Four former engineers, rehired over about three years. Three of them were among the best hires I made in the years they rejoined: productive in a fortnight, honest about what had been broken when they left, and visibly better than when they had gone. The fourth lasted seven months and left again, worse for both sides than the first departure. The interesting part is that I could not have told you at offer stage which one was which, and the step that would have caught it comes after the offer, not before.
This is the method I use now, written as seven steps in the order I run them. It assumes the Singapore market, where a senior engineering search commonly runs a quarter and notice periods compound the delay, which is precisely why the returning engineer is worth a process rather than a hunch.
Why the Returning Engineer Is the Most Underused Channel Here
The arithmetic is not subtle. A former engineer who knows your domain, your codebase and your deployment process skips the part of onboarding that cannot be compressed. They also come with something no reference check provides: you have watched them work for a year or more, under pressure, on your actual problems.
Against that, two real risks. The first is that you rehire a performance problem because you remember them fondly and forget specifically. Memory is extremely kind to people who left. The second, less discussed, is the signal to the team: if resigning turns out to be the fastest way to come back at a higher band, you have taught your organisation something you did not intend to teach.
The seven steps exist to capture the upside and shut down both risks.
The 7 Steps, in Order
Step 1: Build the alumni list before you need it
Almost nobody does this, and it takes about fifteen minutes per departure. For every engineer who leaves, record six fields at the time they leave, not from memory two years later: role and level, the systems they owned, departure date, stated reason, your honest assessment of whether you would rehire them, and a note on anything specific that would have kept them.
Write the rehire field while the feeling is accurate. Six months later you will remember the person who left as either a hero or a traitor, and neither memory is useful. The point of the written note is that it was made by someone who had the facts.
This pairs directly with running departures properly in the first place. Our method for running developer exit interviews in Singapore produces exactly the raw material this list needs, and the two processes should be one process.
Step 2: Separate the push from the pull
This is the single most predictive distinction I have found, and it is worth being rigorous about.
A pull departure means something better appeared: a bigger role, a different domain, a company they had always wanted to work at, a relocation, a founder opportunity. Nothing here drove them out. They are a strong rehire candidate, because the reason they left is a reason that has now been satisfied, and what they come back with is a comparison.
A push departure means something here made staying worse than leaving: a manager, a stalled career path, a project they could not stand, compensation that had drifted below market, an on call rota that was eating their life. These are rehireable only if the specific thing that pushed them has changed, and only if you can name the change in one sentence without using the word culture.
Most departures are a mix, and the honest version of this step is asking which was dominant. Here is the uncomfortable part: the exit interview answer is frequently the polite version. People in Singapore, in my experience, routinely give a pull reason for a push departure, because the market is small and burning a bridge is expensive. So read the exit interview, then read the eighteen months before it.
Step 3: Apply the three question disqualifier before you reach out
Ten minutes, three questions, and one clear no stops the process. Do this before any contact, because an approach you have to walk back is worse than no approach.
- Would you have kept them if they had never resigned? Not would you have been sad. Would you have actively fought to retain them. If the honest answer is that their departure solved a problem for you, stop.
- Has the thing that made them leave actually changed, and can you name it in one sentence? The manager moved on. The career path now has published levels. The rota is now six people instead of three. Compensation was rebenchmarked in March. If your sentence contains the words culture, journey or evolving, you do not have a change, you have a hope.
- Would their old team be glad to see them back? Ask two people who worked with them, informally, before you contact the candidate. This question has saved me twice, and it is the one most likely to be skipped because the answer might be inconvenient.
One no is enough. Two yeses and a maybe is a conversation.
Step 4: Reach out as a person and name what has changed
The approach matters more here than in any other channel, because the candidate already has a complete opinion of you.
A direct message from the person they actually worked for outperforms anything a recruiter sends, by a margin I would describe as embarrassing. Keep it short, name the specific change, and make it a conversation rather than a pitch. Something close to: the rota is six people now and the platform team has its own roadmap, two things you said were broken when you left. I would like to buy you lunch and hear what you have been building, with no agenda about a role.
Two things to avoid. Do not open with money, which signals that nothing else changed. And do not approach someone in their first six months somewhere new: you look opportunistic, they feel compromised, and if they were going to be reachable they will still be reachable at month nine.
Step 5: Interview them properly, on what has changed in them
Skipping the interview is tempting and is a mistake. But interviewing them like a stranger is also a mistake, and slightly insulting.
What you already know: how they work, how they handle pressure, whether they are honest about problems, how they treat juniors. Do not re-test it. What you do not know is everything that happened while they were away, and that is where the whole interview should sit.
- What did they actually own elsewhere? Scope, not title. Titles inflate fast in this market.
- What did they learn that you do not have? The best returners bring back a practice, a tool or a hard won opinion. If they come back with nothing new, they were coasting, and they will coast here.
- Is the role you are offering bigger than the one they left? If not, you are asking them to go backwards and the honest conversation is step 6.
- Involve one person who never worked with them. This is the most valuable hour in the process. Someone without the shared history will notice things nostalgia hides.
