Most agencies don’t lose margin on bad clients or weak talent. They lose it in the unbilled space between tools — the contact record someone re-types into the project tracker, the “quick change” that never gets logged, the finished design that sits for two days waiting on a handoff nobody triggered. Below is the audit we use to find those leaks, put a dollar figure on each one, and plug them — without hiring, and without raising rates.

Why the leaks hide in plain sight

Every tool in your stack probably looks fine on its own. The CRM does CRM things, the project tool tracks tasks, the invoicing app sends invoices. The problem is that none of them account for what happens in between: a human copying, chasing, reconciling, and re-explaining. That work never appears as a line on the P&L. It shows up as a feeling — “we’re busy all the time but the margin isn’t there” — which is exactly why owners tend to look in the wrong places first. They audit rates, utilization, and salaries, and skip operations entirely. In our experience, the money is usually leaking through four specific gaps.

The four leaks, and how to spot each one

1. Manual data re-entry

Follow one new client from signed proposal to first deliverable and count how many times a human re-types the same information. Across the audits we’ve run, four to six times is typical: into the project tool, the invoicing system, the client portal, the shared drive structure, maybe a reporting sheet. Each entry is only ten or fifteen minutes, which is why nobody flags it. But at five new projects a month across a team, you’re looking at 40–60 hours a year of pure transcription — before you count the cost of the inevitable typo that sends an invoice to the wrong contact.

The tell: ask anyone on your team “what do you re-type every week?” If they answer instantly, you have this leak.

2. Untracked scope creep

Not the dramatic scope battles — those get escalated and negotiated. The leak is the small stuff: the “can you also tweak the header” Slack message, the extra revision round agreed on a call, the export a client requested by email. Each one gets done, because your team is helpful, and none of it lands on the project record.

Run the math on one retainer. If a $6,000/month client absorbs three unlogged hours at a $120 blended rate, that’s $360 — six points of margin on that account, gone silently. Across ten retainers, it’s often the difference between the margin you planned and the margin you got.

3. Idle handoff time

We’ve written separately about the handoff between closed-won and kickoff; this leak is about what happens after kickoff, once the project is already in delivery. It doesn’t cost labor hours; it costs capacity. Map a recent 30-day project and mark every period where the work was simply waiting: for a brief to be approved, for a revision round to come back, for the person who “owns” the next step to notice it’s their turn. In the projects we’ve mapped, four to six idle days out of 30 is common, and it means the same team ships fewer projects per quarter than it could. That’s revenue you were staffed for and never collected.

4. Duplicate admin work

Status reports assembled by hand from three tools. Timesheet chasing. Screenshotting dashboards into a client deck. If a project manager spends Friday afternoons building updates from data that already exists somewhere, you’re paying senior rates for copy-paste. We regularly see PMs losing four to six hours a week to this — half a day of billable capacity per PM, every week. It’s almost always the first place we point an owner who wants a quick win.

The audit: one week, no consultants

You can locate all four leaks in about a week without disrupting delivery.

Pick three recent projects — one that went great, one average, one painful. Three is enough to see patterns; one is not.

Map every system touch. Walk each project chronologically and write down every moment information moved between tools via a human: re-entered, exported, pasted, or verbally relayed. Don’t judge yet, just list.

Interview, don’t survey. Surveys get polite answers. A ten-minute conversation per person with two questions — “what do you re-type every week?” and “where does work sit waiting on you or someone else?” — gets the truth. Your team already knows where the leaks are; nobody has asked.

Log the quick asks. For five working days, have everyone drop client requests that arrived outside the project record into one channel. The volume usually surprises owners more than anything else in the audit.

Putting a number on each leak

Turn hours into margin, or the audit stays an interesting document instead of a decision.

For re-entry, scope creep, and duplicate admin: hours per month × blended hourly cost, divided by monthly revenue. That gives you margin points per leak. A 10-person agency billing $100k a month that finds 90 leaked hours at a $95 blended cost is leaking roughly 8.5 points — on work that was already sold.

Idle time works differently, because it’s capacity rather than hours: estimate how many extra projects per quarter the team could ship if handoff gaps were cut in half, and value them at your average project price. Take that same agency: if it runs ten projects a quarter with five idle days each, halving those gaps recovers roughly 25 project-days — close to a full extra project slot. At a $30,000 average project, that’s $30,000 a quarter, or about 10 points of monthly revenue. It’s a rougher number than the hourly leaks, but even a conservative version of it is usually the largest of the four.

Your numbers will differ. The point isn’t precision — it’s ranking the leaks so you fix the biggest one first.

Plug the leaks: consolidate before you automate

The instinct is to buy a tool per leak: a form builder for intake, an integration layer to sync systems, another dashboard for reporting. We’ve watched agencies do this and end up with more gaps than they started with, because every new tool creates two new seams — plus a stack of subscriptions and a sync layer someone now has to babysit. The goal isn’t to work faster inside each tool; it’s to remove the seams between them.

The durable fix is fewer seams, not more patches. When CRM, project delivery, client communication, and billing share a single client record, most of the leaks close structurally: there’s nothing to re-type, scope requests land on the project because that’s where the client already is, and handoffs trigger automatically instead of waiting for someone to notice. That’s the operating model we built OpenAva around — one hub where the client record, the work, and the money reference the same source of truth, with automation layered on top of a shared record rather than duct-taped between disconnected apps.

Two honest caveats. First, consolidation won’t rescue a pricing problem or chronic under-utilization — it recovers margin you’ve already earned, not margin you never charged for. Second, some tools should stay separate: your accounting ledger and your creative software are fine where they are. Consolidate the operational core — the path a client and their money travel — and integrate the rest.

Mistakes that undo the gains

Migrating everything at once. Move one workflow — usually client onboarding, since it touches every system — prove the leak closed, then expand. Big-bang migrations stall and breed tool-hoarding.

Keeping the old tools “just in case.” If both systems stay live, people update both, and you’ve reintroduced duplicate entry with extra steps. Set a shutdown date.

Automating a broken process. If your scope-change process is “whoever the client messaged deals with it,” automating it just makes the chaos faster. Fix the process on paper first; automate second.

Where to start this week

Pick one recent project and run the mapping exercise on it — an hour, maybe two. Count the re-entries, the unlogged asks, and the idle days, and multiply them out. If the number is small, your operations are tighter than most. If it isn’t, you now know exactly which leak is costing you the most — and which one to plug first.