पाठशाला Pathshala · दल Dal, The team · Lesson 13 · Build

Onboarding: the first thirty days that decide retention

Most early attrition is decided in the first month. How to plan it with a buddy, a first win inside ten working days and a written review on day thirty, and what halving early exits is worth.

Pathshala, The Founder Library · 11 October 2026 · 8 min read

Old metal keys on rings hanging from hooks on a weathered wooden board. — Photograph: ClickerHappy · Pexels

Most early attrition is decided before the new hire has done anything that matters. A person who has no first win by the end of week two, no one to ask the small questions and no clear verdict at day thirty starts looking, and in India the next offer is usually one phone call away. Onboarding is the cheapest retention programme a startup will ever run.

This lesson builds a thirty-day plan with three parts that do the work: a buddy who is not the manager, a first win sized to land inside ten working days, and a review on day thirty with a written verdict. It shows what early exits cost and closes with the ritual that keeps the plan running when the founders are busy.

Why the first thirty days decide it

Gallup found that only 12 per cent of employees strongly agree that their organisation does a great job of onboarding new people. The same piece cites the Society for Human Resource Management on the cost of getting it wrong: turnover can reach 50 per cent in the first eighteen months, and replacing someone takes six to nine months of their salary once the search, the interviews and the ramp are counted. Startups are not better at this than large firms. They are usually worse, because nobody owns it and the founders assume a sharp hire will work it out.

The new hire is deciding three things in the first month, mostly without saying so. Whether the job is the one they were sold. Whether they can succeed here, which they judge from whether anything they have done has mattered yet. And whether they belong, which they judge from who talks to them when nobody has to. A company that answers all three by day thirty keeps most of its hires. One that leaves them unanswered finds out at the resignation.

The month decides something for the company too. Molly Graham, who built teams at Facebook and Quip, says that a decision on fit usually takes only a couple of months and that a no should be acted on quickly. A deliberate first month is how the company reaches an honest answer in that time rather than in a year.

Before day one: the gap between offer and joining

In India the gap between an accepted offer and a first day is often a notice period of one to three months, and it is where many hires are lost before they start. The current employer counter-offers, a competitor calls, or the excitement fades into doubt. Treat the gap as part of onboarding. A founder call every two weeks, an invitation to one team lunch or demo, and the first week’s plan sent in advance keep the decision warm. None of it should be work: an unpaid assignment during someone else’s notice period says the wrong thing about how the company treats time.

The paperwork belongs in the gap as well. Since the labour codes came into force in November 2025 an appointment letter stating the job, the wages and the social security is mandatory for every worker, so send it with the offer rather than on day one. Have the laptop ordered, the email and tool accounts created, the payroll and provident fund details collected and the IP assignment signed before the person arrives. The [offer letters lesson](/library/employment-contracts-offer-letters-that-hold-up) has what the documents must say. A new hire who spends day one waiting for a login learns that the company does not plan, which is the opposite of the lesson you meant to teach.

The buddy, the first win and the manager’s first week

The buddy is a peer, not the manager, who has been at the company for at least six months and does similar work. Their job is the questions a new hire will not ask a manager: where the real documentation lives, who actually decides, what the founder means by a terse message at midnight. Give the buddy thirty minutes a day in week one and three short check-ins a week after that, and write it into their goals for the month so that it is work rather than a favour. Choose buddies who like the company. A cynical buddy onboards people into cynicism.

Hands measuring a timber plank with a ruler on a carpenter’s bench beside a hammer and tools. — A first win is a piece of real work small enough to finish in ten days and real enough that someone notices. Photograph: Ono Kosuki · Pexels

The first win is a real piece of work, sized to ship inside ten working days, that a customer or a colleague will notice. For an engineer it is a small fix in production. For a salesperson it is a live demo run alongside a founder and then alone. For an operations hire it is one broken process mapped and fixed. Choose it before the person joins and write it into the plan. The first win is not a test. It is how the hire learns the path from work to impact in this company, and how they answer their second question: can I succeed here.

