways
- FAQs
- References
- Related reading on OneStopNDT
Getting found now means two different search engines, not one
An NDT company used to have one job: rank in Google for the inspection method or service a buyer typed in. That job hasn’t gone away — traditional organic search still drives most B2B research — but a second channel has opened alongside it. Buyers now also ask ChatGPT, Google’s AI Overviews, or Perplexity to summarize, compare, and shortlist vendors before they ever open a search results page. A procurement engineer scoping a corrosion-under-insulation inspection contract might ask an AI tool to list qualified vendors and summarize their capabilities — and if a company’s site doesn’t give that tool clear, structured, well-sourced information to work from, it simply won’t be mentioned.
This guide covers both channels for NDT marketers and owners: the technical SEO fundamentals that still decide whether Google can find and rank a site, the structured data that helps both Google and AI tools understand product and application pages, and the newer practice of optimizing for AI answer engines — variously called GEO, AEO, or AI search optimization. It is deliberately honest about which of these is mature and measurable (traditional SEO) and which is still new and loosely instrumented (AI search) — and it links out to OSNDT’s existing, more detailed articles on general SEO strategy, local SEO, website speed, responsive design, and site features, which this piece builds on rather than repeats.
Technical SEO basics, in brief
The fundamentals that decide whether Google can crawl, index, and rank an NDT company’s site haven’t changed, and OSNDT already covers each in depth — this section is a short orientation, with the fuller treatment linked below.
- Crawlability and indexing. A clean sitemap, a logical URL structure (service pages, application pages, resource pages kept in distinct, consistent paths), and no orphaned pages let Google find and index everything worth ranking.
- Mobile-friendliness and page speed. Google indexes and ranks primarily on the mobile version of a site, and slow-loading pages lose both rankings and visitors before they finish loading. OSNDT’s own importance-of-website-speed article covers the tools (PageSpeed Insights, GTmetrix, Pingdom) and tactics (image compression, caching, minified CSS/JS) in detail.
- Responsive design. A site that reflows correctly across phone, tablet, and desktop is now a baseline ranking and usability factor, not a nice-to-have. See OSNDT’s importance-of-responsive-design-for-ndt-websites for the technical approach (fluid grids, flexible images, media queries).
- Internal linking. Linking service pages to the application pages that use them, and both to supporting resource content (FAQs, glossary, guides), spreads authority across the site and helps both Google and AI tools understand how a company’s offerings relate to each other.
- Local SEO. For NDT companies that serve a geographic territory — field inspection, calibration, on-site services — a complete Google Business Profile, consistent name/address/phone (NAP) data, and location-specific pages matter as much as on-page content. See OSNDT’s importance-of-local-seo-for-ndt-companies.
- General on-page and off-page strategy. Keyword research, content depth, backlinks, and site UX are covered fully in OSNDT’s top-seo-strategies-for-ndt-websites and top-website-features-every-ndt-website-needs.
Get these right first. Structured data and AI-search optimization, covered next, build on top of a technically sound, fast, well-linked site — they don’t substitute for one.
Product and application pages: the pages that actually sell
Most NDT company websites under-invest in two page types that do the real work of getting found and chosen: product pages (an eddy-current instrument, a phased-array probe, a calibration block) and application pages (corrosion-under-insulation inspection, weld inspection for a specific code, tank-floor scanning). These are where a buyer’s specific, high-intent search lands — and where structured data has the most direct payoff.
A well-built product or application page gives both Google and an AI answer engine the same things: a clear, specific title and heading (the method and the application, not just a brand name), a description that states what the service or product does, for what materials or defects, to what standard, a list of specifications or deliverables, and a clear path to contact or quote. Thin pages that just restate a brand name and a stock photo give a search engine nothing to rank and an AI tool nothing to cite.
Structured data — covered next — is the machine-readable layer on top of that content: it tells Google explicitly “this is a Product” or “this is a Service,” with price, availability, or service-area fields filled in, rather than leaving the engine to infer it from prose.
Structured data: what still earns a rich result, and what doesn’t
Structured data (schema.org markup, typically added as JSON-LD) is a direct way to tell Google and AI crawlers what a page is about. But the rules on what structured data actually earns a visible reward in Google changed materially in 2023, in a way most NDT company sites don’t yet reflect.
