Where to Spend the Next Growth Dollar

Most growth budgets get split the way you’d cut a cake at a kid’s birthday — everyone gets a slice so nobody cries. Region A gets its share, Region B gets its share, the new market gets a hopeful sliver. It feels fair. It feels safe. It is neither.

Even allocation is what you do when you haven’t decided. You’re paying to keep options open instead of paying to win somewhere. The next dollar isn’t a vote you owe each team — it’s a bet, and bets have a single best home. The job is to find that one place, not to be polite to the org chart.

So I stopped asking «how do we divide this?» and started asking «which one question, answered, would tell me where the dollar goes?» That turns budget into a prioritization problem. And prioritization problems have decision trees.

The first node is not «which unit is biggest». Big is often just old. The first node is share of the local audience — penetration. A unit reaching a high fraction of the people it could plausibly serve has proven the offer works there. A unit at a low fraction in a bigger city has noise, not signal.

High penetration tells you the market said yes. That’s the cheapest information you’ll ever get, because the market paid for it. So I rank candidates by local share first. The leader on penetration earns the right to be considered — not the leader on raw volume, who may simply have been around longer and is now coasting on inertia.

Penetration usually leaves you with two or three contenders that look close. Now you walk the funnel, one step at a time, and let it break the tie for you.

First tie-break: active → power-user conversion. Lots of people show up — fine. Do they cross into real, habitual usage? A market that turns actives into power users has a working core loop; a dollar there compounds. The one that can’t is a leaky bucket, and I won’t pour into a bucket.

Still tied? Second tie-break: power-user → payment conversion. This is where engagement becomes P&L. Power users who don’t convert to paying tell you the value is real but the capture is broken — a different, slower fix. Power users who pay tell you the whole chain holds. Penetration earns the look, active→power decides the short list, power→pay names the winner.

Here’s the counter-intuitive move. Once the tree points you at one strong market, don’t automatically pour into its dominant unit. Look inside that market and feed the weaker one.

The dominant unit is dominant because it’s near-saturated — you’ve already taken the easy share, and every further point costs more than the last. The weaker unit in the same proven market has the same demand conditions but real headroom left. The marginal dollar buys more growth there precisely because nobody’s wrung it out yet.

It feels wrong to starve your winner and back your laggard. But you’re not rewarding the laggard — you’re buying the cheapest remaining growth, and the cheapest growth lives where the headroom is, not where the trophy is. Spend on the gap, not the glory.