In this guide
Most retention budgets are spent company-wide on a problem that lives in one or two teams, and spent after the exits rather than before them. Wanderfly exists to fix the timing: it shows people leaders which departments are carrying too much, what is driving it, and where it is heading over the next 14, 30 and 60 days, without ever showing anyone how one individual is doing.
Free to start. ₹249 per user per month for risk scores and forecasts. Live in an afternoon.
The short answer
Everybody already knows the retention advice. Pay fairly, promote clearly, train your managers, be flexible, say thank you. It is on every page that ranks for this, it has been the same list for twenty years, and it is broadly correct.
It is also almost useless on a Monday morning, because it does not tell you which of those things is wrong in your company, in which team, or whether you have about three months or about three weeks.
That is the gap this page is about. Not what to do in general. How to find out what to do here, early enough that doing it still matters.
Retention is a measurement problem first
Ask a founder how retention is going and you will usually get one number for the whole company. That number is the least informative thing in the building.
A company sitting at 14% normally has several departments in single digits and one somewhere near thirty. Spending the retention budget evenly across all of them is spending most of it where there was no problem. And a company-wide figure that mixes voluntary with involuntary exits mixes an outcome you are trying to prevent with a decision you made on purpose.
Four splits turn it into something you can act on.
- Voluntary against involuntary. Only one of them is a warning.
- Regretted against not. Marked by the manager at the time of the resignation, not reconstructed at quarter end.
- By tenure. Exits inside 90 days are about hiring and onboarding. Exits at eighteen months are about something else.
- By team. Where the whole thing usually turns out to live.
The attrition rate formula and calculator will get you the arithmetic in twenty minutes, and what attrition means in HR covers the vocabulary if your reporting uses the words loosely.
What actually predicts an exit
Every number in the previous section is about people who have already gone. Useful for diagnosis; useless for prevention. For prevention you need something that moves while the person is still there.
Gallup, in research published in March 2019, found that 52% of people who left voluntarily said their manager or organisation could have done something to keep them, and 51% said that in the three months before they left, nobody had spoken to them about their satisfaction or their future.
That is the whole problem in two numbers. The exits were mostly preventable, and the three months in which they could have been prevented passed in silence.
India adds its own pressure to that silence. Gallup’s 2026 State of the Global Workplace report, covering 2025, puts employee engagement in India at 23% (a seven-point fall against the prior three-year average), with 17% of people thriving in life and 28% reporting they felt a lot of stress the previous day. Meanwhile 53% say it is a good time to find a job where they live. Quiet people and an open market is not a comfortable combination.
What fills those three months, more often than anything else, is burnout: it builds slowly, it shows up in how people answer ordinary questions, and it can be measured without reading anybody’s messages. That argument, and the places it stops being true, is set out in burnout and attrition.
How Wanderfly measures retention risk
No single signal is worth much on its own. A bad mood on a Tuesday is a bad Tuesday. Wanderfly combines four per department, over time.
| Signal | Weight | Where it comes from |
|---|---|---|
| Stress | 35% | The PSS-10 perceived stress scale, run monthly |
| Meeting workload | 30% | Calendar timing from Google Calendar or Outlook, read-only, never the contents |
| Mood | 20% | Daily check-ins, in the app or in Slack |
| Engagement | 15% | Whether people are still answering at all, often the first thing to go |
It recomputes nightly over the last 28 days and produces a score from 0 to 100 with a confidence level, banded low, medium or high, plus a forecast at 14, 30 and 60 days. The forecast is the part that matters for retention: a department at 45 and climbing is a more urgent problem than one at 60 and falling.
What you do with a signal
A score nobody acts on is an expensive way of being surprised. The response is deliberately unglamorous and it is almost always about the shape of the week rather than its length.
- Look at what is driving it. The score breaks down by input, so “this team is struggling” becomes “this team is in meetings for 58% of its week and the stress score followed it up”.
- Change one condition, within a fortnight. Kill the recurring meeting. Move the deadline. Rebalance the on-call rota off the same two people. Say which thing you changed and why.
- Make the private help genuinely private. Expert sessions are paid for by the company and booked without anyone being told, because a counsellor you have to ask your manager about is a counsellor nobody uses.
- Watch whether it moved. Four to eight weeks for the leading measure. Two to three quarters for the attrition rate itself.
Which retention strategies are worth the money
Being direct about this, as a company that sells into this budget. Based on what we see across the teams we work with.
| Strategy | Verdict |
|---|---|
| Fixing below-market pay | Do this first. Nothing else on this list works underneath a pay problem. |
| Changing the shape of the week | The highest-value thing most startups can do, and nearly free. Fragmentation and unpredictability cost more than hours. |
| Manager one-to-ones with a real agenda | Enormous effect, almost no cost. The direct manager is the single biggest factor in a small company. |
| Stay interviews | Worth it, twice a year, run by the manager, not attached to the appraisal. |
| Counter-offers | Buys a quarter. The reason they looked is still there and now costs more. |
| Annual engagement survey | Worse than nothing if nothing follows it. It teaches people that describing a problem achieves nothing. |
| Offsites and engagement events | People enjoy them. They do not stay for them. |
| Activity monitoring | Actively harmful. Reduces exactly the discretionary effort you are trying to keep. |
The step-by-step version of all of this, sized for a team of 20 to 200, is in the 90-day plan.