- Product and Service schema remain fully active. Google’s own documentation confirms Product/Offer markup still drives rich results (price, availability, review snippets) in search [Google: Merchant listing structured data]. This is the highest-value schema for an NDT company’s equipment or service pages.
- Google has now fully retired FAQ and HowTo rich results — for every site. HowTo rich results were removed entirely from September 2023. FAQ rich results were first restricted (August 2023) to one Q&A per page, for authoritative government/health sites only, then phased out for those remaining sites too: full removal across all sites took effect 7 May 2026, with Search Console’s FAQ report and the Rich Results Test dropping FAQ support that June and API support ending August 2026 [Google: FAQPage structured data]. In practice, no NDT company site — regardless of how correctly it’s marked up — can earn a visible FAQ or HowTo rich snippet in Google search results any more, for any site.
- FAQ/HowTo schema still has value — just not the one most marketers expect. It no longer buys a Google rich snippet, but clean, explicit question-and-answer markup is exactly the structure an AI answer engine parses most easily when assembling a response. Keep FAQ markup for that reason, not for a search-results snippet that Google has stopped awarding to sites like this one.
- Organization schema is worth adding sitewide. It doesn’t drive a rich result by itself, but it gives both Google’s Knowledge Graph and AI tools an unambiguous entity record — legal name, logo, services, locations, social profiles — which matters for the entity-clarity signals AI answer engines rely on (see below).
Practical priority for an NDT company site: Product/Service schema on every equipment and service page first; Organization schema sitewide second; FAQ/HowTo schema on key pages third, framed internally as an AI-parsing aid rather than a rich-snippet tactic.
AI answer engines: being cited, not just ranked
Traditional SEO optimizes for a ranking position a person clicks. AI answer engines — Google’s AI Overviews and AI Mode, ChatGPT, Perplexity — synthesize an answer from multiple sources and present it directly, often without the user clicking through to any of them. This practice of optimizing to be one of the sources an AI tool cites or draws on goes by several overlapping, not-yet-standardized names: GEO (generative engine optimization), AEO (answer engine optimization), AIO, LLMO. None has settled as the industry standard term, and the underlying tactics overlap heavily with good SEO — clear structure, authoritative sourcing, entity clarity — rather than replacing it [eMarketer: FAQ on GEO and AEO, 2026].
A few findings worth an NDT marketer knowing, stated honestly rather than oversold:
- AI tools don’t cite the sources you’d expect. As of an October 2025 dataset, Reddit, LinkedIn, and YouTube were among the domains most frequently cited by major LLMs in response to queries — not traditional trade media or industry publications. Being published and discussed in those channels, not just on a company’s own site, is part of being found by AI tools.
- Citations are unstable. 40–60% of the sources an AI tool cites for the same query change from one month to the next across Google AI Mode and ChatGPT. No tactic “locks in” AI visibility the way a stable Google ranking can.
- AI referral traffic is small in volume but converts disproportionately well. Some major publishers — Reuters and The Guardian, per Similarweb’s 2026 GenAI Brand Visibility Index — report well under 1% of their referral traffic coming from AI platforms despite frequent citation, yet The Washington Post has reported subscription-conversion rates 4–5x higher from AI-referred visitors than from ordinary search visitors. Low volume, high intent.
- This connects to how vendors actually get shortlisted today. Buyers are increasingly using generative AI tools to help build vendor shortlists before they ever visit a company’s site directly — a shift the OSNDT marketing-leader community has discussed in its own channels.
The practical takeaway: entity clarity (Organization schema, a consistent, well-documented identity across the site and in third-party-citable channels), clearly structured application and FAQ content, and a presence in the industry discussion that happens off-site — not just on-site copy — are what AI answer engines currently have to work with.
Measuring visibility: mature for Google, still immature for AI
Traditional search is well-instrumented. Google Search Console remains the first-party source of truth — impressions, clicks, average position, and query data by page. Third-party rank trackers (Ahrefs, SEMrush, Moz, already referenced in OSNDT’s own SEO strategy article) add competitor comparison and historical tracking. This toolkit is mature and standardized; an NDT company with no other analytics in place should start here.
