What decision does the First-Function Matrix frame?
Which function a team's first Digital FTE should own. The wrong first choice is the most common reason pilots fail: a team picks the most exciting function, usually a judgement-heavy one, the record builds slowly, the approvals are hard, and the pilot is judged a failure before a routine class could earn anything. The matrix replaces excitement with two measurable properties and one check, and it points most teams at the same unglamorous corner, which is where the first month goes well.
The frame works for any team on ftes.ai and for any vertical; the industry pieces in the library apply it to each.
What are the axes and the check?
Evidence clarity, scored high to low: does the work have a checkable right answer that lives in a system you already run, so that a Digital FTE can propose with the evidence attached and a person can verify in seconds? Reconciliation against a bank feed is high; a positioning decision is low. Hours recovered, scored from a fortnight's baseline: how many hours per week does the function currently consume, and at whose rate? The check: can you write the always-escalate list for the function today, naming what must always reach a person and who receives it? If the list needs a workshop to write, the function is not ready to be first.
Place each candidate on the two axes and mark whether the check passes.
- Axis 1: evidence clarity, high to low
- Axis 2: hours recovered per week, from the baseline
- Check: escalate list writable today
How does the worked example read, with the inputs left blank?
Candidates: [function A], [function B], [function C]. For each: evidence system, [system]; right answer checkable in seconds, yes or no; clarity score, [high, medium, low]. Baseline hours per week over the fortnight from [date] to [date]: [hours] at [rate], [hours] at [rate], [hours] at [rate]. Escalate list writable today: for [function A], yes, [items] to [receiver]; for [function B], no, the list needs the [team] workshop; for [function C], yes, [items] to [receiver]. Placement: [function A] high clarity, [hours] hours, check passed; [function B] medium clarity, [hours] hours, check failed; [function C] high clarity, [hours] hours, check passed. First function: the top-right candidate with the check passed, [function].
If two candidates tie, choose the one whose baseline was easier to record; it will be easier to measure too.
Where do the common functions land?
Top-right for most teams: bank and processor reconciliation, order-status and how-to support in one category, CRM hygiene, interview scheduling, receivables follow-up drafts, and pre-arrival guest questions. High clarity, moderate hours: knowledge base upkeep, listing content from a property record, payout matching. High hours, medium clarity: inbound qualification, first-touch outreach drafts, product content; strong second roles once the approval habit exists. Bottom-left: market entry, research synthesis, campaign strategy, anything where the right answer is a judgement; third or fourth roles, managed closely, if at all.
A team whose top-right corner is empty may not need a Digital FTE yet, and the matrix says so honestly.
How is the frame applied?
Record a fortnight's baseline for each candidate function using the baseline time template: hours per task, at whose rate, with the tasks named. Score clarity by asking, for the most common task in each function, where the right answer lives and whether a person could verify a proposal in seconds from an attachment. Attempt the escalate list for each in ten minutes. Place the candidates, choose the top-right one that passed the check, and write its remit. The first-run wizard will ask only for what the platform does not already know.
Re-run the matrix when the Second-Role Test says the team is ready for another; the corner will have shifted because the first role now produces evidence for adjacent functions.
What does the frame refuse?
Starting with the function the leadership finds most interesting. Starting with a function whose evidence lives in people's heads. Starting with anything whose escalate list needs a policy decision nobody has made. Starting with several functions because they all look good. And starting without a baseline, because the matrix's second axis is then a guess and the first review will have nothing to compare. The refusals are why the first month is boring in the way a good first month is.
The persona pieces for founders and COOs in the library apply the same corner from those seats.