I used to believe retention was the reward for being inspiring. I held town halls, painted big visions, and shared cheerful updates. People clapped and posted emojis. Then two of our best engineers resigned in the same week and a third asked for a reference. The Slack channel went quiet in a way that felt personal. It was not just churn. It was trust thinning to transparency.
From the outside we looked like momentum. New logos. Fresh capital. Headcount that made our deck look mature. Inside, our pace outran our systems. Roles overlapped. Escalations were unclear. Feedback surfaced only when something broke. Turnover was not a random storm. It was a bill for design debt, issued late and paid in talent.
Departures rarely happen all at once. People leave a company one decision at a time. A promotion that took half a year to confirm. A meeting where a junior PM was cut off and no one intervened. A weekend fire drill that slid from rare event to routine. Each moment was small on its own. Together they formed a path to the door.
The first pattern I missed was calendar fatigue. High performers spent their weeks explaining work rather than doing it. Every handoff required a meeting because ownership was cloudy. When people do not feel true ownership, they will not fight to keep what they cannot claim. The second pattern was uneven standards. We praised speed in Kuala Lumpur yet punished it in Riyadh where partners needed more process. People began optimizing for survival rather than outcomes. When energy goes into reading the room, attention leaves the customer. Pride in the work drains out, and the idea of being replaced starts to feel like relief. The third pattern was kindness without clarity. We softened hard conversations until course corrections sounded optional. Underperformance lingered. High performers carried the gap and then burned out. In the moment, softening felt humane. It was not. It was avoidance dressed as empathy.
The sentence that woke me up did not come from a survey. It came in a resignation call. I like the mission, but I do not know how to win here. People do not need constant applause. They need visible ways to win and fair ways to lose. If neither is clear, they leave.
What followed was not a miracle tactic. It was a series of dull but decisive choices. We started by mapping ownership like an engineer would map a system. Not a glossy job description, but a living diagram of recurring decisions and who was accountable for each one. Contributors were listed, escalation paths were explicit, and overlapping claims were treated as leadership bugs to fix at once. Within two sprints the meeting load dropped because people stopped relitigating authority. Work started to move the way water moves when you clear the channel.
We then wrote a standards charter that could travel across markets without shrinking expectations. Good looked the same in Singapore, Malaysia, and KSA, but the rituals flexed. In KSA we added pre mortems to honor partner pace. In Malaysia we shortened planning cycles to honor vendor reality. The team felt seen, and quality stayed consistent. Predictability began to replace guesswork.
Reflection became scheduled, not crisis driven. Each team had a fixed hour for honest post mortems with two rules that never changed. Speak to the work, not the person. Leave with one decision, not ten tasks. We stopped waiting for fires to learn how to handle heat. The point was not to be nice. The point was to make learning a cost of doing business that we paid on time.
We removed mystery from rewards. Bands were not enough. Criteria mattered more. We wrote what merited a raise, an equity refresh, or no change. We shared sample cases so a marketer in Penang could see how a peer in Dubai progressed. Once people understood the rules of the game, resentment dropped. High performers tolerate hard paths. They do not tolerate hidden ones.
Access to leadership changed shape. I blocked weekly office hours and trained managers to run the same practice two months later. Anyone could book fifteen minutes to escalate or clarify. The one boundary was simple. Come with the decision you would make if I were unreachable. The habit produced better thinking and fewer emergencies that were really updates in disguise.
Leaders sometimes ask for the secret to retaining people. There is no trick. Retention is the result of design. Start by writing what you expect by role in a way a new hire can grasp on day three. Train managers to give specific feedback within two days of an issue rather than saving it for quarter end. Track time to first independent win during onboarding, not only time to hire. Run exit interviews like you paid for the truth, because in every sense you are paying for it.
In talent constrained markets, it is tempting to counteroffer and move on. That may buy a month and sell out your culture. If you pay to keep someone, also pay with attention to fix why they were ready to leave. Most exits are not about money alone. People leave for meaning, mastery, and fairness. Those are leadership variables, not market destiny.
Across Southeast Asia, I see another recurring confusion. Founders use family language to signal warmth and loyalty. Early on it feels generous. Later it becomes pressure, and it is often invoked to justify missed reviews or shifting goalposts. Replace family with team. Teams have roles, seasons, standards, and a scoreboard. That frame is healthier and scales across culture and time zone.
Context also matters. In Saudi Arabia, national growth pulls talent upward at high speed. Promotions come early, so coaching must arrive even earlier. If you elevate people without a manager development plan, turnover will look like ambition when it is more often exhaustion. In Malaysia, many strong operators have been burned by chaotic startups and place a premium on predictable rhythms. In Singapore, mobility is high and brand signal carries weight. If you cannot offer stability or prestige, you must offer clear growth and responsibility. Show the path in the first month, not as a promise but as a plan.
I have not mentioned perks because perks do not close exits. Clear ladders, honest feedback, fair rewards, and sane calendars do. When you can sketch your org design on a napkin and your team can repeat it without you in the room, you are ready to measure retention again. Until then, departures are data. Listen.
If you are staring at a quarter that already includes three resignations, take one hard action now. Run stay interviews with your top five people. Ask what would make them leave in the next ninety days. Do not argue. Do not sell. Map the risks. Then fix one or two structural issues that surface. Maybe it is a messy handoff between sales and delivery. Maybe it is a manager who has never been trained to manage. Maybe it is you saying yes to everything and pushing chaos downhill. Treat each insight like a product bug. Patch now. Re architect next sprint.
If I were starting again, I would ship the ownership map before the product. I would train managers before raising the next round. I would set compensation clarity before hiring ten more people. I would never call a company a family if I intended to run it with standards. A culture that depends on the founder’s constant presence is not culture. It is dependency. Build something sturdier than that. People stay when the game is fair, the rules are visible, and the work produces pride they can feel at the end of the week. Address high employee turnover by designing such a game, and the scoreboard will take care of itself.
WhatsApp
Instagram
Threads