π‘ Our Expert Take
The question I now ask every returning candidate, and the one that has told me the most: what did we do badly that you only understood after you left. A good returner answers this immediately and specifically, because leaving gave them a comparison and they have thought about it. A weak returner says everything was fine and they just missed the team. That second answer sounds warm and it is a warning, because it means either they are managing you, or they genuinely learned nothing from a year somewhere else. Either way you are not getting the upgraded version of the person you hired, you are getting the same one with a gap in the middle.
Step 6: Negotiate the level honestly, including the uncomfortable part
Benchmark the offer against what the Singapore market pays for who they are now. Not against what they earned when they left, which is both out of date and the exact anchor that makes returners feel they are being punished for having gone.
Then do the thing most hiring managers avoid: say explicitly whether this is a return at a higher level or at the same one, and why. The failure mode is ambiguity, not disappointment. A returner who believes they are coming back more senior, into a role that is in fact identical to the one they left, disengages within a quarter and you will not know why. People accept a flat level readily when it is named, and resent it enormously when they discover it.
If the level is the same, say what would change it and roughly when. If it is higher, say which part of the scope is new, so the rest of the team can see the reason and does not conclude that resigning is a promotion strategy. Where the bigger role genuinely exists on a different team, a structured internal move after they land is often the cleaner answer than inventing seniority at offer stage.
Step 7: Re-onboard them as a new joiner, not as a returning hero
Here is where my one failure actually happened, and it had nothing to do with the interview.
He had been strong before, left on a clear pull, came back on a bigger title. We gave him access on day one and nothing else, because he knew the place. Except the codebase had moved substantially in eighteen months, two services he had never seen had become central, the deployment process had changed twice, and half the team was new to him. None of which he asked about, because asking would have meant admitting that the returning expert did not know. So he guessed, quietly and competently, for about two months, and by the time it surfaced in a review he had already decided the problem was him.
Give a returner the full programme. Access on day one. An architecture walkthrough covering what changed while they were gone, delivered as a session rather than a document. A named buddy, ideally someone who joined after they left, which fixes the direction of the relationship nicely. And a written thirty day plan with the same structure every new joiner gets.
It costs a week and it is the difference between my three successes and my one failure. If you are filling a seat that has been empty for a while, pairing this with a proper engineering backfill plan stops the returner from inheriting a pile of undocumented decisions on day three.
No alumni list to work from?
Most teams reach this article having realised they never kept one. HireDeveloper.sg sources pre-vetted engineers for Singapore employers while you build that habit for next time, and we are happy to help you reconstruct the list and sanity check which former colleagues are genuinely worth an approach. If the honest answer is that none of them are, we will say so and get on with the search.
Get startedThe Difference Between My Three Successes and My One Failure
Laid side by side, the pattern is almost dull. All four passed the interview. All four left on reasonable terms. The three who worked out got a re-onboarding and the one who did not, did not.
π‘ Our Expert Take
If you take one thing from this, make it step 1, and start it this week with the people who left in the last two years rather than waiting for the next departure. Writing six fields per leaver takes an afternoon, and it converts a channel that currently depends on somebody remembering a name into something you can work deliberately when a role opens. The teams I see doing this well treat the alumni list as part of the sourcing pipeline, reviewed whenever a requisition is approved, alongside referrals and the open market. The teams I see doing it badly discover six months later that the perfect candidate for an open role spent a year on their own payroll and nobody thought to call.
FAQ: Rehiring Former Developers
Is rehiring a former developer a bad signal to the rest of the team?
It depends on why they left and how the return is handled. Someone who left for a bigger role, learned something real and came back into a genuinely larger role reads as a compliment to the company. Someone who left loudly over a problem that still exists, and returns to the same seat at a higher salary, teaches your team that resigning is the fastest route to a raise. The difference is whether the original cause was fixed and whether you can name the fix in one sentence.
How long should you wait before rehiring someone who left?
There is no fixed interval, but the useful signal is whether they completed something where they went. Someone wanting to return after eight weeks has usually discovered the new role was mis-sold, which tells you about the other employer and nothing about them. Someone returning after eighteen months or two years has typically shipped something, built a real comparison and can say concretely what they want to do differently. The second conversation is far more useful than the first.
Should you rehire a developer at their old level or a higher one?
Benchmark against who they are now and what the Singapore market pays for that person, then state explicitly which it is and why. The failure is ambiguity rather than disappointment. A returner who believes they are coming back more senior, into a role that is actually identical, disengages inside a quarter. If the level is flat, say so and say what would change it and when. People accept a flat level readily when it is named and resent it enormously when they discover it.
Do returning developers need onboarding?
Yes, and skipping it is the most expensive mistake in this whole process. Codebases, deployment processes, team composition and decision rights all move in a year, and a returner is the one person who will not ask, because asking means admitting they do not already know. Give them the full programme: access on day one, a session on what changed while they were away, a named buddy who joined after they left, and a written thirty day plan. It costs a week and it saves a quarter.
Ready to work this channel properly?
Start the alumni list this week, then tell us which roles are open. We will run the open market search in parallel while you work the returners, so you are not choosing between the two. Where a former colleague is the right answer, we will tell you that and step back.
Start the conversation