What each round is for
Every round is the same fifteen-minute conversation with the same promises. What changes is the question you're trying to answer. Here's what each one is for.
- A survey asks everyone the same forty questions. We ask nine, and change the tenth depending on what you said.
- We talk to one team this month, because something’s wrong there.
- You get two pages with the quotes, not a dashboard with a score.
- A survey comes from HR. We don't work for you, and we say so before the first question.
1 / 9
Stay interview
- When to run it
- Something's moving on a team: a new supervisor, a schedule change, three resignations in a month, a shift that's gone quiet.
- What it finds
- Why people stay and what would make them leave, while they still work for you.
- Why your other systems can't
- Exit interviews arrive after the decision is made. Engagement surveys ask everyone forty questions once a year. Neither hears the reason in time.
- What you get
- Three numbers (likely to leave, would stay if one thing changed, and concerns about a supervisor), the themes five or more people raised, three things to do this quarter. Run the same team again and the report shows what moved.
Honest limit: On a team under about eight, most themes will print as "fewer than 5."
See a sample →2 / 9
Exit interviews
- When to run it
- Open for 12 months, from people in their last weeks. Hand over the link when you accept the notice. Exit interviews: priced on your yearly leavers, open for 12 months, renew each year. Once five have finished, open rounds refresh on the 1st of each month when there are new completions. If completions cross the tier you paid for, pay the difference before the report updates. At month 11, you can renew through a new paid round; nothing renews or charges automatically.
- What it finds
- The real reason people left, said to someone who doesn't work for you, compared with what the people who stayed are saying.
- Why your other systems can't
- An HR exit interview gets "a better opportunity." A neutral one gets the schedule, the lead, the dollar.
- What you get
- The reasons that cleared five, and what leavers say against what stayers say.
Honest limit: It needs time to build up five answers; small teams may take a quarter or two.
See a sample →3 / 9
Ask before a change
- When to run it
- The week before you decide a schedule, overtime or shift-pattern change.
- What it finds
- What the change would cost people in practice (childcare, a second job, getting to work), which objections come from people who'd otherwise stay, and what they'd want in return.
- Why your other systems can't
- The decision is made in the office and the cost shows up as resignations a month later.
- What you get
- The objections ranked, the practical costs, the trade-offs people would accept, and an estimate of what backfilling would cost, using your replacement cost or a standard figure if you haven't set one.
Honest limit: It's input, not a vote. On a represented team it goes through the union.
See a sample →4 / 9
Floor check
- When to run it
- Every quarter, or before a busy season.
- What it finds
- What slows the work: the machine that fails every afternoon, the supply that's always short, the step everyone works around, with how often it happens.
- Why your other systems can't
- Tickets only exist for what someone reports. The small, daily friction never becomes a ticket.
- What you get
- The problems ranked by how many people raised them, and three fixes in order of how many people they'd help.
Honest limit: If someone describes a hazard, the report notes it even if only one person raised it. It isn't a safety reporting system.
See a sample →5 / 9
First 30 days
- When to run it
- Hand the link to each new hire in their third or fourth week.
- What it finds
- What the job ad or interview overstated, what the first week should have included, and the moment people nearly quit.
- Why your other systems can't
- Your HRIS knows the 90-day quit rate. It doesn't know why.
- What you get
- Two plain lists to fix the posting and the first week, and material for a crew-written job preview, a draft you approve before anyone sees it; and, once five or more people came from the same place, which recruiting sources bring people who stay.
Honest limit: It needs at least five new hires in the window.
See a sample →6 / 9
Benefits in plain words
- When to run it
- Four to six weeks before open enrollment.
- What it finds
- What people don't understand, what they'd use if they knew how, what they pay for and never use, and, where five or more people say, why.
- Why your other systems can't
- Benefits platforms record elections. They don't record whether anyone understood them.
- What you get
- The questions to answer in the packet, the benefits people skip and why, and three things to explain or change before enrollment.
Honest limit: It explains confusion; it never gives plan advice. The plan documents govern.
See a sample →7 / 9
Did they understand it?
- When to run it
- A week after a training or a new policy, or two weeks after open enrollment closes.
- What it finds
- Which parts the crew took the wrong way, and what to say instead.
- Why your other systems can't
- Training systems record that someone completed a course, not what they took away from it.
- What you get
- A short list of what to re-explain and the sentence to use.
Honest limit: It checks the explanation, never the people, and can't be used to show anyone was trained.
See a sample →8 / 9
Schedule constraints
- When to run it
- Before you build or change a shift pattern.
- What it finds
- When people can actually start and stop, which days can't move, and why (childcare, a second job, transport), as counts for the team.
- Why your other systems can't
- Scheduling software assumes availability. It never asks.
- What you get
- A count for each constraint and three ways to build a pattern people can work.
Honest limit: Most useful on teams of 20 or more; on smaller teams the counts mask. On a represented team it's input to bargaining, not a substitute.
See a sample →9 / 9
How pay works here
- When to run it
- When pay comes up as a reason people leave, or before pay ranges become public under transparency laws.
- What it finds
- What the crew believes about raises, differentials and getting to the next rate, including what they've got wrong and what's true.
- Why your other systems can't
- Payroll knows every rate and a market study knows what others pay. Neither knows the story the crew tells each other.
- What you get
- Each belief marked as matching your pay facts or not, the questions people can't answer, fixes where an unfair belief turns out to be true, and a draft plain-language explainer in English and Spanish for you to review before it can be downloaded.
Honest limit: It doesn't tell you what to pay anyone, and if a comparison is accurate, the report says so.
See a sample →Stay and First 30 days include a skippable job-recommendation question that gives employers a simple round-to-round measure of how the crew feels; a team count is never a company-wide rating.
After the report
The part that makes people answer next time.
Hear back
Post three lines: what you heard, what you'll do, by when. The team sees it on the same link they used, and the next round opens by asking whether it happened.
Practice the huddle
Rehearse telling the crew, with a crew built from your own report, before you do it for real. It tells you what to say first and what to stop saying.
Say it for the floor
Paste the memo nobody reads and get a version a tired person on a break can follow, in English and Spanish, with every rule still in it.
All three use only what’s in your report. Nothing about your practice or your memos is kept. Between rounds, the open line keeps the door open.
What we won't do for any customer
- No white-labeled or employer-branded interviewer.
- No questions about a named person.
- No lowering of the five-person rule.
- No access to raw conversations.
- No changes to the promises employees see.