AI-search measurement is genuinely new and not yet standardized. The most concrete development: Google Search Console began rolling out a dedicated “Search Generative AI performance” report starting 3 June 2026, separating out impressions for AI Overviews, AI Mode, and generative Discover features, with breakdowns by page, country, and date, and device breakdown for Search (not Discover) — though it does not yet include a clicks metric and is rolling out to a subset of sites first [Google Search Central Blog, June 2026]. Beyond that first-party report, a handful of vendor-published “AI visibility” benchmarking tools have appeared (for example, a January 2026 ChatGPT-vs-Perplexity-vs-Google-AI-Mode B2B citation benchmark from averi.ai, itself an aggregation of several third-party studies), but these should be treated as emerging and vendor-interested, not as settled methodology.
What’s realistically measurable today, and what isn’t. Citation frequency (how often a brand is mentioned when manually prompting AI tools with relevant queries), AI referral traffic as a segment in standard web analytics, and rough sentiment of how a brand is described are workable proxies. Prompt volume and the exact weighting an AI model gives to any one source are not reliably measurable by anyone outside the model providers themselves. An honest answer for an NDT marketer asking “how do I know if this is working”: track the proxies above, expect AI-search measurement to lag traditional search measurement by a few years, and don’t let a vendor’s dashboard imply more precision than the underlying data supports.
Traditional search vs. AI search, side by side
Traditional (Google organic) | AI answer engines (GEO/AEO) | |
|---|---|---|
What you’re optimizing for | A ranking position a user clicks | Being one of the sources an AI tool cites or synthesizes |
Core levers | Technical SEO, content depth, backlinks, local SEO | Entity clarity, structured FAQ/application content, off-site discussion (Reddit, LinkedIn, YouTube) |
Structured data payoff | Product/Service schema remains eligible for rich results (not guaranteed); FAQ/HowTo rich snippets fully retired industry-wide (2026) | FAQ/Organization schema aids AI parsing even without a Google rich-result reward |
Measurement maturity | Mature — Search Console, established rank trackers | Immature — GSC’s generative-AI report began rolling out June 2026, no clicks data yet |
Stability | A good ranking can hold for months | 40–60% of cited sources change month to month |
Traffic vs. intent | Higher volume | Lower volume, reportedly higher conversion intent |
Putting this into practice
The companies that benefit most from this guide’s priorities are NDT equipment distributors and service providers with genuine product and application depth to show — companies like SPA InnoVision, an OSNDT customer in this space.
In general, the NDT companies most likely to show up well in both Google and AI answer engines share the same traits: specific, well-structured product and application pages rather than generic brand copy; Product and Organization schema correctly implemented; and a visible presence in the industry conversation — webinars, panel discussions, trade-press commentary — that happens outside their own website, which is increasingly where AI tools are drawing the material they cite.
Common mistakes
- Expecting a Google rich snippet from FAQ/HowTo markup. Google fully retired FAQ rich results industry-wide as of 7 May 2026 (HowTo rich results were gone entirely from September 2023). Keep the markup for AI-parsing value, but don’t budget for a search-results snippet Google no longer awards to any site.
- Treating AI-search visibility as measurable with the same precision as Google rankings. The tooling isn’t there yet. A vendor dashboard that implies exact, stable AI-citation tracking is overselling an immature category.
- Writing thin product pages that restate a brand name instead of the application, material, defect type, and standard. Neither Google nor an AI tool has anything specific to rank or cite from a page like that.
- Skipping Organization schema because it doesn’t produce a visible rich result. It’s one of the clearest entity-clarity signals available, and entity clarity matters more, not less, for AI answer engines.
- Assuming AI-search tactics replace SEO fundamentals. They layer on top of a technically sound, fast, mobile-friendly, well-linked site — they don’t substitute for one.
- Attributing a specific customer or case-study claim without a citable source. Any customer example used in marketing content should be backed by material the customer has actually published or confirmed, not inferred.
Key takeaways
- Technical SEO fundamentals — crawlability, speed, mobile, internal linking, local SEO — are unchanged and still the foundation; see OSNDT’s existing articles for the detail.