The manager’s first week holds four meetings. On day one, the plan for thirty days, with what success looks like at day thirty written down. On day two or three, the history of the company told honestly, including what went wrong. On day five, a first one-on-one, which starts the weekly habit that the [one-on-ones lesson](/library/one-on-ones-and-feedback-that-lands) describes. And one meeting with a founder, whatever the hire’s level, about why the company exists. Ben Horowitz, writing about hiring executives, advises defining what the hire should accomplish in the first thirty days before the search even starts. The same discipline works at every level.

What early exits cost

The case for all of this is arithmetic. A company making twenty hires a year that loses three in ten inside six months has six early exits. At an average cost to company of ₹12 lakh and a replacement cost of six months’ pay, each exit costs ₹6 lakh and the year costs ₹36 lakh, before counting the work not done while the seat was empty or the effect on the people who stayed. Halve the early exits and the company keeps three people and ₹18 lakh. Set your own numbers.

Read the green squares as the prize. A plan that costs a buddy two hours a week and a manager five for one month is cheap against a single replacement. The figure leaves out the larger cost: early exits teach a team that new people do not last, and a team that believes that stops investing in the next one.

Onboarding is the cheapest retention programme a startup will ever run. Its cost is a month of attention. Skipped, its price is a replacement.

Weeks two to four: from shadowing to owning

Structure the month as a hand-over of ownership. Week one: shadow and learn, with the first win under way. Week two: ship the first win and take one recurring responsibility, such as a weekly report, a place on the support rota or a set of accounts. Week three: own a small area, with the buddy as back-up rather than guide. Week four: work as a full member of the team, with the manager watching for where the plan was wrong.

Two habits help. Ask the new hire to keep a list of everything that confused them, and fix three items on it by day thirty, because a newcomer sees the gaps in the company’s documentation that everyone else has learned to walk around. And introduce them to one person outside their team each week, by name and with a reason. Google, which surveys its new hires at thirty, ninety and 365 days on tools, productivity and connection, redesigned its programme partly because new people said they wanted relationships beyond their own group. A startup of thirty people can do the same with a list and a calendar.

The thirty-day review

On day thirty the manager and the new hire meet for an hour with the day-one plan on the table. Three questions go each way. To the hire: is the job the one you were sold, what would have made the month easier, and what have you seen that we should fix. From the manager: here is what you did well, here is what we need more of, and here is our view of the next sixty days. Write the verdict in three lines: on track; on track with named changes; or not on track, with the specific gap and a date.

The third verdict is rare and it matters most. A gap seen at day thirty can be closed by day ninety with a clear plan. The same gap discovered at a six-month review usually ends in an exit that surprises nobody but the person leaving. Horowitz suggests letting go of an executive who is not up to speed within thirty days. For most roles that is too fast, but the principle holds: by day thirty the company should know which way the hire is heading and should say so.

Run a short version again at day ninety with the same questions and compare the answers. If several hires in a row say the job was not the one they were sold, the problem is in the interview rather than the onboarding, and the [interviewing lesson](/library/interviewing-for-judgement-not-pedigree) is the place to fix it.

The onboarding ritual

For every hire the manager writes the thirty-day plan before the offer is signed: the first win, the buddy, the four first-week meetings and what success at day thirty looks like. During the notice period: a founder call every two weeks and every document signed. On day one: the laptop, the accounts and the plan ready. On day thirty: the review and the three-line verdict. On day ninety: the short version again.

Once a quarter one person, a founder until there is someone better placed, looks at every hire of the last two quarters: who left inside six months, what their thirty-day review said and what was on their list of confusions. Count the early exits and compare the rate with the quarter before. The number is small enough to see each person behind it, which is the point. Every early exit is the story of a month the company did not plan.

Survey figures are averages across large employers; your own early-exit rate is the number to watch. Nothing here is legal advice.