The retainer that looked profitable until you mapped the handoffs

I have sat with agency owners who could tell me their retainer fee to the penny and their deliverable count by heart. Ask them how many times a single piece of content changes hands between people or systems before it leaves the building, and the room goes quiet. Nobody had counted. Nobody thought to.

That is usually where the money went.

What retainer drift actually means

Retainer profitability does not collapse in one meeting. There is no invoice where you can point and say: that is where it went wrong. It goes wrong across thirty small moments across twelve weeks, none of which individually triggers an alarm.

A client asks for one more round of amends on a social carousel. Reasonable. A strategist jumps into a copy doc to fix a brief gap that was never formally scoped. Helpful. The monthly report takes three hours instead of one because the performance data lives in Google Analytics, the paid data lives in Meta Ads Manager, the SEO data is in a Notion doc someone built six months ago, and the client wants it all in a single slide deck. Manageable, once.

The problem is that each of these moments is a handoff. Work changes hands between a person and a system, or between two people, or between a tool and a template. Each handoff carries a cost: context-switching time, re-reading time, the small decisions about what version is current. Individually invisible. Collectively, they are why a retainer that looked like 60% margin in month one is running at 30% by month four, and nobody can explain where the hours went.

The deliverable count stayed the same. The handoff count doubled.

Where handoff debt hides in your tooling

ClickUp and Notion are the two places I see this most clearly, because they are flexible enough to become whatever a team needs them to be. That flexibility is the problem.

In ClickUp, a typical content retainer might have a task for each deliverable. Looks clean. But inside that task: a description written by the strategist, a comment thread where the copywriter asked three clarifying questions, a linked Google Doc where the actual draft lives, a second comment thread on the doc itself, a Loom from the account manager explaining the client's verbal feedback, and a final status update in the task that may or may not reflect what actually happened. That is five or six context switches for anyone picking up the work mid-flight. A new team member does not know where to start. A senior person covering for someone on leave spends twenty minutes just orienting.

Notion compounds this differently. It is excellent for documentation and genuinely poor for work-in-progress tracking, but agencies use it for both because it is already open in a tab. A content brief lives in a Notion database. The draft lives in Google Docs. Feedback comes in over email because the client does not have Notion access. Someone pastes the feedback into a comment on the Notion page. Someone else acts on the email version. Now there are two feedback records and one of them is wrong.

The tool is not the cause. The process is. But the tool is where the evidence accumulates, if you know how to read it.

The handoff map: how to run it in under an hour

Pick one deliverable from your current highest-value retainer. A blog post, a paid social set, a monthly report. Something that completed in the last four weeks.

Trace it backwards from delivery to brief. Write down every moment the work changed hands: person to person, person to tool, tool to template, draft to feedback to revision. Do not filter for significance. Write them all down.

handoff_audit:
  deliverable: "Monthly performance report - April 2026"
  total_handoffs: 11
  chain:
    - step: 1
      from: "GA4 dashboard"
      to: "analyst spreadsheet"
      method: "manual export"
      time_cost_mins: 20
    - step: 2
      from: "Meta Ads Manager"
      to: "same spreadsheet"
      method: "manual export"
      time_cost_mins: 15
    - step: 3
      from: "spreadsheet"
      to: "Notion summary doc"
      method: "copy-paste"
      time_cost_mins: 25
    - step: 4
      from: "Notion doc"
      to: "account manager"
      method: "Slack message with link"
      time_cost_mins: 5
    - step: 5
      from: "account manager"
      to: "slide deck template"
      method: "manual reformat"
      time_cost_mins: 40
    - step: 6
      from: "slide deck"
      to: "client"
      method: "email attachment"
      time_cost_mins: 10
    - step: 7
      from: "client"
      to: "account manager"
      method: "email with tracked changes"
      time_cost_mins: 5
    - step: 8
      from: "account manager"
      to: "analyst"
      method: "Slack summary of email"
      time_cost_mins: 10
    - step: 9
      from: "analyst"
      to: "revised slide deck"
      method: "manual edits"
      time_cost_mins: 30
    - step: 10
      from: "revised deck"
      to: "account manager review"
      method: "Slack link"
      time_cost_mins: 15
    - step: 11
      from: "account manager"
      to: "client"
      method: "email"
      time_cost_mins: 5
  total_time_mins: 180
  scoped_time_mins: 60
  untracked_overrun_mins: 120

When agencies run this exercise honestly, the number of handoffs on a single deliverable is almost always between eight and fourteen. The scoped time assumed four or five. The gap is where the margin went.

Now look at which handoffs are structural (they happen every single time, for every client) versus incidental (they happened because of a gap in the brief, a missing integration, or a tool being used outside its intended purpose). Structural handoffs can be engineered out. Incidental ones are symptoms of a process problem that will recur.

Where scope creep is structurally hiding

Scope creep is usually described as a client behaviour problem. Clients ask for more than they contracted. That happens. But the more expensive version is internal scope creep: work that expands not because the client asked but because the process requires it.

The strategist who covers a brief gap is not doing anything wrong. They are being professional. But if the brief template does not capture that information in the first place, the gap will appear on every retainer, every month, and someone will always cover it informally. That informal coverage never appears on a timesheet category that triggers a review. It is just absorbed.

The same logic applies to reporting. If the data infrastructure requires four manual exports to produce one slide, the report will always take three hours. Scoping it at one hour is not optimism. It is a recurring loss built into the contract.

This connects directly to the brief problem I wrote about in the brief bottleneck piece: when intake is weak, every downstream handoff carries the cost of the missing information. Fix the brief, and three handoffs downstream disappear automatically.

The question to ask for each structural handoff is: what would have to be true for this step not to exist? Sometimes the answer is a better integration. Sometimes it is a conversation with the client about data access that nobody has had because it felt awkward.

Where to start

One retainer. One deliverable. This week.

1. Choose the retainer you feel least confident about on margin. Not the one you think is fine.
2. Pick one deliverable that completed in the last thirty days.
3. Map the handoff chain from brief to delivery. Every step. No filtering.
4. Count the handoffs. Write the time cost against each one.
5. Separate structural from incidental.
6. Identify the one structural handoff with the highest time cost. Ask what would have to be true for it not to exist.
7. Do not try to fix everything. Fix that one handoff before the next delivery cycle.

The goal is not a perfect workflow by next quarter. The goal is a number: handoffs per deliverable, tracked across retainers, reviewed monthly. Once you have that number, you can see drift happening in real time instead of discovering it when the margin report lands.

Retainer health is not a client management problem. It is a process visibility problem. The clients are not the variable. The handoff count is.

If you have run this exercise and the structural handoffs are deeper than a single integration fix, the AI Workflow Audit is where I work through this with agency teams directly: mapping the full handoff chain, identifying where automation can remove steps rather than just accelerate them, and building the tracking that makes drift visible before it becomes a write-off.

For a related view on where invisible process debt accumulates in AI-assisted work, the piece on why most AI pilots fail covers the same pattern from a different angle. The tool gets blamed. The workflow design is the actual problem.