How to Have an Honest Conversation With Your Marketing Agency About AI

Most business owners do not need a dramatic break-up with their marketing agency because of AI. They need a better conversation about what they are paying for, what results matter and which parts of the work can now be done differently. If you are wondering how AI and your marketing agency should fit together, begin with the work itself rather than a demand for a cheaper retainer. An agency may be using AI to speed up research, first drafts or reporting. That does not make its strategic judgement, specialist skills or accountability worthless. Equally, faster production should not remain a mystery to the client. ## Start with the business outcome Bring three things to the meeting: the enquiries or sales you want, the customers you most want to reach and examples of work that has or has not helped. Ask which activities contribute to those outcomes and which exist because they have always been in the monthly plan. A sensible opening question is: “Where can AI remove repetitive work without weakening the quality of our marketing?” It invites a useful answer. “How much can you cut your fee now you have ChatGPT?” usually does not. ## Agree the division of work Separate direction, production and approval. Your business knows its customers, offers and constraints. The agency may bring channel expertise, execution capacity and outside perspective. AI can help with research organisation, variations and repetitive preparation. Someone still has to verify claims, check tone, protect confidential information and decide what deserves to be published. Ask for a simple responsibility map: who sets the brief, who uses AI and for what, who checks the output, who signs off, and what happens when the work misses the mark. If your internal team can take on a defined part of production, discuss how the agency's scope and fee should change. If you need someone to decide the direction before anyone produces more material, a consultant can help with that planning. ## Measure value, not content volume A calendar full of AI-assisted posts is not automatically progress. Agree a few business measures: qualified enquiries, useful sales conversations, conversion quality and what you are learning from customers. Review the work against those measures and stop what is not helping. The aim is a healthier partnership, not an agency-versus-AI contest. Our [AI marketing support](/ai-marketing/) starts with the commercial problem and works out which people, processes and tools should do the job. ## FAQs ### Should my marketing agency disclose when it uses AI? Yes, ask for a clear policy covering which tasks use AI, what data goes into tools and how people review outputs. You need accountability rather than a list of every keystroke. ### Should AI make my agency cheaper? It may reduce the time spent on some production tasks. Discuss the scope, quality and outcomes before assuming the whole fee should fall; strategy, expertise and responsibility still have value. ### What if my agency refuses to discuss AI? Ask for its approach in writing and propose a review of the work and results. A refusal to explain quality controls or scope is a more useful warning sign than whether it uses a particular tool. ### When should I bring in an independent consultant? When the underlying problem is unclear direction, positioning, priorities or a difficult division of work. A consultant can define the plan without automatically replacing the agency.

This article was written by Peter Lowe. The ideas and opinions are his own; AI was used to assist with drafting and editing.