Case study · JoVE Visualize

How JoVE Visualize got 6.7× more AI citations in 4½ months

Visualize is JoVE's library of science articles and videos. In my first 4½ months as Sr. SEO Strategist, we fixed how search engines and AI assistants crawl it, rebuilt how its pages get indexed, and tied the traffic to sales pipeline.

Role
Sr. SEO Strategist
Period
May – Sep 2026
Site
visualize.jove.com

The results

  • $100KPipeline generatedorganic-sourced, tracked in CRM
  • 6.7×AI answer citations (Bing)209K → 1.4M
  • 3.4×Organic users (GA4)37.9K → 127.8K
Growth by metric, before vs after 4½ months before (= 1×) First 4½ months

Hover or tap a row for the raw numbers.

  • Bing clicks9.1×
  • AI answer citations6.7×
  • Bing impressions6.1×
  • Pages cited by AI5.1×
  • Google video clicks4.3×
  • Organic users3.4×
  • Google web clicks1.9×
  • Google web impressions1.5×
From AI assistants11.4%of Visualize's users now arrive from AI assistants like ChatGPT: 16,408 people in 4½ months.GA4 · first-user channel

Where users come from now

GA4
88%organic search
  • Organic Search127,784 users88.4%
  • AI Assistants16,408 users11.4%
  • Referral717 users0.5%

May 11 – Sep 23, 2026. A user can arrive through more than one channel.

See the raw numbers
Before and after, by metric
MetricBeforeAfterChange
Bing clicks11.2K101.8K9.1×
AI answer citations209.2K1.4M6.7×
Bing impressions851.3K5.2M6.1×
Pages cited by AI6263.2K5.1×
Google video clicks3.5K15K4.3×
Organic users37.9K127.8K3.4×
Google web clicks48.9K94.5K1.9×
Google web impressions19.9M29.9M1.5×

Visualize launched in September 2025. I joined in May 2026, so these results compare my first 4½ months with the 4½ months before I joined. Both periods are after launch; there's no like-for-like year earlier.

What we changed

  1. Fixed what bots actually see

    Server-side rendering was serving fallback API content to Googlebot and blocking AI crawlers like GPTBot and ClaudeBot from JavaScript-dependent content. Bot log analysis showed where crawl budget was being wasted.

  2. Cleaned up crawl errors

    Fixed 404s, 500 errors and broken redirects, so crawlers spend their time on pages that should rank.

  3. Rebuilt sitemaps for indexing

    Split URLs into smaller, focused sitemaps and redefined how they're submitted, then built a pipeline (Python, Playwright, Search Console, BigQuery) to track indexing per sitemap. Indexed pages grew 300%.

  4. Corrected the structured data

    Moved article pages to ScholarlyArticle schema (MedicalScholarlyArticle, with MeSH topics and author data), so search engines and AI assistants understand what each page is and who wrote it.

  5. Fixed international SEO

    Resolved hreflang conflicts across 5M+ pages, caused by two systems generating mismatched URLs for the same page.

  6. New content and page structure

    Published new content with a clearer page structure, and tested three article layouts against each other on pages that were seen often but rarely clicked.

  7. Used jove.com's authority

    Added valuable content on the main jove.com site, drawing on its long-established authority to support Visualize.

We ran these as a series of experiments. I found and prioritised the issues and wrote the technical requirements and tickets; JoVE's developers built them, and I tracked the impact in Search Console, Bing Webmaster Tools and GA4.

From search to pipeline

How organic search turns into pipeline

$100K pipeline

How it's measured

  • Bing Webmaster Tools · AI Performance (AI answer citations, pages cited)
  • Bing Webmaster Tools and Google Search Console · visualize.jove.com
  • GA4 · users by first-user channel
  • CRM · organic-sourced pipeline

Technical SEO lead in Jaipur, India. Open to roles across EMEA and to consulting.

--:-- in JaipurProbably at the desk

© 2026 Nikhil Sharma

Designed & built by Nikhil Sharma