Bastille Networks

Bastille Networks makes invisible wireless threats visible for the Fortune 100, the Department of Defense and the intelligence community, and I was the engineer who ran their website, their analytics and the answer-engine optimization program that put them ahead of Cisco and Fortinet in AI share of voice.

Key Takeaways

  • Took Bastille to 45% AI share of voice across tracked buyer prompts, ahead of Cisco at 37% and Fortinet at 14%, with 100% visibility on US-compliance wireless monitoring prompts.
  • Moved average Google position from the 20s and 30s into single digits, with clicks up roughly 20% and brand searches up 59% per day over the comparison period.
  • Implemented Consent Mode v2 under outside counsel review with three layers of enforcement and no drop in lead flow.
  • Rewrote the research library for answer engines: 34 research pages recategorized, 125 video pages given rebuilt structured data, and long-form technical guides published across the wireless threat landscape.
  • Built a custom AI visibility tracker as a WordPress plugin, querying six AI providers daily and matching responses against brand and product keywords so visibility is measured, not guessed.
  • Stood up a BigQuery warehouse with daily ETL from HubSpot, Google Analytics and Search Console, feeding dashboards inside WP Admin where the team already works.
  • Cut image weight 60 to 80% sitewide and rebuilt asset loading so each page pulls only what it needs.

Organization

Bastille Networks is a Series C wireless security company. Its platform monitors 100 MHz to 7.125 GHz across 50+ protocols through ceiling-mounted software-defined radio sensor arrays, making rogue devices and wireless attacks visible inside secure facilities. Customers include the Fortune 100, the Department of Defense and the intelligence community.

Challenge

The buying audience for wireless security, CISOs, SOC managers and physical security directors, was increasingly starting research inside AI assistants rather than Google, and Bastille was structurally invisible there. Its deepest technical content sat behind email-gated PDFs that machines could not read, its original research carried no structured data, and the site itself had launched unfinished, with performance and analytics gaps to match.

Solution

  • An answer-engine optimization program: a published llms.txt company profile, verified social profiles, solution pages rewritten in answer format, and gated PDFs converted to indexable web content.
  • Consent Mode v2 with server-side GeoIP defaults, a consent banner, and per-tag GTM gating, documented for finance and legal with open decisions marked for counsel.
  • A custom AI Tracker plugin measuring brand visibility across six AI providers daily.
  • Analytics rebuilt end to end: BigQuery warehouse, Looker and in-admin dashboards, automated performance reporting, and custom event tracking down to every button on the site.
  • A full performance pass: images resized and converted to WebP, fonts subset, stylesheets and scripts split and loaded on demand, video embeds converted to click-to-load.
  • Launch support for time-sensitive marketing: partnership announcements, landing pages, a podcast hub and a careers page, delivered to fixed release windows.

Results

  • 45% AI share of voice, first in category.
  • Organic clicks up roughly 20%, homepage click-through more than doubled.
  • Consent rollout completed with lead flow unchanged.
  • Mobile page speed score up 30 to 50 points.
  • 8 to 15 hours a month of marketing and developer time removed.