What is the approval rate?
The approval rate is the share of a Digital FTE's proposals in a period that a person approved without changing them. Proposals that were edited before approval and proposals that were sent back count against it, each visibly. It is computed per class of action, such as exact-match reconciliations, first-touch outreach, or shipping-status replies, because a role's classes differ in difficulty and blending them hides the one that matters.
It is one of the three universal measures of a Digital FTE, with outcomes against targets and cost per outcome, and it is the one that says whether the manager agrees with the work.
How does the approval rate work on ftes.ai?
On ftes.ai every proposal's outcome at the approval gate, approved, edited, or sent back with a reason, is logged with the approver's name. The Performance tab computes the rate per class and shows the send-back reasons grouped beside it. Auto-approved actions in a promoted class are counted separately, because they were not reviewed, and any send-back on one of them is flagged because it triggers a step-back.
The monthly review reads the rate with the reasons: a rising rate means the remit and memory are converging on what the team wants; a falling rate in one category means something changed and the reasons say what.
Why does the approval rate matter?
It matters because it is the honest quality measure and the honest cost measure at once. A proposal approved unchanged was right and cost the minutes of reading it; one edited or sent back was not quite right and cost more human time. So a class's rate tells the manager both whether the work meets the standard and how much of their own time it consumes.
It also matters because it is the evidence the autonomy ladder runs on. A class promoted to auto-approve should have a long run of unchanged approvals behind it, reviewed by the manager, and the rate is where that run is visible.
How should it be read?
Read it per class and with its reasons. A class at a high rate for a long run is a candidate for promotion. A class at a low rate with one repeated reason is a remit gap: fix the rule and watch the rate rise. A class at a low rate with scattered reasons may be a class that genuinely needs a person each time, which is fine. A class whose rate falls after being stable is drift: a changed source, a changed input, or a changed expectation.
Never read one blended number for a role. A role with an excellent rate on routine matches and a poor one on adjustments is two stories, and the blend tells neither.
- High and stable, long run: promote the class
- Low with one repeated reason: fix the remit
- Low with scattered reasons: keep gated, that is fine
- Was stable, now falling: drift, find the change
What is it often confused with?
It is confused with accuracy, which nobody can measure directly; approval rate measures agreement with the manager's standard, which is the standard that matters. It is confused with throughput; a role can approve many proposals at a poor rate. And it is confused with a target; it is a diagnostic, and a manager chasing a high rate by approving carelessly has broken the instrument.
It is sometimes read without the reasons. The rate without the reasons is a number without a meaning.
Where do you see it in the workspace?
In the Performance tab per class of action with send-back reasons grouped beside it, in the monthly review the FTE writes, and in the autonomy ladder's promotion view in Setup, where a class's run is shown when a manager considers promoting it.
Related terms: send-back, autonomy ladder, cost per outcome, targets, monthly review.
In the first month, expect the rate to be modest and to rise, and read it weekly with the reasons rather than monthly; the early send-backs are where the remit gets written properly. A rate that is high from the first week is worth a second look, because it usually means the manager is approving faster than they are reading, and the record it builds will not support the promotion decisions that later depend on it.