Holding all else equal: how comprehensive schema drove +45% MoM organic clicks for a mid-market cybersecurity SaaS.

The client was a mid-market cybersecurity SaaS in the XDR space: a competitive category, enterprise-adjacent buyers, commercial-intent queries dominated by rich SERPs. Organic traffic had been growing slowly, but the site was systematically under-indexed relative to its content surface.

Two things were underneath it. The first: fragmentation. The foundational content (what the market calls a "learn" or "resources" hub) lived under 26 disconnected URL folders, many overlapping in topic, none acting as pillars. Google saw a sprawl, not a structure.

The second: illegibility to the engine. Across BOFU and blog surfaces, structured data was inconsistent or absent. Pages eligible for rich listings (product pages, FAQ blocks, review modules) were rendering as plain blue links. On commercial queries where competitors had stars, prices, and expandable FAQs, this client didn't. Impressions were stacking up; clicks weren't.

If the system is underperforming, the cleanest diagnostic is to change one thing and watch what moves. So I ran it that way: schema only, comprehensively. No new content. No internal-linking revamp. No technical cleanup wave.

Eight schema types, rolled out across the BOFU and blog page sets: Product, SoftwareApplication, FAQ, Review, Organization, Breadcrumb, Article, HowTo. The decision wasn't "which plugin." It was treating structured data as a systemic surface and implementing it everywhere it applied, not page-by-page when someone remembered.

In the same window I restructured the fragmented resources hub. 26 URL folders consolidated into 8 pillars under a single "Security Foundations" parent: Cybersecurity, Managed Service Providers, Attack Techniques, Endpoint Security, Incident Response, AI and Machine Learning, Threat Intelligence, and Managed Detection and Response. Each pillar maps directly to the client's product surface, XDR + MDR + MSP, not generic SEO topics.

The framing I used with the CMO was deliberately narrow: we are going to make your site more legible to the engine, and we expect the lift to show up first on commercial queries, where enhanced SERPs have the most leverage. Setting that expectation mattered. Senior stakeholders have seen every "aggressive content plan" and rightly discount them. A single-variable test that predicts a specific shape of outcome is a different kind of promise.

The last piece was measurement discipline. I instrumented the MoM window specifically so the read would be honest: Jan 2026 as the clean baseline, Feb 2026 as the first full month post-rollout. I also pre-registered the BOFU regex I'd use for the commercial-query cut (platform | software | tool | provider) so the result couldn't drift toward cherry-picked language after the fact.

The schema window came after a project phase that produced no content at all. It produced the map: who buys, what they search, what already ranks, and what the engine couldn't read.

1. Read what ranks. For the money keywords, competitors ranked their educational hubs, where you might expect product pages. Only two "managed" terms showed results willing to rank a product page, so those two pages got targeted optimization and everything else pointed at a hub. That finding is why the 8-pillar Security Foundations hub exists.
2. Size the prize. The priority money keywords added up to about 572,000 searches a year at an average position near 22. Moving them into positions 3 to 10 was worth roughly 32,000 more high-intent clicks a year.
3. Baseline the AI answers. In tracked prompts the brand appeared in 20% of responses. Its own pages supplied 3% of the citations; third-party sites supplied 86%.
4. Audit what the engine sees. Long titles and meta descriptions, H1s duplicating the title tag on 174 pages, redirecting URLs listed in the sitemap, and almost no structured data.
5. Set the north-star metric with RevOps. Organic sales-accepted leads, with LLM-referred traffic deliberately left out of the forecast.

From there each change got its own window where possible: schema and the hub consolidation in January and February, the URL migration in April with robots rules to clear crawl waste beside it, then an internal-link pass pointing every link at the new paths. Reporting ran every two weeks on a fixed shape: pipeline first, branded and non-branded traffic split apart, the bottom-of-funnel filter, and a set of money terms with their AI Overview status.

The pages that grew against a falling sitewide line were the refreshed ones: a tools comparison page climbed every month of the quarter and returned to the top two, and new vendor-comparison pages entered at positions 4 to 6. Two programs carry the plan into the next quarter. An FAQ retrofit across the hub's 280 pages, 222 of which had no FAQ at all, takes it to 1,501 questions built from tracked AI prompts and real search questions, rolled out 50 pages at a time. And the platform pages split from 7 to 17 by capability, each backed by a guide, a buyer's guide, a listicle, and thought leadership built on the client's own threat data, because comparison content converts but is the easiest content type to copy.