Client reporting is quietly eating your marketing agency's margin

The reporting and pitch treadmill nobody priced in

Two things are squeezing marketing agency retainers at the same time. Client-side procurement teams keep running formal pitch and fee-review processes even for relationships that have been in place for years, so retainers get re-priced more often than they used to. And the Advertising Standards Authority's guidance on AI-generated advertising content is a reminder that using a tool changes nothing about who is accountable. The CAP and BCAP Codes have no AI-specific rules - the same rules that apply to any other ad apply here, and the agency and advertiser remain responsible for every claim, however it was produced. Neither problem gets solved by working faster. Both point to the same fix: knowing exactly where the hours in a retainer month go, so the ones the client is actually paying for get protected and the ones that aren't get automated.

Most agency owners can name the symptom without naming the cause: a retainer that used to run at a healthy margin now runs thin, the account team that used to have slack for new business is permanently behind on it, and account handlers spend more of the week chasing timesheets and approvals than actually managing the account. None of that is the creative or the strategy the client signed up for, but it all still has to happen, every month, on every account.

The usual response is to absorb it - longer hours, thinner margins, a founder who does the pitch decks at the weekend. That gets a client roster through one busy quarter. It does not fix the fact that the retainer was priced on the work, not on everything that surrounds it.

Where the hours actually go

Ask any agency owner to walk through a typical week and the same four things show up, in some order, on almost every account.

• Client reporting. Pulling numbers from GA4, Meta Ads Manager, Google Ads and whichever CRM holds the pipeline, then writing the commentary that makes the numbers mean something to a client who is not going to read a raw dashboard. On a multi-channel account this typically runs to half a day a month, sometimes closer to a full day when the client wants a narrative deck rather than an export.
• Scope creep. The extra revision, the out-of-hours request, the "quick addition" to a brief that never gets logged against the scope of work. It is rarely billed, because nobody wrote it down at the time, and it is usually where a meaningful chunk of the month's margin quietly disappears.
• Timesheets and the project-management board. Chasing the team to log time properly, keeping Asana, Monday or ClickUp current, moving cards so the board actually reflects where a project stands rather than where it stood last Tuesday. An hour or so per account handler in a typical week, considerably more once a project falls behind.
• Pitches and proposals. Every new-business opportunity needs a tailored deck - approach, deliverables, timeline, price - built from a brief and a template, and most of that work never converts. A full working day is typical for a mid-sized pitch, layered on top of the retainer work still owed to existing clients that week.

These are typical ranges rather than a fixed rule - a content-led shop loses time differently to a paid-media or full-service agency, and the mix shifts with account size and channel count. What is consistent across very differently shaped agencies is that these four things, not the creative or the strategic thinking, are what quietly eats a retainer. In the agencies we've assessed, they routinely add up to a third or more of an account handler's working week - not a headline figure we'd put on your business specifically, but a pattern worth checking against your own.

What is safe to hand to AI, and what still needs a person

The instinct, once a task is flagged as repetitive, is to automate all of it. That is the wrong move for at least two parts of an agency's work, and getting the line right matters more here than in most sectors, because the output goes out under the agency's name and the client's brand.

Safe to automate: the first draft of a reporting narrative, built from a templated data pull rather than a blank page every month. The structural skeleton of a pitch deck - sections, layout, boilerplate case studies - drafted from a brief so a strategist edits rather than starts cold. Keeping a project-management board and timesheet record current from calendar and task data, so the chasing stops being a manual weekly job.

Not safe to hand over without a person checking it first: any AI-generated or AI-manipulated imagery or copy going into a live ad without someone confirming it meets the same CAP Code standards as any other ad before it runs - responsibility for every claim sits with the agency and the advertiser, not with the tool that produced the draft, and disclosure that AI was used is only expected where not disclosing it would mislead the audience. Client data pulled from an ad account or a CRM and fed into a consumer-grade AI tool without a proper business-tier data agreement - that is a UK GDPR question, and it needs answering before the workflow gets built, not after a client asks. And the final pricing or scope decision on a proposal, which should stay with a person who understands the account, however good the drafted structure underneath it is.

Reporting-dashboard tools built specifically for agencies - AgencyAnalytics and Databox are two commonly used examples - already pull the raw platform data automatically. The time an agency actually loses is almost always in the commentary layer sitting on top of that data, not the pull itself, which is exactly the part that drafts well from a template and a clear brief on how a particular client likes to be told things.

One task, before and after

Take the monthly client reporting deck for a mid-sized paid-media account. Before: an account handler exports data from GA4, Meta Ads Manager and Google Ads separately, reconciles it against the CRM's pipeline figures, and writes the narrative from scratch - typically half a day, more if the client comes back with follow-up questions that need a second pass.

After: the dashboard tool the agency already pays for pulls the same data automatically, and an AI drafting step produces a first-pass narrative from that pull, against a template built around how this specific client likes to be told things. The account handler reads it, corrects anything the data doesn't fully explain, and adds the one or two lines of strategic judgement the client is actually paying for. In agencies that have set this up properly, that half-day task typically comes down to under two hours - the account handler is still doing the part that needs a person, just none of the copy-paste sitting around it.

That is a description of the fix, not a promise about your account specifically. Whether this task looks like a half-day drain or something smaller in your agency depends on how many clients you report to, how many channels each one spans, and whether the tools you already pay for are actually pointed at the job.

Find your own number

Every agency runs a different mix of retainers, projects and channels, so the honest answer to "how many hours could we get back" is never a sector average. It is whatever your account handlers, delivery team and founder are actually doing this week, measured properly rather than guessed at.

That is what the AI workflow assessment for marketing agencies is built to find. A 60-minute session maps where your hours genuinely go across reporting, scope, timesheets and pitches, then hands back a report built around your agency's own tools and workflow - not a generic list - with a prioritised plan for what to fix first. It is guaranteed too: if the report doesn't show at least 5 hours a week worth reclaiming, the fee comes back.

If you want a rough sense of where the biggest number in your agency is likely sitting before you book anything, the admin cost calculator gives you a starting estimate in a couple of minutes. And if part of your admin load sits inside a CRM or PM tool that half the team avoids logging into properly, the piece on the CRM and workflow gaps that cost UK service businesses is worth reading alongside this one - it covers the drop-out points that tend to sit right next to the reporting and pitch problems above.

For a faster, free starting point, the AI Readiness Quiz takes about two minutes and flags where the friction in your agency is highest right now.

Sources: the roughly one-third-of-a-week figure for reporting, scope creep, timesheets and pitch admin combined is drawn from patterns observed across HoursBack's own assessments of UK service businesses, not a published industry study; individual agencies will vary. Task-level hour ranges (reporting, timesheets, pitches) are typical ranges, not fixed benchmarks. References to the ASA's guidance on AI-generated and AI-assisted advertising content and to UK GDPR obligations on client data are general regulatory context, not legal advice - agencies should check current CAP Code guidance and their own data processing agreements directly.