Getting named in 10 AI answers in 90 days

Before and after block

VISIBILITY CHANGE

Before and after

Brand mentions across the tracked prompt set, at programme start and at 90 days.

BEFORE Baseline mention count to be added

AFTER Cited across 10 tracked prompts

Add the starting figure before this page goes live. If the brand was mentioned in zero prompts at the start, say zero. That is a far stronger left-hand side than a blank, and it is the number this whole page rests on.

---

Brief about the Problem

Buyers in fintech research before they shortlist, and a growing share of that research no longer happens on a results page. It happens in a conversation with an assistant, where the buyer asks which providers are worth considering and gets back a short list of names.

If your brand is not in that list, you are not in the running, and you will never see the impression in any analytics tool. The client was investing in content and ranking respectably in traditional search, but was absent from the answers being given to the exact questions their buyers were asking.

---

Case Study Unveiled

The problem was not content volume. It was that the content had been written to rank rather than to be quoted, and those are different jobs.

Pages built around keywords tend to circle a topic broadly. Pages that get cited answer one specific question with enough precision that an assistant can lift the answer and attribute it. The client had plenty of the first kind and almost none of the second.

There was also no measurement in place. Without a fixed set of prompts to track, there was no way to know whether anything was working, which meant no way to justify continuing.

---

What Did We Do

We started by fixing the measurement. A set of ten prompts was defined, covering the questions a real buyer asks before shortlisting a provider in this category, and the brand's presence in the answers to each was recorded as a baseline.

Content was then rebuilt around those questions rather than around keyword targets. Each priority question got a page that answers it directly, specifically and early, with the kind of detail that makes a passage worth quoting rather than paraphrasing.

Underneath that, we worked on the signals that decide whether a brand gets named at all: consistent entity information across the web, third-party presence in the places these engines draw from, and structured content that makes the source of an answer unambiguous.

SEO and GEO ran together throughout rather than as separate workstreams, because the same depth that earns a citation also earns a ranking, and splitting them would have meant doing the work twice.

---

The Results

Within three months the brand was being cited across all ten tracked prompts, having started from a materially lower base.

Lead flow is now attributed to organic and AI-led discovery as separate sources, so the client can see what the AI visibility work contributes rather than having it disappear into a general organic number.

The tracked prompt set continues to run, which means visibility is monitored as a trend rather than claimed once and forgotten.

---

Before this page goes live

Add the baseline. The before column is the most important number on this page and it is currently blank. Ten out of ten means little without knowing whether the starting point was zero or eight.

Add the lead figure if you have it. "Lead flow attributed to organic and AI-led discovery" describes a capability, not a result. If there is a number behind it, it belongs in The Results.

Keep the industry descriptor at fintech. Anything more specific about the product category would narrow the field enough for someone in the sector to identify the client.

Check this against the Khatabook page. Both are fintech, both are confidential-adjacent, and both sit in your results grid. Make sure the imagery and layout make clear these are two different clients.

Confirm which assistants were tracked. If the prompt set covered ChatGPT, Gemini, Perplexity and AI Overviews, saying so makes the claim far more concrete than "AI answers" alone.