Template platforms and the AI discoverability gap: a vendor-by-vendor assessment

The question I'm most often asked, on the first call of a new engagement, is whether the firm's existing template platform is good enough or whether a rebuild is warranted. The answer, in most cases, is platform-specific. Each major template platform has a different position on AI-search readiness, and the gap between what the platform exposes by default and what the discipline now requires varies materially.

This assessment is a frank, vendor-neutral view of where the major platforms stand. I have no commercial relationship with any of the platforms named below and no affiliate revenue from any link. The assessment reflects work done across client engagements over the past eighteen months.

The criteria are the same for each platform:

Schema.org coverage. Does the platform render the core types (Organization, Service, Person, WebSite, BreadcrumbList, FAQPage) by default, with the right level of completeness?
llms.txt support. Does the platform let the firm serve a llms.txt at the domain root?
Content depth tolerance. Does the platform's editor make it easy or hard to publish 1,500-word pieces with structured headings, citations and internal links?
Custom code access. Can the firm add custom JSON-LD, custom meta tags, custom robots directives, and custom head content where the defaults are inadequate?
Indexability discipline. Does the platform produce clean URLs, sensible canonicals, and a properly structured sitemap.xml?

The assessment is current at the time of writing. Platform features change; an annual revisit is warranted.

Squarespace

Schema.org coverage. Squarespace renders a baseline Organization and WebSite markup on most templates, with a modest set of properties. LocalBusiness markup is exposed where the site has a configured business address. The platform does not expose Person, per-service Service entities, or FAQPage markup by default. Custom JSON-LD can be injected via the Code Injection feature, but the work has to be done by hand for each page that needs it.

llms.txt support. Squarespace does not natively support serving a llms.txt. The available workarounds (a workaround URL on a subdomain, a hosted static file via Cloudflare, an unsightly /llms-txt/ page) are all imperfect.

Content depth tolerance. The blog editor handles long-form content acceptably. The page editor is harder for non-blog content over 1,000 words; the section-based layout system makes it awkward to flow large amounts of prose.

Custom code access. Available on Business and higher plans. The Personal plan doesn't include it, which is a material limitation.

Indexability discipline. URLs are reasonably clean. Sitemap is auto-generated. Canonicals are present but, on some templates, point to URLs the firm didn't intend.

Verdict. Acceptable for firms that have already chosen Squarespace and aren't willing to migrate, provided they're on the Business plan or higher and willing to do meaningful custom work. Not the platform I'd recommend for a new build that takes AI discoverability seriously.

Wix

Schema.org coverage. Wix has, in the last two years, materially improved its structured-data defaults. Organization, WebSite and (where appropriate) LocalBusiness markup is rendered automatically. The Wix Editor exposes some schema.org settings through its SEO panel, but the properties exposed are a subset of the useful set. Custom JSON-LD can be injected through the SEO panel's advanced section.

llms.txt support. Wix added basic llms.txt support in 2025 through its SEO settings. The implementation is limited (the firm can supply the content but the file is generated and served by Wix rather than directly editable as a Markdown file), and the workflow is awkward, but it exists.

Content depth tolerance. The Wix editor is structurally hostile to long-form. The editor is built around drag-and-drop section blocks rather than flowing text, and pages over 1,500 words become difficult to edit. The Wix Blog editor is more forgiving but limited in formatting options.

Custom code access. Available on the Business and higher plans. The editing workflow is, by template-platform standards, reasonable.

Indexability discipline. URLs are clean, sitemap is auto-generated, canonicals are present.

Verdict. Acceptable for service firms whose content strategy is centred on shorter-form pages and a modest blog cadence. Not appropriate for firms that intend to publish substantively long-form content.

Webflow

Schema.org coverage. Webflow's default schema.org coverage is, frankly, thin. The platform doesn't render Organization, WebSite or any other core markup by default. The firm must hand-author all JSON-LD and inject it via the per-page or site-wide custom code fields.

llms.txt support. Webflow doesn't provide a native llms.txt mechanism. A llms.txt can be served by hosting it through the platform's static assets feature, with some workaround on the build configuration. Not a clean solution.

Content depth tolerance. The CMS is, for blog and article content, one of the better template platforms. Collection-based content allows for long-form pieces with proper structure.

Custom code access. Excellent. Per-page and site-wide custom code is exposed, with no plan-tier restrictions on content-modifying access. The CMS allows custom embeds and rich text with fewer restrictions than competitors.

Indexability discipline. URLs are clean, sitemap is auto-generated, canonicals are exposed.

Verdict. A platform that requires the firm to do almost all the structured-data work by hand, but rewards that work with broad customisation. Acceptable for firms with a designer-developer on hand or with the budget to retain one. Not appropriate for firms that need the platform to ship sensible defaults.

WordPress

Schema.org coverage. Varies entirely with the theme and plugins in use. The base WordPress install ships almost no structured data. A site running Yoast SEO or Rank Math will have meaningful Organization, WebSite and (with configuration) Person and LocalBusiness markup. A site on a poorly configured commercial theme can have markup that's partial, contradictory or, in some cases, actively wrong.

llms.txt support. Available through several specialist plugins as of 2025. The implementations vary in quality; manual hosting via the theme's static assets is also viable.

Content depth tolerance. Excellent for long-form. The Gutenberg editor is, in its mature form, a workable long-form environment.

Custom code access. Full. The platform is open-source and exposes every layer.

Indexability discipline. Varies with theme and plugins. A properly configured site is excellent; a misconfigured site can be a discoverability disaster.

Verdict. With the right plugin stack and a careful configuration, WordPress is fully capable. The variance in practice is enormous, and a meaningful share of audited WordPress sites have non-trivial discoverability problems.

Shopify

Schema.org coverage. Strong on Product, Offer, Review and AggregateRating. Shopify's commercial focus shows. Weaker on Organization, ProfessionalService and the firm-as-a-source markup that matters for service businesses. Acceptable for product businesses; suboptimal for service businesses using Shopify as a general website.

llms.txt support. Not natively supported. Workarounds through theme files exist but are awkward.

Content depth tolerance. Adequate for blog content. Not designed for long-form non-commerce content.

Custom code access. Available on the Basic plan and above for theme code. Liquid template language requires some familiarity.

Indexability discipline. Strong on the commerce side. Acceptable elsewhere.

Verdict. The right choice for commerce businesses exclusively. Not the platform a professional services firm should be on, regardless of the AI-discoverability question.

The summary

For a professional services SMB choosing a platform today:

The cleanest position is a modern static site (Eleventy, Astro, Next.js) custom-built and hosted on a CDN. This is the position the Build package delivers.
WordPress with a careful plugin stack is the second-best position for firms that need a CMS-driven workflow.
Wix and Squarespace are acceptable for firms on those platforms today, with meaningful custom work.
Webflow is acceptable for firms with design-development capacity in house.
Shopify is, for service firms, the wrong tool.

The audit framework covers the specific signals to test on whichever platform the firm is on. The retrofit work, where a rebuild isn't warranted, focuses on the cost-effective interventions within the constraints of the existing platform.

The decision of whether to rebuild is, ultimately, a strategic one rather than a technical one. A firm whose website is otherwise serving its commercial purposes well shouldn't rebuild lightly. A firm whose website is already under-performing and whose platform actively constrains discoverability has a clearer case to act.