AI Pulse · Business Models & Winners

Retainer pricing has become the standard way AI agencies charge clients. New data from Duet shows 78% of AI agencies now lead with retainers, up from 64% last year. The shift comes with a clear margin formula: charge three times your blended hourly cost to hold a 60 to 70 percent delivery margin.

The numbers explain why this matters. One agency, kipps.ai, ran a one-off $5,000 project that cost them $4,000 to deliver, a loss on day one. The same work sold as a $500 monthly subscription broke even by month 21 and turned profitable after that. The difference is not just revenue timing. Retainers let you staff predictably, refine the work over months, and avoid the constant pitch cycle that kills cash flow.

This tracks with what we see in the field. One-off AI projects look appealing because the price tag is big, but they hide cost blowout, scope creep, and the fact that most clients need ongoing tuning, not a handoff. Retainers align incentives. You stay close to the work, the client gets continuous improvement, and you can plan your team around stable revenue. The 3x cost rule is not arbitrary. It covers delivery, account management, platform costs, and the buffer you need when a model changes or a client asks for a tweak.

If you are running AI work in-house, the same logic applies. Treating AI capability as a recurring operational layer, not a series of projects, is how you build something that compounds. That is the thinking behind tools like the Omni Command Centre, where the model is ongoing orchestration, not one-time deployment. Retainers work because AI work is never really finished. The agencies that figured this out early are the ones still operating profitably today.