Agency retainer burn: see it before the invoice, not after

The retainer is the agency’s favourite contract: predictable revenue, a standing relationship, no re-selling every month. It is also where margin disappears quietly. A client on forty hours consumes fifty-five because the account team likes them, nobody notices until finance reconciles timesheets on the twenty-eighth, and by then the month is unbillable. This post is about noticing on the twelfth.

Why retainers run over without anyone deciding to

Over-servicing is rarely a decision. It is a hundred small ones: a quick extra round of amends, a call that ran long, a designer helping out because the client was nice about the last delay. Each is reasonable. None is visible, because the hours are written down later — often on Friday, from memory — and the total only exists when someone adds the month up.

The structural cause is that the retainer limit lives in a contract, the hours live in a timesheet, and the two meet once a month in a spreadsheet. If the limit and the hours lived in the same place, the comparison would be free and continuous.

The three numbers you need mid-month

1. Hours used against the period allowance. Not the month’s hours in total — the hours logged to this client in this retainer period, against the allowance the contract sets.
2. Cost, not just hours. Ten senior hours and ten junior hours are not the same burn. Costing each hour at the person’s real rate turns the gauge into money, which is what margin is measured in.
3. Projected end-of-period position. At the twelfth with 60% of hours used and 40% of the period gone, you are heading over. That projection is the early warning; the invoice is the post-mortem.

Making hours arrive by themselves

Everything above depends on hours being recorded as they happen and attached to a client. The friction of a separate time-tracking form is exactly why Friday reconstruction happens. The fix is to make the attendance clock the time tracker: when someone clocks in or switches task, they pick the client project, and the entry is done. Inactivity detection pauses the clock when the desk goes quiet, so the record stays honest without anyone babysitting a timer. We compare the three kinds of time record in attendance, time tracking or timesheets.

In Orbiz that clock is AMS. Members clock in and log time against the client project; the entries flow into Trackline, where each client is a project with retainer periods. Trackline costs every hour at the member’s rate from their HR record — costs are never stored, they are computed at read time, so a rate change is reflected everywhere at once — and shows hours and margin for the current period against the allowance.

What to do when the gauge turns amber

• Re-scope, do not absorb. Move the extra round of amends into next period, or agree a one-off addition. Both are conversations the client accepts on the twelfth and resents on the invoice.
• Check the mix. If senior hours are eating the allowance, the plan needs a junior on the routine work. This is a staffing decision you can only make mid-month.
• Look at estimation, not just burn. If the same client runs over every period, the retainer is mis-sized. Trackline’s estimation accuracy view shows whether the pattern is one client or the whole book.

Fixed-price projects have the same disease

Everything here applies to fixed-fee work with milestones instead of periods. The gauge becomes hours and cost against the milestone estimate, and the amber moment is when a milestone has consumed its effort before its scope is done. Consultancies live in that world as much as agencies do — the professional services page walks through the engagement view, and the agency page covers the retainer set-up with client logins and files alongside.

A one-week set-up

1. Create each retainer client as a project with its period length and hour allowance.
2. Make sure every account team member has a rate on their HR record — this is what turns hours into margin.
3. Ask the team to pick the client project when they clock in. Do not introduce a separate timesheet.
4. Give account managers the portfolio dashboard and agree an amber threshold, say 70% of hours at 50% of the period.
5. Review the first period’s gauges together. The first surprise is usually a client everyone assumed was profitable.