About the author

Robin Laires is a marketing tech and AI engineering specialist and the founder of Nest Content – a UK-based SEO and AI search optimization agency. He founded it after six years at Jellyfish, a marketing agency, where he built marketing tech for global brands.

Background

Nest Content is a small UK agency doing SEO and AI search optimization for service businesses. I run it and build most of the systems myself. More of our work this year has been AI search: checking whether Google’s AI Overviews, ChatGPT, and Perplexity mention a client when a buyer asks for a recommendation.

Every client gets a weekly report and a live dashboard on their own Search Console data, and the AI search part of the report runs on DataForSEO.

What I use DataForSEO APIs for

Weekly tracking for clients

I use the SERP API to run each client’s buyer questions on a schedule. It records what Google’s AI Overviews and AI Mode say, which sources they cite and whether the client is one of them. As it’s an API, I can track the exact questions each client’s buyers ask.

For a UK executive search firm, I track 54 buyer searches in Search Console and the same questions in AI answers. Between July and September 2026, the firm went from 23 to 36 of those searches on Google’s first page, and from 5 to 12 in first place. Its Google clicks went from about 44 a day to 72. The site was relaunched in July as well, so not all of that is down to the SEO work.

On the AI side, ChatGPT, Gemini, Perplexity and Claude named the firm in 55 of 104 answers to 26 buyer questions on 1 October, and Perplexity linked to its own site in 21 of its 26 answers. For pages published on 28 September, Google’s AI Overviews cited the firm on 4 of the 5 questions within three days. These are snapshots, each one a single run with web search on.

A pipeline for an agency

I also built an AI-visibility measurement pipeline for an agency. It runs 76 buyer prompts across ChatGPT, Gemini, Claude, Perplexity and Google AI Overviews, so 380 checks a run, and DataForSEO’s LLM Responses and SERP APIs capture the answers. They’re scored against a fixed rubric and checked claim by claim against a 39-item fact bank, and anything uncertain gets a human review.

One-off audits

Before taking someone on, I run the same checks as I would in a one-off audit. It shows them who Google’s AI and the other assistants name for their searches today, and which sources those answers cite.

What I learned

AI visibility moves around more than I expected. For another client, one check named them in 37 of 64 AI answers, and a week later, it was 24 to 29, mostly because one assistant stopped citing them. An audit can be out of date within weeks, so I check every week.

I’ve written a longer guide on what to look for when hiring help with this. Read it here.

Robin Laires, Founder, Nest Content