Case study: 3.5× more AI traffic in the e-commerce CRM niche
How a niche SaaS CRM started appearing in AI comparisons next to the major platforms — and grew from 14 to 96 demo requests a month from the AI channel.
- Niche
- SaaS CRM for e-commerce
- Period
- Oct 2025 — Mar 2026 · 6 months
- Coverage
- 7 AI models · 280 prompts
- Status
- Ongoing · NDA
A niche product in the shadow of the generalist giants
The client is a SaaS CRM built for online stores: ready-made integrations with marketplaces, delivery services and telephony. In its own scenario the product is genuinely stronger than the generalist platforms — but the AI models barely knew it existed.
For broad queries like “best CRM”, the models named the large generalist platforms. The client's brand showed up in 11% of answers, and non-branded AI traffic — from comparisons and shortlists — was close to zero.
Objective
- Appear in comparisons and recommendations for niche “CRM for an online store” prompts
- Grow beyond branded AI traffic into non-branded — from comparisons and shortlists
- Grow demo requests from the AI channel with clean attribution through to the deal
Betting on niche comparisons
- Month 1 — Mapping the comparisons
- A map of 280 comparative “CRM for e-commerce” prompts across 7 AI models
- Analysis of which platforms models cite when recommending SaaS
- Technical audit: Schema.org (SoftwareApplication), llms.txt, feature pages
- Months 2—3 — Content built for what models get asked
- “X vs Y” comparison pages and shortlists targeting the niche prompts
- Integration guides: marketplaces, delivery, payment services
- Expert breakdowns on Habr and vc.ru with real client substance and numbers
- Months 4—6 — Ratings and attribution
- Reviews and ratings on the platforms buyers use to pick SaaS
- End-to-end analytics: from a mention in an answer to a demo request
- Monthly SoV measurements and prompt-map corrections
Six-month trajectory
Aivo Lab AI visibility data across a map of 280 comparative prompts, plus Yandex.Metrica and the client's own CRM data. The project is ongoing.
- AI Share of Voice: 11% → 29% (+18 pp in 6 months)
- AI traffic / month: 880 → 3,128 (3.5× in 6 months)
- Citation Rate: 12% → 27% (+15 pp in 6 months)
- AI demo requests / month: 14 → 96 (6.9× in 6 months)
What worked
Niche beats scale
On broad “best CRM” prompts, beating the giants is close to impossible. The growth came from narrow comparisons — “CRM for an online store” — where the product is genuinely stronger.
Comparison pages work
Models like structured comparisons. “X vs Y” pages with honest tables became the most-cited content of the project.
AI traffic converts better
A visitor arriving from an AI answer already has a recommendation in hand: conversion to a demo request is 1.8× higher than from classic search.