The re-brief tax is the time your team spends explaining a client to an AI tool before it can do useful work. Most chats start from zero, so the tax repeats on every task. You can measure yours with 1 formula: minutes per re-brief, times chats, times accounts, times people, times hourly cost.
AI Did Not Remove the Brief. It Multiplied It.
Before AI, you briefed a person once. They remembered. Next week, they still knew the client.
Now your team opens a new AI chat for each task. Most of those chats know nothing about the client. Even tools with memory or project folders usually hold that context per person and per tool.
So you did not get rid of the brief. You traded 1 brief per person for 1 brief per chat. Multiply that by every account and every person, and you have a tax nobody budgeted for.
What the Tax Looks Like in a Normal Week
It hides inside small moments. Each one feels too short to matter.
- Pasting brand guidelines into a chat. Again.
- Typing "our client sells to CFOs at mid-size banks" for the 40th time.
- Fixing a draft that broke a rule the AI never heard about.
- Pinging the account manager because the strategy is not written down anywhere.
- Rewriting output so it sounds like the client and not like a robot.
None of that is skilled work. It is work about work. Asana's Anatomy of Work research found that 60% of a person's time at work goes to work about work, not skilled work. It also counted 209 hours a year on duplicative work for the average knowledge worker.
Those are broad numbers across knowledge workers, not agency numbers. Do not borrow them. Measure your own.
The Formula
Here is the whole thing. 5 inputs and a calculator.
Monthly re-brief hours = (M × C × A × P × 4.33) ÷ 60
Monthly re-brief cost = Monthly re-brief hours × H
| Input | What it means | How to get it |
|---|---|---|
| M | Minutes spent re-briefing per chat, including fixing output that missed context | Time 20 chats with a phone timer and average them |
| C | AI chats per account per week | Have the team tally new chats on client work |
| A | Accounts per person | Count active accounts per account manager |
| P | People using AI on client work | Count them |
| H | Loaded hourly cost per person | Salary plus taxes and benefits, divided by hours worked |
4.33 is the average number of weeks in a month. Dividing by 60 turns minutes into hours.
A Worked Example (Made-Up Inputs)
This is an example, not data. The inputs below are assumptions picked to show the math. Your numbers will be different.
Assumptions:
- M = 6 minutes per chat
- C = 10 chats per account per week
- A = 6 accounts per person
- P = 5 people
- H = $60 per hour
Weekly minutes = 6 × 10 × 6 × 5 = 1,800 minutes. That is 30 hours a week.
Monthly hours = 30 × 4.33 = about 130 hours.
Monthly cost = 130 × $60 = $7,800.
Yearly, that is 30 hours × 52 weeks = 1,560 hours, or $93,600 at the same rate.
Again, every one of those inputs is made up. Change M from 6 to 3 and the cost drops by half. Change C from 10 to 20 and it doubles. That is why you measure.
The Bigger Cost Is Capacity
The dollar figure is the small problem. The big one is accounts you cannot take on.
Here is the second step. Divide your weekly re-brief hours by the hours 1 account needs each week.
Accounts lost to re-briefing = weekly re-brief hours ÷ weekly hours per account
Staying with the made-up example, say 1 account needs 5 hours a week. 30 hours ÷ 5 = 6 accounts' worth of time. That is a full book at the common 4 to 8 account ceiling, spent on explaining clients instead of serving them.
There is a quality cost too. A rushed brief gets a rushed draft. Off-brand work slips through. The client notices before you do.
How to Measure Yours in 5 Days
- Pick 2 people who use AI on client work every day.
- Give each a simple tally sheet. 1 line per new AI chat on client work.
- Time the setup on each chat. Stop the clock when the AI has enough context to do real work.
- Add rework time when a draft misses something the AI was never told.
- Average it out and plug it into the formula.
You will have a real number by Friday. Not a guess. Not a vendor's stat.
Then take it a step further. Our agency capacity page shows how to turn saved hours into accounts per person.
How to Stop Paying It
The tax exists because the client lives in people, not in a system the AI can read. Better prompts will not fix that. Neither will a bigger AI budget. It is not your team's fault. The tools were never built to carry client context across people.
Portable Delivery Intelligence fixes the storage. Each client's strategy, rules, priorities and approvals sit in 1 place. That place connects to the AI tools your agency already uses, like ChatGPT, Claude and Gemini. The context travels with the work.
Nobody re-briefs the AI. Nobody re-explains the client. Your team opens a chat and starts on the real work.
The install runs in 3 steps. Land: 1 service goes live in days, on real client work. Expand: every account moves on. Scale: your team or ours runs it, priced per account.
I built this because I lived the problem. I scaled a 25 person marketing agency the hard way, and the bigger it got, the less profitable it became. So I documented every single job. Then I rebuilt each one as AI.
Our internal mapping calls it "25 to 1." 1 operator plus the system carries the work a 25-person marketing team used to carry. The same delivery system has supported $7M in client revenue.
See the proof: 42 verified Upwork reviews, 5.0, Top 1% Expert-Vetted, with clients including Bowen eBikes and Kajeet.
With Portable Delivery Intelligence in place, our target is 18 to 25 accounts per operator, up from the common 4 to 8. Same team. No new hires.
FAQ
What is the re-brief tax?
It is the time your team spends giving an AI tool client context before it can do useful work. It repeats on every new chat, every account and every person.
Don't custom GPTs or project folders fix this?
Partly. They usually hold context for 1 tool and often 1 person. Agencies use several tools across a team, so the context needs to live in 1 place every tool can read.
How do I measure re-brief time?
Time the setup on 20 real chats, including rework from missed context. Average it, then plug it into the formula with your chat count, accounts, people and hourly cost.
Is the worked example in this post typical?
No. Every input in it is an assumption chosen to show the math. Your real cost could be much higher or much lower, which is why you should measure your own.
Want your real re-brief number? Bring one client account. We map where your team's time leaks and how many more accounts each person could carry. You keep the map either way. Book your mapping session here.
