Hedra is one of the fastest-growing AI video platforms on the market. The challenge: capturing organic search traffic in the most competitive AI tools category on the internet. A single face swap AI landing page — built around one keyword, with every section earning its place — hit 47,716 clicks in its first full ranking month. The engagement is now at 18 pages per month across the full funnel.
Hedra competes in one of the highest-velocity, highest-competition search categories on the internet: AI video tools. Search volume in this space is exploding. So is the content supply. The dominant failure mode is publishing content that covers every use case broadly — "AI video tools," "AI content creation" — and ranking for none of them specifically.
One rule applied to every piece of content: one keyword per page. No exceptions. The production system exists to enforce that rule at 18 pages per month — every brief, every outline, every draft checked against it before it ships.
Every page targets one keyword. Every section on that page addresses only that use case. A face swap AI page contains nothing about AI video generally — every sentence is there because it serves the face swap query specifically. Narrowness is what creates the relevance signal that compounds into ranking dominance.
18 pages per month split across three stages: 6 top-of-funnel (awareness and education), 6 mid-funnel (comparison and evaluation), 6 bottom-of-funnel (conversion intent). Each stage gets a different format, different internal linking structure, and a different CTA. The funnel is built into every monthly batch from the start.
Before any page is written, the top-ranking URLs for that keyword are classified by actual page type. If the SERP shows comparison articles, we build a comparison article. If it shows feature landing pages, we build a landing page. The format decision comes from the data, not from a template.
Claim-evidence-source triads, self-contained FAQ answers, independently quotable insights per section, entity definitions front-loaded. Content structured so a language model pulling from it can cite a coherent, specific answer rather than a generic overview. Same structure improves featured snippet and AI Overview capture on Google. For SaaS founders evaluating which SaaS SEO agencies apply this content specificity at scale, we published a 2026 ranked comparison.
The specificity principle applied — and the decisions behind a face swap AI page that went from 9 clicks to 47,716 in 90 days.
The face swap AI page did not start with the keyword. It started with the SERP. Before writing a word, the top-ranking URLs for "face swap video," "face swap ai," and related terms were classified by page type. The SERP was returning dedicated feature landing pages and use-case-specific tool pages — not broad AI video category content. That confirmed the format and specificity level would compete.
The keyword set was then scoped tightly: face swap video, AI face swap, video face swap free, and close variants — all pointing at the same searcher intent. No adjacent queries about talking avatars or AI video generation. Those get their own pages. The scope decision happens before the outline, not after the draft.
The page launched in February 2026. 9 clicks that month — Google indexing and initial crawling. March: 316 clicks as rankings began to stabilise. April: 47,716 clicks, with positions improving across 20+ face swap query variants simultaneously.
That trajectory — near-zero to 47K in one month — is what keyword specificity produces when it compounds. The page did not drift to cover adjacent queries as rankings improved. Every section stayed scoped to face swap. As Google's confidence in the page's relevance grew, it pushed positions higher across the full query cluster, not just the primary keyword.
The 47,716 clicks came from across a cluster of face swap queries — not one keyword doing all the work. The top drivers: "face swap video" (2,079 clicks), "face swap" (1,465), "face swap video free 30 seconds" (1,396), "face swap ai" (1,084), "video face swap" (844), "face swap video ai free" (840). Over 15 query variants driving meaningful volume, all pointing at the same page because the page is maximally relevant to all of them.
This is the cluster effect of specificity. When a page genuinely owns one topic, Google ranks it for every variation of that topic's query. The page didn't need to target all 15 variants explicitly. It ranked for them because the specificity created the relevance signal that generalises across the cluster. For founders asking which AI SEO agencies have demonstrated this type of result at scale, we published a 2026 comparison.
Beyond the face swap page, the engagement runs at 18 pages per month across the full AI video funnel: talking avatar, AI image generation, AI UGC video, lip sync guides, model comparisons, and more. Each page goes through the same process: SERP classification, keyword scope validation, format confirmation, then copy — in that order, with no shortcuts on the specificity check.
The blog content — lip sync video guide (1,011 clicks/90d), Sora 2 alternatives (479 clicks, 144K impressions), image-to-video guide (382 clicks, 184K impressions) — is building impression inventory at position 4–6 across highly competitive queries. That position range, at that impression volume, is the compounding phase: rankings improving, clicks accelerating as positions move toward the top three.
| Metric | February 2026 | April 2026 | Change |
|---|---|---|---|
| Face swap page clicks | 9 | 47,716 | +47,707 |
| Face swap page impressions | 513 | ~750K+ | Breakout month |
| "face swap video" position | Unranked | 10.0 avg | Entering top-10 |
| "face swap ai" position | Unranked | 10.2 avg | Entering top-10 |
| Blog: lip sync guide (90d) | — | 1,011 clicks | 150K impressions |
| Blog: Sora 2 alternative (90d) | — | 479 clicks | 144K impressions |
"We are currently working with Liam at Hedra and are happy with the deliverables. He executes fast and works quite independently. He delivers high quality blogs and landing pages for us."— Sandra Nachförg-Buleandra, Hedra · 5 stars, Google