How organic search became Sova's

The situation in March

Sova was running an outbound-heavy GTM motion: a 14-person SDR team, ~$11k blended CAC, and a pipeline mix that was 71% cold outbound, 23% partner-sourced, and 6% inbound. Of that inbound, less than 1% was attributed to organic search. The category, pre-hire talent assessment for enterprise, wasn’t competitive in SEO terms. It was empty.

That was the opportunity. And also the risk: empty SERPs usually mean the buyer isn’t searching.

The buyer-search audit that changed the plan

Before we wrote a single page, we ran 90 minutes of buyer-style searches against Google, ChatGPT, and Perplexity. We posed as a CHRO shopping for an assessment platform, every query we could imagine a real one running.

The pattern was unambiguous: buyers were searching, frequently and in detail, but the queries weren’t head terms. They were comparison (“Sova vs Criteria”), workflow (“validity coefficient pre-hire assessment”), and skeptic (“are pre-hire personality tests legally defensible”). The platforms answering those queries were either academic sources or platform-incumbent thought leadership, both adjacent to the buyer, not aligned with them.

What we built

A buyer-mapped content tree. We grouped the 184 surfaced queries into 11 clusters and committed to one cornerstone page per cluster, each backed by 3-5 supporting articles. The cornerstone pages ran 1,400-2,200 words; the supporting articles ran 600-900. No fluff, no “what is pre-hire assessment?” intro paragraphs.

A legal-defensibility microsite. The skeptic queries were the highest-intent and the lowest-trust. We spun a small subfolder (/legal-defensibility/) authored with Sova’s general counsel, covering EEOC compliance, validity reporting, and adverse-impact analysis. Eight pages, three months of legal review, and the entire microsite ranked top-3 within ten weeks.

Programmatic comparison pages. Sova-vs-competitor pages, one per realistic alternative, each with a structured frame: capability matrix, honest assessment of where Sova lost, ICP fit signal. We didn’t write hit pieces; we wrote pages a real CHRO would forward to procurement.

“The legal-defensibility microsite was the unlock. The minute we started showing up for compliance queries, our pipeline mix flipped within a quarter.”

The numbers behind the +167%

Demo requests in month one of engagement: 18. Month four: 48. Month eight (handoff month): 78, sustained for the following three months post-handoff. The lift wasn’t linear; it inflected sharply around week 16 when the legal-defensibility cluster started compounding.

CAC moved with it. Blended CAC dropped from $11k to $6.8k over the same period, driven entirely by inbound mix shift, outbound CAC was unchanged.

What we’d do differently

Underweighted ChatGPT and Perplexity citation work in the first two months. By month four it was clear that 30% of high-intent searches were happening on AI platforms before Google. We retrofitted AEO patterns into the existing pages in month five, should have been parallel from week one. Now standard procedure on every new engagement.