3 Tactics That Increased Our Client's Visability in AI Search by 40%

"Why doesn't ChatGPT know we exist?"

That was the frustrated question we were asking ourselves after we had just asked Claude to recommend an embedded AP/AR automation API, and watched it enthusiastically suggest Bill.com, Tipalti, and even Stripe. But Monite - our client who'd built arguably the best API in the space - didn't even get a mention.

The Wake-Up Call

We spent the next hour testing every AI tool we could find. The results were frustrating:

ChatGPT: Recommended Bill.com in 9 out of 10 queries

Claude: Knew Tipalti's entire feature set by heart

Perplexity: Could quote Stripe's documentation verbatim

Monite: Cricket sounds

Monite's API was technically superior to most of these options. Better documentation, faster implementation, more flexible. But none of that mattered if AI couldn't find them.

Traditional SEO wasn't the problem. They ranked well on Google. But we knew that B2B buyers (particularly in the SaaS space) had moved on from Google to AI chat interfaces.

Tactic 1: We Turned Case Studies Into Technical Goldmines

Our first hypothesis: AI tools trust technical content more than marketing fluff.

So we took Monite's generic case study about "increasing efficiency by 40%" and transformed it into a implementation deep dive including code snippets such as:

Within two weeks, Claude started citing these technical details when asked about implementation complexity. We'd given AI the ammunition it needed to recommend Monite confidently.

Tactic 2: We Created Competitor Pages That Actually Helped

This felt counterintuitive - why should you create pages about your competitors? But here's what we learned: AI tools look for balanced comparisons to seem unbiased.

Instead of "Why Monite is so much better than Bill.com" (which AI would ignore as propaganda), we wrote:

"Monite vs Bill.com: Honest Comparison for Developers"

We included a real comparison table:

The twist? We actually recommended Bill.com for certain use cases. This honesty made AI trust the entire page more.

Tactic 3: The LLM Optimisation That Almost Backfired

We restructured every page to be scannable by LLMs:

Before: Long paragraphs about "synergising payment workflows"

After:

• Clear bullet points

• Headers that answer questions

• Code snippets every few paragraphs

• "People Also Ask" sections

But then something weird happened. After two months, our content started reading like robot food. Bounce rates increased. Humans hated it.

So then we decided to rewrite parts of the post with more of a paragraph, narrative structure to keep it engaging for users. Thankfully, AI still loved the structure, but humans actually enjoyed reading it.

The Results

Three months later:

• ChatGPT recommended Monite in 4 out of 10 queries (up from zero)

• Claude could explain Monite's entire webhook system

• Perplexity cited our technical docs directly

• Overall AI search visibility up 40%

But here's what really mattered: Inbound leads mentioning "ChatGPT recommended you" increased 2.3x.

People Also Ask

Q: How did you measure AI search visibility?

A: Weekly programmed queries across 5 AI tools with 20 different phrasings of our target keywords. Tracked mentions, position, and detail level. Built a simple spreadsheet, nothing fancy. We've tried a few of the AI citation tools out there and not excited by any of them. We're sure something else will come along soon!

Q: What content performed best for AI optimization?

A: Technical implementation guides with actual code snippets. AI tools seem to weight these 3x higher than marketing content.

Q: How long before you saw results?

A: First mentions after 6 weeks. Significant presence by month 3. ChatGPT was fastest to update, Claude took longest.

Q: Did traditional SEO metrics drop?

A: Surprisingly no. Human engagement actually increased 15% after we added back the conversational tone.

Q: What's the one thing that didn't work?

A: Keyword stuffing for AI. We tried repeating "embedded AP automation API" everywhere. AI tools actually penalised us for this - they're smarter than we thought.