Case study · 90-day teardown
A SaaS discovery platform serving over 100,000 monthly software buyers discovered ChatGPT was recommending five comparison sites for every category query — and they weren’t one of them.Ninety days later, +34 visibility points across six AI models. This is exactly what we shipped.
AI visibility composite · 6 models
High-intent prompts won
High-authority citations
New organic signups from AI traffic
The wake-up
The founder noticed it first inside their own buying behavior — when evaluating new software for the company, they were typing prompts into ChatGPT instead of opening a Google tab. The same shift was happening to their audience: SMB software buyers, founders, ops leads, the people their entire platform was built to serve.
When we ran the diagnostic, the picture was uncomfortable. ChatGPT was answering “best CRM for a 30-person company” with five named sources — G2, Capterra, a handful of tech blogs, and one comparison site that wasn’t them. Of 247 buyer-intent prompts we scanned, their brand appeared in none of the top 20.
A platform built specifically to be the answer to those questions wasn’t in the answer.
Day 0 · the baseline
/ 100The work
No mystery work. No quarterly “deliverables” report. Every week shipped real changes that we’d already mapped to specific prompts in the target list. The dashboard tracked which prompts moved as each piece landed.
Weeks 1–2
Ran 247 buyer-intent prompts across all six AI models. Mapped the categories where competitors were winning (G2, Capterra, software review aggregators) and the uncontested prompts nobody owned. Identified 18 high-intent prompts as the 90-day target.
247 prompts scanned · 18 priority targets identified
Weeks 3–6
Schema markup on every category page (SoftwareApplication + Review + AggregateRating). Restructured comparison pages for citation extractability. Got the brand named correctly in their public footer + author bios. Crawlability fixes for GPTBot, ClaudeBot, and PerplexityBot.
142 pages updated with schema markup · 4 crawl blocks fixed
Weeks 5–10
Built citation-grade content on third-party sites: 6 guest features on industry publications, 4 podcast appearances, contribution to 3 industry "top sites" lists. Every piece engineered to be cited as a source by AI — not just to drive referral traffic.
7 high-authority citations earned
Weeks 8–12
Published 22 "Software A vs Software B" pages targeting the exact phrasings buyers use with AI. Each engineered for citation: structured comparison tables, named experts quoted, FAQs that mirror prompts AI receives. These pages became the most-cited assets by week 10.
22 comparison pages · 11 ranking as AI sources
90 days later
"best CRM for a 30-person company"
"affordable Salesforce alternatives"
"best project management tool for marketing teams"
"top SaaS comparison sites for software buyers"
"where to compare HR software for startups"
"is HubSpot or Salesforce better for early-stage"
"best free CRM alternatives to Salesforce"
"top SaaS review sites that are not G2"
We didn’t need another content agency. We needed the team that understood the AI search shift was already happening to our audience. The Radar dashboard showed us which prompts we were losing, in real time, every week — and the work shipped against that list, not against an opinion.
The math
Twelve-week retainer at $5,000/month — $15,000 total. The work shipped on schedule. The score moved every week the dashboard refreshed.
Six months in, SpotSaaS attributed ~22% of new organic signups to AI-referred traffic — measured via GA4 referrer tracking for chat.openai.com, perplexity.ai, and claude.ai — in categories where they’d been functionally invisible at engagement start. The compounding has continued: by month nine they were the cited source for three more categories that weren’t even on the original target list.
This is now their organic moat. SEO compounded over years. AEO compounds in months because the field is empty.
What this would look like for your brand
Free audit. Same diagnostic that opened this engagement. Snapshot of your visibility across the six AI models we track — and the three highest-ROI fixes we’d ship first.
Run my free audit