Answered from these docs only, by a model that cannot see your account. Check the pages it cites.

Running a workforce impact study

Before the Workforce impact sign-off: the groups in scope, hours freed by group, what they'll go to, who signs, what to attach, and consultation.

View as Markdown

Answered from these docs only, by a model that cannot see your account. Check the pages it cites.

Checked against the product on November 9, 2026

A workforce impact study says which teams an AI initiative changes the work of, how many hours it frees in each, and what those hours will go to. It's what whoever signs the Workforce impact sign-off signs against. See Required sign-offs before approval.

Roiva doesn't run the study. It sends no survey, connects to no HR system and scores no role or person. It records who signed the study off and the hours and FTE they agreed. The study you attach is kept as the sign-off's evidence, like any document on the initiative, so it should hold groups and counts, never names. Your own team runs it, or a partner does.

The workforce impact worksheet lays out the four parts below on one page. It needs no account. A spreadsheet with the same columns works as well.

1. The scope

List the groups the initiative changes the work of, by team or role group, never by name: "Claims intake, Phoenix", "Tier 1 support agents". For each group, record:

  • the headcount in scope, as people and as FTE;
  • the tasks the AI changes, in a line each: "first read of an inbound claim", "drafting the reply to a billing ticket";
  • what the AI does in each task: drafts, checks, routes, or does it end to end.

A group of one or two people names them. Fold it into a wider group.

The study is about tasks and hours. It isn't a list of roles to replace, and it shouldn't score any.

2. Hours freed, by group

Work the hours out from the bottom up, task by task:

  1. How many times a year the group does the task.
  2. The minutes each one takes today, and with the AI. Time them where you can. See Running a time study.
  3. Minutes saved × times a year ÷ 60 = hours freed a year.
  4. Take off the time the AI adds: checking its output, fixing what it gets wrong, training.

Add the tasks up for each group, then add the groups. For example, a group that reads 30,000 claims a year and saves 6 minutes on each frees 3,000 hours a year, about 1.4 FTE. That's 3,000 hours spread across everyone who reads claims, not one role.

Then compare it with the figure from the top down. The sign-off starts from the business case: its planned freed capacity a year, divided by an hourly rate, is hours a year. Those hours divided by 2,080 (40 hours a week for 52 weeks) are FTE. The rate is the one the initiative's capacity formulas read, a role's own rate such as a developer's where that's what they price, and otherwise the labor rate: the initiative's own where one is set on its Health tab, else the account's, under Method → Assumptions, or Roiva's default where neither is set. The sign-off drawer says which it used. Where the formulas read other hourly rates, it uses the labor rate and says so, so check the hours and correct them on the sign-off.

When the two differ, say why in the study. Whoever signs can correct the hours and the FTE on the sign-off, and it keeps what they signed. Change the FTE too when you change the hours: left as it was, it's worked out from the new hours at 2,080 each. If the business case is what's wrong, change it there. See Writing an initiative's business case.

3. What the hours will go to

For each group, split its hours across what they'll go to. Unknown is an answer.

Goes to What it means In Roiva
Redeploy The same people do other work. Freed capacity
Absorb growth More volume, same team. Freed capacity, unless a planned hire is canceled
Avoid hires A planned hire isn't made. Cash, once shown
Cut contractor or overtime spend A bill that falls: contractors, outsourcing, overtime. Cash, once shown
Reduce roles Fewer roles in the group, so its payroll falls. Cash, once shown
Undecided Not yet known. Freed capacity

Only avoided hires, closed roles and cut contractor, outsourcing or overtime spend become cash, and only once someone shows they did. Everything else stays freed capacity: hours given back, priced at what they cost. Payroll is the same afterwards, so it's never cash. Growth absorbed is an avoided hire only where a hire planned for that growth is canceled, with a date on the decision. "We would have hired someone" stays capacity.

A closed role counts from the month its payroll stops, not the month it's decided. What closing it costs is a cost of the initiative, recorded on its Costs tab under the Severance cost type for the month it's paid, so ROI nets it. Record an amount, not a name. Consultation duties most often apply here: see below.

Roiva doesn't hold this split as figures. It's in the study you attach.

4. Who signs it off

The business owner of the groups in scope, or HR. Usually not the initiative's owner: an owner can sign their own initiative's sign-off only where the account allows it, or where nobody else in the account can approve: no other Owner, Admin or named approver.

  • A member signs it themselves with Sign off, or is asked to with Request sign-off.
  • For someone without a seat, such as an HR business partner, an Owner or Admin uses Record for someone else, with the evidence that shows their sign-off.

Attach the study as the evidence. Add it to the initiative's Documents section as a review record and choose it on the sign-off, or link to where it lives. A sign-off takes one piece of evidence, so add what the study rests on to the Documents section too: the time study, the volume report, the hiring plan. Put the split from part 3 in a line in the sign-off's note, so an approver reads it without opening the file.

Consultation

In some jurisdictions, an employer has to inform or consult employee representatives, such as a works council or a union, before some changes to work or jobs. Whether that applies, and when, is a question for your employment counsel. Roiva gives no legal advice.

Roiva doesn't track consultation itself. Add its record to the initiative's Documents and say in the sign-off's note that it happened. To make approval wait on it, an Owner or Admin can add a required sign-off of its own, such as Employee consultation, under Settings → Approvals, signed against that record.

After approval

Freed capacity is booked month by month, as the initiative's formulas measure it or as someone records it. Where some become cash, a person records it each month as Capacity realized as cash and cites what shows it: a canceled requisition, a closed position with payroll before and after, a contract not renewed, an overtime report, a vendor's invoices before and after. For a closed position, cite a record by position or by group, such as the team's payroll summary, rather than one that names the person. It's approved only up to that month's approved freed capacity, and a dollar is counted once. See Adding costs and value entries. Payroll that fell by more than the month's freed capacity has its own type for the part above it: see Adding costs and value entries.

What the study shows, and what it doesn't

It shows the groups the initiative changes, the hours it frees in each and how they were worked out, and what each group's hours are planned to go to.

It doesn't show:

  • what happens to any one person;
  • that the planned split will happen;
  • that any hours became cash;
  • that the minutes timed hold for every task.