You're not missing a growth hack. Something in your funnel is leaking and the dashboard isn't showing it, because the number the dashboard optimises for isn't the number in your bank account.
We find the leak first. Then we scale.
Live store, real revenue, growth stopped. Or growing, but the profit isn't. Usually somewhere between ₹20L and ₹2Cr a month, where the things that got you here have run out.
Platform ROAS ignores RTO, cancellations, discounts and the COD mix. Real CAC is often 20% higher than the dashboard says. Brands optimise the reported number and lose money faster.
A theme update, an app conflict, a consent banner. Events stop firing or fire twice. The algorithm learns on fiction and your CPMs climb for reasons nobody can explain.
The same three ads for four months, refreshed with new colours. Frequency climbs, CTR drops, CPM rises, and it reads as "Meta got expensive."
No retention flows, or flows written eighteen months ago. Every rupee of growth costs full price.
Six-second mobile load, a checkout asking for things it doesn't need, product pages that don't answer the two objections stopping the sale.
Around month eight of scaling, you've bought everyone your ads convert cheaply. Nothing bends that curve except people who already know your name.
Tracking rebuilt properly — CAPI, GA4, server-side. Contribution margin modelled per SKU after RTO. Full funnel walked from ad click to repeat purchase. Then we fix what's cheap and broken: pixel, checkout friction, mobile speed, abandoned cart flows.
This is where most of the early gain sits. It's unglamorous and it's usually worth more than the campaign work.
New creative built on customer insight, not the last winner recoloured. Audience segmentation that actually differs. Landing pages matched to the ad. Retention flows on email and WhatsApp. SEO and CRO running in the background.
Budget increases in increments that don't reset learning. Creative killed before fatigue, not after. Always one experiment live. Weekly numbers you can read without a translator.
We've run this at scale, not just launched something new and walked away. Ajay owned the direct channel P&L at Akasa Air — pushing it from a small share of overall sales to a meaningfully larger one over time. That's not a launch, that's grinding a plateau. At Apollo's e-pharmacy business, order volume was high enough that small conversion improvements moved real revenue, not vanity metrics.
Robin spent years at Orientbell turning a legacy tile brand into the most-visited site in its category — a slow, compounding fight, not a single big win.
Between us, we've also fixed plateaus that had nothing to do with traffic or creative. At NIIT, growth had stalled not because the ads were bad but because the sales team downstream couldn't convert leads fast enough — fixing it meant fixing what happened after the click, not before it. At IndiaMART, growth came from changing who we worked with — many small creators instead of a handful of big ones — not from spending more on the same channel.
Plateaus are what most of our careers have been. The fix is rarely "spend more." It's almost always something specific and unglamorous that a dashboard doesn't show.
Some plateaus aren't marketing problems. If the product has thin margin, a high return rate, or no repeat purchase, no amount of optimisation fixes it. We'd rather tell you that in the audit than take your money for six months.
And the first month may look worse. Fixing attribution usually reveals that your real CAC was higher than you thought. That's not a step backwards, but it feels like one.
Tracking and quick fixes inside 30 days. Real signal by week six to eight. Compounding after that.
Your call. It rarely works to run two.
Around ₹2L a month on media. Below that, the data is noise.
