Key findings
66% of agencies report increased client demand for AI search services. AgencyAnalytics, n=494.
Enterprises put an average of 12% of digital budgets into answer engine optimization in 2025. Conductor.
66% of marketers have used an AI search tool to evaluate a marketing vendor or agency. Fractl, via PPC Land.
How many clients are asking agencies about AI search?
Most of them. AgencyAnalytics surveyed 494 marketing agency professionals between February and April 2026. 66% reported increased client demand for AI search services, a category the report notes did not exist a year earlier.
The budgets are moving on the client side too. Conductor, which sells a platform in this market, surveyed more than 250 enterprise digital leaders and found they put an average of 12% of digital budgets into answer engine optimization in 2025, with 94% planning to increase it. GNW Consulting and Demand Metric found that 92% of B2B organisations were either experimenting with generative engine optimization or running it, from 225 responses in early 2026. Gartner’s 2026 CMO Spend Survey found CMOs allocating 15.3% of marketing budgets to AI initiatives while total budgets stayed flat.
What are clients actually asking for?
Usually one thing: why a competitor is named in ChatGPT and they are not. That is a measurement question before it is a content question. The client needs to see which questions were asked, on which engine, which businesses were named, and which pages the engine read to decide.
Clients also use these tools to choose agencies. A Fractl survey released in July 2026, as reported by PPC Land, found that 66% of marketers had used an AI search tool to research or evaluate a marketing vendor or agency. Your own agency is being described by the same engines.
What does the work involve?
A fixed set of questions. Written in the customer’s words, by stage of the customer journey, and run daily on each engine. Change the questions and you cannot tell whether the answer moved.
A source record for every answer. Which page, listing or article each claim came from.
Corrections first. Most wrong answers trace to an outdated third-party page. Fixing those is faster than publishing.
Publishing against gaps. Pages for the questions where no business is named yet.
Earned sources. The directories, review sites and publications the engines already retrieve for the category.
Reporting by engine. A blended score hides where the client is losing.
It touches SEO, content and PR, which is why agencies are well placed to sell it and why it is hard to staff as a side project.
How to offer it, and sources
There are three routes: build a practice, buy tools and train the team, or partner with a specialist. We work with agencies two ways. You can refer a client and we run the engagement, or you can white-label the service and deliver it under your name. In both, our team does the daily reading, the source work and the reporting. How SourceNode partners with agencies. The engagement we document in full is on our results page.
The researcher
Odin Lund is co-founder and COO of SourceNode and leads its data analytics and research, including the daily tracking behind this work across ChatGPT, Gemini and Google AI Overviews. Odin founded First Query, the AI search practice that became SourceNode, and co-founded The Pixel Paper.
Meet the research team
About this research
Published by SourceNode
SourceNode is a data and research-first AI search firm. We record what AI models say about a business every day, trace each claim to its source, and use that data to get businesses named in AI answers. Every figure we publish carries its method and its sample.
Share Blog
Related research
AI search
AI search statistics for 2026: the numbers we trust, with sources
Read the research
AI search
What McKinsey’s AI search research says, and what to do about it
Read the research
