What is cost per outcome?
Cost per outcome is what one result from a Digital FTE cost: the tokens the role consumed over a period, converted at the plan's rate, divided by the outcomes it delivered over the same period. A Finance Controller FTE's outcome is a reconciled transaction; a Search Visibility FTE's is a page published that passed the fact gate; a Customer Support FTE's is a ticket resolved at first line with a cited source; a Recruiter FTE's is a shortlist delivered.
The term exists because per-seat pricing cannot produce it. A seat's cost attaches to a person, and what that person produced with the tool is invisible to the invoice. Metered work attaches to the work, and the work is what was bought.
How is it computed on ftes.ai?
On ftes.ai usage is reported per Digital FTE, and outcomes are recorded per item in the Targets tab with the evidence behind each. The Performance tab sets the two side by side for a cycle. As a labelled example with your own figures to fill in: if a role consumed a certain number of tokens in a month and Targets records a certain number of reconciled transactions for that month, the cost per reconciled transaction is the first divided by the second, at the plan's rate.
The figure is per role by construction, because the numerator and the denominator are per role. A total across roles that produce different outcomes would be meaningless.
Why does it matter?
It matters because it is the comparison a finance leader can actually make. The baseline cost of the same outcome, recorded before the pilot from the team's own hours and rates, sets against the cost per outcome after two cycles, measured the same way, is a decision. A savings percentage from a vendor is not.
It also matters as a trend. A role's cost per outcome should be flat or falling as its remit settles, its memory is corrected, and routine classes earn autonomy; a rising cost with flat outcomes points at the model tier or at scope creep, and the monthly review reads it for exactly that.
What counts as an outcome?
An outcome is an item in the role's Targets with evidence, not an activity. A draft nobody approved is not an outcome; a page published after passing the fact gate is. A reply drafted is not; a ticket resolved at first line is. The distinction is the same one that keeps task counts out of the productivity measures: activity is cheap, and only results carry value.
Outcomes are valued conservatively when the comparison is made: at what they cost to produce before, from the baseline, never at a notional higher worth.
- Finance: transactions reconciled, exceptions resolved
- Support: tickets resolved at first line with a cited source
- Recruiting: shortlists delivered within the target window
- Search visibility: pages published that passed the fact gate
What is it often confused with?
It is confused with cost per seat, which prices access rather than work. It is confused with total spend, which says nothing about what was produced. It is confused with return on investment, which needs the baseline and the valuation; cost per outcome is one of the inputs to that calculation, not the result. And it is confused with a benchmark; it is your figure, from your record, and only comparable with your own baseline.
It is sometimes computed across roles. Do not; a page and a reconciled transaction are not the same outcome.
Where do you see it in the workspace?
In the Performance tab per role for each cycle, alongside usage and the approval rate, and in the FTE's monthly review. The baseline it compares against is yours to record before the pilot; the pilot and payback guides in the library describe how.
Related terms: token wallet, usage per role, targets, baseline, payback period.
A practical note for the first cycle: the figure will be higher than the steady state, because the first cycle includes the closer reading, the send-backs, and the remit tuning that the record later removes. Record it anyway, label it as the first cycle, and treat the second cycle's figure as the one to compare against the baseline; the difference between the two is itself useful, since it shows how quickly the role settled.