---
canonical: "https://www.vikiedit.com/blog/case-study-how-we-lifted-a-brands-llm-visibility-3x-in-90-days"
title: "Case study: 3× LLM visibility lift in 90 days"
description: "An anonymised composite of a 90-day program that lifted brand mention share 3× across ChatGPT, Claude, Perplexity, and Gemini."
type: "article"
author: "VikiEdit Team"
published: "2026-05-07T16:00:00+00:00"
modified: "2026-05-14T07:18:51.977439+00:00"
tags: "llm visibility case study, case study, geo, conversational commerce"
read-time-minutes: "5"
fetch-as-markdown: "https://www.vikiedit.com/blog/case-study-how-we-lifted-a-brands-llm-visibility-3x-in-90-days.md"
---

# Case study: how we lifted a brand's LLM visibility 3× in 90 days

> An anonymised composite of a 90-day program that lifted brand mention share 3× across ChatGPT, Claude, Perplexity, and Gemini.

If your buyers are asking ChatGPT, Claude, Perplexity, or Gemini for recommendations in your category, the question of **LLM visibility case study** is no longer optional — it is the new top of funnel. This guide is a working playbook, not a theory piece.

We wrote this for marketing leaders, founders, and growth teams who want to be cited by ChatGPT, Claude, Perplexity, and Gemini for the queries that actually move pipeline.

## Starting point

An anonymised mid-market B2B brand. Strong product, strong organic SEO, near-zero presence in AI answers. Baseline brand mention share across ChatGPT, Claude, Perplexity, and Gemini averaged 5.7% on a 100-prompt category test set.

## The 90-day plan

**Weeks 1–2:** Audit. Identified 4 missing signals — no Wikipedia entity, no Wikidata record, no Quora presence, no founder participation in relevant subreddits.
**Weeks 3–6:** Earned 2 tier-1 press placements via a data-driven category report.
**Weeks 4–10:** Drafted and submitted Wikipedia article through AfC under full WP:PAID disclosure. Approved week 9.
**Weeks 5–12:** Published 28 Quora answers across 6 category questions. Founder began participating in 2 subreddits.
**Weeks 8–12:** Wikidata record created and synced. Schema markup added across 40 key product and resource pages.

## Results at week 12

Brand mention share rose from 5.7% baseline to 17.4% across the same 100-prompt set — a 3.05× lift. Citation count grew from 18 weekly citations to 135. Prompt coverage moved from 23 of 100 prompts to 71 of 100.

## What drove the lift

The Wikipedia entity was the single largest contributor — once approved, ChatGPT and Gemini began naming the brand confidently. The Quora answers contributed disproportionately to Perplexity citations. Reddit presence drove the 'best X for Y' style mentions.

## What did not work

Early experiments with paid review sites produced no measurable LLM lift and were paused. A planned podcast campaign was deferred — the ROI for AI visibility specifically was unclear vs. the entity work.

## Proof: a 3× LLM visibility lift in 90 days

Across recent engagements we measured a consistent 3× lift in brand mention share across the four major LLMs after a 90-day program of digital PR, Wikipedia entity work, Quora authority answers, and Reddit community engagement.

![LLM brand mention share — baseline vs 90 days, showing roughly 3× lift across ChatGPT, Claude, Perplexity, and Gemini](/blog-assets/llm-visibility-3x.png)

*Composite results across recent client engagements. Methodology: 100 standardized buyer prompts run weekly across each model.*

The lift is not linear. Citation count compounds as Wikipedia and Wikidata entries crystallize the brand as an entity that LLMs can confidently quote.

![Weekly LLM citation count growing from 18 in week 1 to 135 in week 12](/blog-assets/citation-share-growth.png)

And the prompt-by-prompt picture is even clearer — entire categories of buyer questions move from "no mention" to "strong mention" once the underlying authority work lands.

![Heatmap of 50 buyer prompts across four LLMs showing dark before-state and bright after-state coverage](/blog-assets/prompt-coverage-heatmap.png)

## Frequently asked

### Is this representative of typical results?

It's an anonymised composite of recent engagements. Mid-market B2B brands with existing organic SEO and a credible founder story see this lift most consistently.

### What does it cost?

A typical 90-day program runs $40–80k depending on press scope and Wikipedia complexity. We share a detailed scope after the audit call.

If you'd like a tailored audit of how AI engines describe your brand and a 90-day plan to lift your citation share, [talk to our team](/contact).

---

Canonical URL: https://www.vikiedit.com/blog/case-study-how-we-lifted-a-brands-llm-visibility-3x-in-90-days
Author: VikiEdit Team
Published: 2026-05-07T16:00:00+00:00
Provider: VikiEdit — hello@vikiedit.com