- Product and Service schema remain eligible for Google rich results (not guaranteed); FAQ and HowTo rich results have been fully retired for every site — HowTo since September 2023, FAQ industry-wide as of 7 May 2026.
- FAQ/HowTo and Organization schema remain worth implementing — not for a Google snippet, but because they’re exactly the structure AI answer engines parse most easily.
- AI answer engines synthesize and cite rather than rank and click; being cited depends on entity clarity and off-site presence (industry discussion, trade press, forums) as much as on-site content.
- AI citations are unstable (40–60% turnover month to month) and AI referral traffic is currently small in volume, though it reportedly converts at a higher rate than ordinary search traffic.
- AI-search measurement is genuinely immature; Google Search Console’s generative-AI performance report (rolling out from June 2026) is the most concrete first-party tool available so far, and it still has no clicks data.
FAQs
Does FAQ schema still help SEO for an NDT company site? It no longer earns a Google rich snippet at all — Google fully retired FAQ rich results for every site as of 7 May 2026 (after first restricting them to government/health sites in August 2023). It’s still worth adding because AI answer engines parse clearly structured Q&A content easily when assembling a response.
What’s the single highest-value structured-data type to add first? Product or Service schema on equipment and service pages, since it’s the one type confirmed to still drive visible rich results (price, availability, reviews) in Google search.
Can I track how often my company is cited by ChatGPT or Google AI Overviews? Partially. Google Search Console’s generative-AI performance report (rolling out from June 2026) gives impression data for AI Overviews, AI Mode, and generative Discover — but no clicks yet. Beyond that, manual periodic prompting and referral-traffic segmentation are the realistic proxies; precise, automated AI-citation tracking isn’t available industry-wide yet.
Is GEO/AEO a replacement for SEO? No. The tactics overlap heavily — clear structure, authoritative content, entity clarity — and AI-search optimization sits on top of sound technical SEO rather than replacing it.
Why would an AI tool cite Reddit or LinkedIn over my company’s own website? Because AI answer engines currently weight certain high-discussion, frequently-updated platforms heavily in what they cite. Being discussed in those channels — not only publishing on your own site — is part of being found by AI tools today.
References
- Google Search Central Blog, “Changes to HowTo and FAQ rich results” (Aug 2023). developers.google.com/search/blog/2023/08/howto-faq-changes.
- Google Search Central, “FAQPage structured data.” developers.google.com/search/docs/appearance/structured-data/faqpage.
- Google Search Central, “Merchant listing (Product) structured data.” developers.google.com/search/docs/appearance/structured-data/merchant-listing.
- Google Search Central Blog, “Introducing Search Generative AI performance reports in Search Console” (3 June 2026). developers.google.com/search/blog/2026/06/gen-ai-performance-reports.
- eMarketer, “FAQ on GEO and AEO: Where AI search and SEO overlap” (2026, citing October 2025 data). emarketer.com/content/faq-on-geo-aeo–where-ai-search-seo-overlap-2026.
- OSNDT, “Top SEO Strategies for NDT Websites” (15 Jan 2025). onestopndt.com/ndt-articles/top-seo-strategies-for-ndt-websites.
- OSNDT, “Importance of Local SEO for NDT Companies” (7 Feb 2025). onestopndt.com/ndt-articles/importance-of-local-seo-for-ndt-companies.
- OSNDT, “Top Website Features Every NDT Website Needs” (4 Feb 2025). onestopndt.com/ndt-articles/top-website-features-every-ndt-website-needs.
- OSNDT, “Importance of Website Speed” (28 Jan 2025). onestopndt.com/ndt-articles/importance-of-website-speed.
- OSNDT, “Importance of Responsive Design for NDT Websites” (17 Jan 2025). onestopndt.com/ndt-articles/importance-of-responsive-design-for-ndt-websites.
- NDT Marketing & Growth Playbook — broader marketing strategy this SEO/AI-search guide fits into
- NDT Marketing Benchmark Report — channel cost and spend benchmarks for digital marketing
- Getting Real ROI From NDT Trade Shows — turning trade-show content into on-site SEO assets
- Digital NDT: AI, Automated Defect Recognition, Digital Twins and Data — how AI is reshaping NDT inspection itself, not just search




There are no comments yet!