Domain authority 0 to 24, and showing up in AI answers

What was done

• Structure first. Service pages mapped to how buyers search, one page per offering, landing pages for each campaign, and an internal linking plan so new pages were never orphans.
• Technical SEO. A full crawl and a 45-issue audit with an owner for every fix: the http version serving without a redirect, thank-you pages sitting in the index, 104 case studies set to noindex by mistake, duplicated security headers, a slow server response and over a megabyte of unused JavaScript per page. Sitemaps submitted to Google and Bing and checked monthly, which is how an empty Bing sitemap was caught.
• Link building through directories, partner mentions, listings and the company's own announcements, measured as domain authority month by month.
• Answer-engine optimisation. Organisation and service schema, plain-language answers near the top of each page, a 12 KB llms.txt, robots.txt open to AI crawlers, and a fixed set of 63 category questions run through ChatGPT search, Google AI Mode and AI Overviews each quarter, so visibility in AI answers is a measured number rather than a feeling.
• Security. A sudden drop in paid search spend led me to a hack: hidden gambling links on every page and a cloaked page served only to Google's crawlers. I documented it with a repeatable check, had the hosting team clean it, re-scanned 150 URLs across five user agents to prove it clean, and excluded the contaminated weeks from every report so nobody planned on bad numbers.

Keeping the numbers honest

A headline organic growth figure once reached the leadership deck. It was true and it was useless: most of it was employees logging into a portal on the same domain and job seekers. Commercial pages had grown under 1%. I corrected the report and set a rule that has held since: no traffic number without a like-for-like comparison and a page breakdown next to it.

To make that rule cheap to follow, I built a marketing data warehouse that pulls GA4, Search Console and Google Ads into one model. Its first finding paid for itself: about two thirds of search spend was going to queries where the site already ranked in the top three organically. That spend was cut.

Where it started

Search is where I learned marketing. During college I built nine content websites from zero to first-page rankings, wrote over a thousand articles, and grew them to 300,000+ monthly organic pageviews with a recurring affiliate income. The same habits apply at a listed company; the stakes are just higher and the reporting has to be tighter.