HomeTasksMarketing AgencyContent drafting
Automating content drafting
Content drafting in Marketing Agency. Capability-gated, 65% repeatability, Bearable-tier risk. Here is what the framework says about the shape of this work, and what supervision it would need if you ran it.
What the shape of this task tells you
With the shape of this task alone, the readiness score lands between 30 and 55 out of 100. Where it falls inside that range depends entirely on your volume: the economics half of the score is worth 45 points and it is earned at the scale you actually run. Enter your numbers to fix the score.
Repeatability 65%. A substantial share of instances deviate from the standard path. That is the part that decides this verdict: the exceptions are the work, and they are what a pilot measured on routine volume will not show you.
Judgment medium, risk Bearable. Together these set both the readiness penalty and the supervision this work needs. Errors here are cheap and quick to fix, so cost can reasonably lead the decision.
The failure mode to test for
Claims or statistics presented as fact with no traceable source
This is the one to build your sample around. Errors that survive review are the ones that look correct, so a review pass that only inspects items flagged as unusual will systematically miss it.
The oversight contract
Delegation level 3 of 4 — Act and sample. Scoped write, own account, audit log, tested rollback.
| Period | Sample | Step down when |
|---|---|---|
| Weeks 1–2 | 14% | Error rate under 5% for the full period |
| Weeks 3–6 | 7% | Error rate under 5% for the full period |
| Week 7 onward | 3% | Steady state — do not reduce further |
Step down on measured performance, never on the calendar date. If the error rate breaches the ceiling at any point, go back up a step.
Sample adequacy. Seeing no faults in a sample of n only rules out error rates above roughly 3 ÷ n. To have evidence for a 5% ceiling you need about 60 reviewed units a month. At a 3% steady-state rate that means a monthly output of roughly 2,193 units before a clean month means anything. Below that, treat a clean month as an operational check rather than proof.
Capability-gated
This verdict rests on what models can currently do reliably. It is re-scored after every frontier release and it is expected to move.
Get the full verdict for your numbers
The score above is the shape of the work. Enter headcount, salary, hours and output and the audit returns the readiness score, three costed scenarios and a supervision schedule sized to your volume.
Audit this taskCommon questions
Should content drafting be automated?
It depends on your volume as well as the shape of the work. On shape alone this task scores between 30 and 55 out of 100, with 65% repeatability, medium judgment and Bearable-tier risk. The remaining 45 points of the score come from your cost per unit and whether your manual spend covers the fixed costs, so a low-volume version of this task can score well below a high-volume one.
How much output would need to be reviewed?
At steady state, about 3% of output, after a ramp starting at 14% for the first fortnight. The error ceiling is 5% — above that, automation pauses.
What is the most likely way this goes wrong?
Claims or statistics presented as fact with no traceable source.
Will a better model change this verdict?
This verdict rests on what models can currently do reliably. It is re-scored after every frontier release and it is expected to move.