Somewhere around 4:40 pm on Day 1, one of our functional heads stopped mid-sentence.
He had been defending a format for ninety seconds. Real distribution. A team that liked it. Then he looked at the scorecard we had handed every leader that morning — four columns, one to five — and said: "It's a two on working capital. A two on pull. It doesn't belong."
Nobody applauded. Someone added it to the kill list. We moved on.
I have sat in a lot of planning rooms. That was the best ninety seconds of my year.
The problem we walked in with
Indian D2C (direct-to-consumer) brands turning into FMCG (fast-moving consumer goods) companies have a graveyard between ₹150 crore and ₹400 crore. Not because founders stop working hard, but because the model that got you to fifty is hostile to reaching two hundred.
At fifty crore you win on hustle. Somebody drives to the plant at midnight. Heroics are a feature. At two hundred crore they are the failure mode — they hide process debt, wreck forecasting, and burn your best people.
The AOP is the biggest culture-setting event in any company's calendar, and most HR leaders are not in the room for it. We arrive afterwards, turn someone else's numbers into goals, then spend a year managing decisions we never shaped. So we designed ours as an OD intervention, built on Jim Collins' Good to Great.
Day 1: killing things we were proud of
Ninety minutes to isolate the core. Collins' Hedgehog Concept: sustained returns live only where deep passion, what you can genuinely be best at, and your economic engine overlap.
Anything outside that overlap comes out of the plan. Not deprioritised. Removed. And we gave the room a number that made people shift in their chairs: kill 20–30% of what we do today.
The instrument was a four-line scorecard. ROIC (return on invested capital) potential. Capability advantage. Consumer pull. Working capital intensity. Score one to five. That's all — and it quietly moves the question from whose idea is this to what does this score.
Take the first item on the kill list: expanding our kiosk format.
On paper, unarguable. Direct consumer contact. Visible brand. Revenue we controlled end to end. A source of real pride. Then it met the scorecard. Capability advantage, two — our edge is sourcing and manufacturing, not retail labour across scattered locations. Working capital, two — every kiosk locks up deposit, fit-out and stock before it sells anything. Pull, three at best, because footfall is not demand.
Nobody in that room disliked kiosks. What survived was narrower — premium-format EBOs (exclusive brand outlets) only, no general expansion.
The rest followed. Kill mints. Kill NPD (new product development) on a line we loved. Double down on fresh sweets, fusion sweets, premium formats in categories we already understand.
Killing something you built is grief work. Anyone who says otherwise has never sat in the room while it happens.
Day 1, later: The Flywheel
Three groups. Thirty-five minutes. Find four to six operational actions that, in sequence, compound return on capital. Then the constraint that changed everything. Do not list corporate goals. "Hire better people" is not a step. Map a loop where A forces B to become cheaper. Hope is not a transition strategy.
Every step had to answer three questions. What improves, by how much, in what time. Fewer vendors, packaging cost down 10%, COGS (cost of goods sold) down 3% — that is a step. "Better sourcing improves margins" is a sentiment.
Day 2: the corridor, and an order that offended nobody
Collins' 20-Mile March rejects growth spikes in favour of paced execution — growth that outruns delivery destroys value. So we built a corridor — a floor we will not fall below, a ceiling we will not chase.
Then we built the P&L (profit and loss) live, in a fixed order. Cost of goods, Sales, Operations, Marketing, Finance, HR. HR going last is not an insult. Cost of goods is the physics of this business; everything downstream inherits its constraints. I would rather build a manpower plan against a locked gross margin than an aspiration.
We listed our leakages out loud, channel by channel — expiry, sampling at 1.4%, returns, packaging churn, forecast error showing up as excess stock. Nobody was blamed. That was deliberate, and it is why the list was honest.
The last session ran forty-five minutes and was the part I care most about. Name the behaviours that must become non-negotiable to hit these numbers. Then the slide that earned its keep: behaviour without mechanism is waste.
Values on a wall are decoration. A behaviour is real only when someone can say how it is measured and where it shows up. So each carried a mechanism. Expiry discipline became a monthly closure. Transport cost control became a hard cap on rail deliveries. Margin over vanity scale became a rule that no cost increase passes unchecked.
What I was actually watching
While the strategy ran, I ran a quieter second track. I watched who spoke first after each brief, because the first voice sets the room's permission level. I watched where defensiveness clustered, because resistance is data, not obstruction. I watched the silences and made myself leave them alone.
Mostly I watched for the moment the group stopped performing for the CEO and began arguing with each other. That is when you have a leadership team instead of an audience. Vinay let it happen. Not every founder does.
The homework nobody talks about
The two days are the easy part. The pre-work decides them.
Get the numbers argued about beforehand. Circulate contribution margin by SKU (stock-keeping unit) and channel, plus fill rates, ten days early. A workshop spent disputing a number is a workshop lost.
Run a DISC read on the room before you design it. DISC maps behaviour across four traits — Dominance, Influence, Steadiness, Conscientiousness. High-D leaders anchor the room's first position, so silent scoring must precede open debate. High-I leaders sell a new idea beautifully, which is exactly why kill decisions need a scorecard, not a pitch. High-S leaders withhold dissent to keep the peace, so ask them by name before the room converges.
Pre-mortem your leaders one at a time. I spoke to each before Day 1 — not to pre-negotiate outcomes, but to learn what they were privately afraid of losing. You cannot facilitate an attachment you did not know existed.
Draft the follow-through before you walk in. Owner, mechanism, review cadence, first review date. The half-life of workshop energy is about eleven days.
Why I am writing this down
Most people don't leave over pay. They leave because the work they poured a year into was quietly abandoned in month seven, and nobody told them why.
We killed a quarter of our plan in two days, in front of each other, with the reasoning said out loud. Everyone in that room knows what we are doing this year, and what we are not.
That is the job. If that sounds like a room you would want to be in, we are building more of them at GO DESi.
