MRR doubles after fixing what assistants said the product was
Buyers shortlist via ChatGPT before ever visiting a vendor site — and the assistants have the category wrong. Correcting it models monthly recurring revenue from $500K to $1.03M.
| Measure | Before | After |
|---|---|---|
| Monthly recurring revenue | $500K | $1.03M |
| Added cash flow per month | — | +$530K |
| Prompts citing the product, of 180 | 11% | 58% |
| Correct category attribution | 34% | 96% |
| Sales cycle length | 90 days | 64 days |
The problem
Sales kept hearing the same thing on discovery calls: the buyer had already narrowed to three vendors before making contact, and half said the shortlist came from an AI assistant. The company was named in 11% of 180 category prompts, and when it was named it was described with the wrong category — “applicant tracking” rather than “onboarding automation” — which pushed it into comparisons it always lost. Nurture email was a single monthly newsletter to every contact regardless of stage.
What the model found
Two years of won and lost CRM notes go to a model with one question: separate the language of deals that closed from deals that stalled. Closed deals talk about implementation time; stalled ones talk about integrations. That reframes both the comparison pages and the nurture sequence around time-to-live.
The plays
Re-anchor the category claim
The single highest-leverage fix: one category phrase, used identically on the homepage, in schema, on review platforms, LinkedIn, Crunchbase and every press mention. Models had been averaging three conflicting descriptions.
Publish the comparisons buyers ask for
Nine honest head-to-head pages including the two giants, with a “choose them if” section. Refusing to write these leaves the field to affiliate sites with no incentive to be accurate.
Turn the docs site into a retrieval asset
Public docs, changelog and an implementation guide with defined terms and time estimates. Technical buyers ask assistants implementation questions; the answers now come from first-party docs.
Rebuild nurture around time-to-live
Five stage-based sequences replacing the monthly newsletter, each opening on the implementation-time claim the CRM analysis surfaced.
Wire citations to pipeline
Prompt-level tracking joined to CRM so we can see which prompts produce opportunities, not just mentions. Content priority follows closed revenue.
The lever
Category consistency. Nothing else moves the number until every source agrees on what the product is — after that, every other tactic compounds instead of fighting itself.
How the model was calibrated
Built on Semrush’s finding that 50% of the links in ChatGPT responses point to business and service sites, and that its citation pool sits well below page-one rankings. ChatGPT’s weekly active users grew roughly 8× between October 2023 and April 2025 to over 800 million, which is the reason a shortlist formed inside an assistant is now a routine part of a B2B buying cycle rather than an edge case.
Sources
- Semrush — AI Search & SEO Traffic Study — ~90% of ChatGPT search citations point to pages ranking 21+; AI search visitors convert at 4.4× ordinary organic; 50% of ChatGPT links go to business and service sites.
- Semrush — AI Overviews Study — 200,000 keywords. AI Overview links overlap the organic top 10 only 20–26% of the time; an average AI Overview carries 11 links and includes the #1 organic result just 46% of the time on desktop, 34% on mobile.
- Litmus — State of Email / ROI of Email Marketing — Return per $1 spent by sector: retail, e-commerce and consumer goods 45:1; agencies 42:1; software and technology 36:1.
Want the same read on your category?
Send a brief and we will run twenty prompts before we talk.