AI for independent PR agencies. Where the hours actually go, and what to build first.

Where the time goes in a PR agency

Ask a PR team what they spend their week on and they describe the delivery: releases, briefing documents, pitches, FAQs. Ask what they would hand off tomorrow and a different list comes out. That second list is where the audit usually lands.

• Media monitoring summaries. Reading the overnight coverage, deciding what matters, classifying it and writing it up for the client. Daily or weekly, every client, often by the most senior person on the account because they know what the client cares about.
• Coverage logging. Moving the same coverage into a spreadsheet with outlet, date, reach, sentiment and message pull-through. Pure data entry, done by hand because nothing connects the monitoring feed to the sheet.
• Status documents. The weekly document that exists so the client call can happen. Assembled from the team's heads every week, and mostly the same shape as last week's.
• Monthly reporting. The evaluation report that re-cuts the coverage log into charts and commentary. Agencies rarely bill for it separately, so every hour is margin.
• Pitch boilerplate. Credentials and case studies reused across decks. Real, but smaller than people assume: in the agency we measured it was well under ten per cent of pitch time, and the rest of a pitch is bespoke thinking that should stay that way.

Notice what is not on the list. The brief, the strategy, the idea, the relationship with the journalist and the relationship with the client. Those are the agency. We are not interested in compressing them, and your clients would notice if we tried.

What the first build usually looks like

The pattern that works in PR is draft and verify. The system reads the monitoring feed, classifies the coverage, drafts the client summary and fills the log. A named person checks a fixed list before anything leaves the building: the sentiment calls, the figures, the client's own messages, anything about a spokesperson. Then it goes.

Two things make this the right place to start. It is where the recurring hours are. And it is already how a good agency AI policy describes acceptable use: AI may classify and summarise, a human verifies named items. You are not inventing a new way of working, you are building the one your policy already permits.

The question your clients are about to ask

Your clients read the same headlines you do. The first version of the question is "where are the AI savings?" The second, which arrives through procurement, is "show us your AI policy, your AI transparency statement and the evidence that staff have read them." Public-sector and regulated clients tend to ask it in writing already. Everyone else is a re-pitch away.

The agencies with an answer keep the margin and the account. The answer is four documents: an internal AI usage policy, a client-facing AI transparency statement, a register of approved tools, and signed acknowledgements from the team. The audit produces the first two and tells you how far you are from the other two. What the policy needs to cover.

What stays human

The value of your agency is human. The brief, the judgement and the client relationship stay with your people, at the front of every piece of work and at the back. We build the systems that compress everything in between: the rework, the reporting, the retrieval, the admin. That is the whole proposition, and it is also the reassurance your team needs to hear before any of this starts.

Questions PR founders ask