You're at Rs.1,000 a day and it's profitable. Now you want more. There are two ways to add spend, and they fail for different reasons. Here's how to use both without torching your margins.
Vertical scaling: raise the budget on what works
You take the campaign that's already profitable and increase its daily budget. That's it. Vertical means pushing more money into the same thing.
It's the fastest and it's the cleanest because you don't introduce new variables. Same campaign, same ads, same audience, more money. Whatever the account learned in the last month keeps compounding, because the delivery system is already calibrated around what converts. That's why vertical is where scaling should always start. New variables are risk; more budget on a proven setup is mostly just more budget.
But there's a limit. Each budget increase pushes Meta to reach beyond your cheapest conversions into colder audiences, so efficiency drifts down as you scale. That's normal, not a sign of failure. The question is whether the return stays above your break-even after each jump.
Go in steps of about 20 to 30 percent, not doubling overnight. Big jumps reset the learning phase and you pay for that reset in ROAS. Small steps, spaced a few days apart, let the campaign absorb the budget and settle before the next push.
Practically, from Rs.1,000 a day: Rs.1,250, hold three or four days, Rs.1,550, hold, Rs.1,900, hold. Within a month you're at Rs.2,500 a day with the learning intact, which beats the founder who jumped straight to Rs.5,000, spent a week in learning noise, panicked, cut back, and restarted from a worse position than where they began.
Watch the marginal return on each step. If the Rs.1,900 step returns above your break-even, take the next step. When a step's return dips below, hold where you are, refresh the creative inside the campaign, and try sequencing another step in a couple of weeks rather than forcing it.
Horizontal scaling: add new campaigns and angles
You keep existing campaigns running and add new ones. A new audience, a new creative concept, a new geographic market, a new funnel stage. Horizontal means getting more spend by adding more lanes rather than making one lane bigger.
It's how you get past the ceiling that vertical hits. When one campaign's efficiency starts slipping, a second and third campaign running in parallel let you spend more while each stays efficient. The lanes deliver independently, so a saturated pocket in lane one doesn't drag lane two down.
Why it works, briefly: your single campaign is limited by the part of the market it serves. A different creative angle speaks to a different reason to buy, which means a different slice of people. A new geography is a fresh pool, and if you want to scale internationally, this is the whole method. Horizontal spend buys reach that vertical spend can't, because there is a cap on how much budget one campaign can absorb while staying efficient, and every account hits it eventually.
The cost is complexity. More campaigns means more to manage and more chances to fragment your budget so thin that nothing learns. Add three lanes to a Rs.5,000/day account at Rs.500 each and you've built three learning-phase prisoners while your main campaign goes quiet. That's decoration, plain and simple. It isn't scaling.
The rule: never add a new campaign unless it has its own real budget to learn. Whatever your per-campaign threshold is, the new lane needs its own, not a leftover. A new campaign on a leftover budget is just another thing stuck in learning.
How to combine them
Start vertical. Push the profitable campaign up in steps until marginal efficiency drops toward break-even. Then switch to horizontal, add a new lane, and give it enough budget to work. Once that lane is profitable, go vertical on it too.
Keep adding lanes, keep pushing each one up, and kill whatever falls below break-even. A rhythm worth copying: one new lane tested per month, funded properly, judged on the same break-even as everything else. Keep it if it clears the line. Kill it without ceremony if it doesn't, and put next month's lane money on what survived.
That's how a Rs.1,000/day account becomes a Rs.10,000/day account without the whole thing collapsing when one campaign tires out. The lanes fail independently, and the ones that work carry the growth while you replace the rest.
"Can't I just jump the budget once and be done with it?"
The shortcut everyone tries. From the outside it sounds reasonable: the campaign is profitable, so why pay the small tax three times, why not take the whole step in one move?
Because the jump pays for two things at once: the colder reach and the reset. Take the numbers. Your Rs.1,000-a-day campaign returns 4x. You double it to Rs.2,000 and hold the steps, and the second Rs.1,000 delivers maybe 3x, so the mixed week lands near 3.5x. Now jump instead, doubling straight to Rs.2,000 overnight, and the reset drags the whole budget, not just the new half, into noisy delivery. The same second Rs.1,000 doesn't come in at 3x, it comes in at 2x with the original Rs.1,000 also down at 3x. Same money in, visibly less out, and you can't even tell whether the idea stalled or the delivery did, because the changed variable was never isolated. The stepped method keeps the experiment readable at every step, which is why it's the only version of scaling I trust on accounts I'm paid to look after.
I'm Hirdesh Matai, founder of HDM GRO-ads, the science of e-commerce advertising.