Two of those rows need somebody to own them, and in most companies past about fifty people that somebody is an HR business partner rather than the founder. We have written separately on what the HRBP role actually covers, which is worth reading if you are deciding whether you need one yet.
Without watching anyone
Retention tools have an obvious temptation: if you could just see who is on LinkedIn, who is logging off early, who is quiet in Slack, surely you could predict the exits.
We do not do that, and not only for privacy reasons. It does not work. People answer honestly only while they believe the answer is not attributable, so the moment you build a tool that names individuals you have destroyed the data it depends on, and you are left measuring how safe people feel being candid.
- No leader can see an individual’s mood, stress score, conversations or expert bookings.
- No group view is shown for fewer than five people, and the floor is enforced in exports and share links too.
- Calendar access is read-only and covers timing, never contents. Wanderfly never reads Slack channel messages.
Wanderfly is an Indian company and handles personal data under the Digital Personal Data Protection Act, 2023. The full detail is on our security page, and the longer argument is in detection versus monitoring.
What it costs
Per user per month, in rupees, before 18% GST.
| Plan | Price | Team size | What you get for retention |
|---|---|---|---|
| Free | ₹0 | Any | Daily check-ins, surveys and polls, the Wellbeing Baseline, team analytics |
| Starter | ₹99 (₹79 yearly) | Up to 25 | Adds department analytics and Kudos |
| Growth | ₹249 (₹199 yearly) | Up to 100 | Adds risk scores and forecasts, calendar workload, retention analytics, Initiatives, Safe Space and Expert Support |
| Enterprise | On request | Unlimited | Adds a full risk audit export and AI-written summaries for leadership |
Every new team starts with a 30-day Growth trial, so you see the scores before you pay for them. Set that against the cost of the thing you are trying to avoid: Gallup put replacing an employee at one-half to two times their annual salary, which on a ₹18 lakh engineer is somewhere between ₹9 lakh and ₹36 lakh. See pricing for the full plans.
When Wanderfly is the wrong choice
We would rather you bought the right thing. Pick something else if:
- You need an HRMS. Payroll, leave, appraisals and exit workflows are a different category. Keka, Darwinbox and Zoho People do that; we do not, and we are not trying to.
- You want to know who is idle. That is productivity monitoring. There is no activity tracking, screenshots or keystroke logging here, by design.
- You want named flight-risk predictions. Some tools offer this. We think it breaks the measurement and the trust at the same time, so we do not.
- You are paying well below market. Fix that first. Measurement will only tell you something you already know.
- Your team is under five people. The anonymity floor means you would see very little. Talk to each other instead; it is genuinely the better tool at that size.
If it does sound like the right fit, book a walkthrough and we will use your team size and stage rather than a demo account.
Questions people ask
What is the best employee retention software for Indian teams?
It depends which half of the problem you have. If you need to record exits, run appraisals and manage leave, that is an HRMS such as Keka, Darwinbox or Zoho People. If you need to see which teams are heading for trouble before anybody resigns, that is what Wanderfly does: team-level burnout and retention risk scores from daily check-ins, the monthly PSS-10 stress scale and calendar meeting load, with forecasts at 14, 30 and 60 days, plus private expert sessions paid for by the company. Pricing is in rupees, from a free plan to ₹249 per user per month. Most companies need both; they are not substitutes for each other.
How do you improve employee retention?
Start by finding out where you are actually losing people, because a company-wide attrition figure almost always hides one or two teams carrying most of it. Then change the conditions in that team (meeting load, fragmentation, who carries on-call, the quality of the one-to-one) rather than adding a company-wide benefit. Then build something that tells you before the next resignation instead of after it.
What is retention risk?
The likelihood that people will leave a team in the near future, estimated from measures that move before a resignation rather than after one. Wanderfly scores it by department on a 0 to 100 scale with a confidence level, recomputed over the last 28 days. It is a statement about a team’s conditions, not a prediction about any named individual.
Does employee wellbeing actually improve retention?
The parts that change working conditions or get people real help do. Benefits nobody uses do not, and a dashboard nobody acts on does not either. The mechanism that works is short: measure often enough to see a team drifting, act on it within a fortnight, and make expert support private enough that people will actually use it. Skip the second step and the first two are decoration.
How long before retention work shows up in the numbers?
Leading measures move in four to eight weeks if you change something real. The attrition rate itself takes two to three quarters, because it is calculated over the last twelve months and because anyone already serving a 60- or 90-day notice period is still counted. Judge the first quarter on what you changed, not on the rate.