Three weeks on a cold Meta ad account, a DTC skincare brand, roughly a $6 average order value, and not one purchase. Been there. Same dashboard refreshing, same reports, convinced the algorithm personally hates me.

Here's what I've learned running this play on low-AOV Indian DTC brands, plus what a few people in this space confirmed when I asked. Four questions, four answers.

Quick context first: beauty and skincare, around ₹600 AOV, brand new ad account and pixel. Converting fine on Amazon and offline, with repeat orders coming in. Meta is the only channel doing nothing.

1. How long does the ramp actually take?

No universal number exists, but what matters is roughly 50 purchases per week inside an ad set before delivery stabilizes. On a fresh pixel you're starting from zero. At a $6 AOV that bar is cheap to clear, which is the good news here.

The bigger tell is your Amazon and offline numbers. The product converts, the buyer exists, repeats are happening. So what Meta is asking someone to do is buy a ₹600 item from a website they've never heard of, when Amazon has the delivery and return policy they already trust. For a new India DTC brand at this price, that's usually the real conversion problem, not the ad account.

Two things to check before writing Meta off. Is cash on delivery enabled at checkout? A huge share of first purchases at this price point in India will not go prepaid on an unknown site. And run a Meta to Amazon test using Amazon Attribution links. Same ads, different destination. If they convert there, your ads work and your site is the gap.

2. What's the minimum daily budget?

Work backward from target CPA, not AOV. Say you can live with ₹250 per purchase. Fifty a week at ₹250 is ₹12,500, so about ₹1,800 a day into a single ad set gets you stable delivery inside a week. Smaller than most people assume, precisely because your AOV is small.

What kills low-AOV accounts isn't underfunding. It's splitting that ₹1,800 across three ad sets and five campaigns so nothing ever sees 50 purchases. One campaign, one ad set, broad targeting, all creatives inside it until you have a winner.

Then upload your Amazon and offline customer list as a custom audience and feed it into Advantage+ audience as a suggestion. On a cold pixel that's the one input that reliably beats open targeting, and it's the thing most new accounts don't have sitting around.

3. Why don't Meta sessions line up with Shopify sessions?

Meta doesn't report sessions at all, so you're actually comparing link clicks or landing page views against Shopify sessions. Shopify only credits a session to Facebook when the URL carried UTM parameters. Instagram's in-app browser strips the referrer, so a chunk of your Meta traffic lands in Shopify as "direct." Clarity fires on page load and ignores attribution entirely, which is why it tracks so much closer to Meta.

Fix: add UTMs at the ad level, compare Meta landing page views instead of link clicks, and accept that Shopify will still read low. What's left over is people who clicked and bounced before the page finished rendering. On Indian mobile connections, run a speed test on your product page. Those five minutes pay for themselves.

Someone in the thread made a sharp point on this. If Clarity matches Meta while Shopify lags, Shopify is undercounting, not Meta overreporting. Shopify strips bot traffic and only counts sessions that complete a full page load after the redirect. Trust Clarity and Meta for delivery diagnostics, use Shopify for settled orders.

4. When do you pull the plug?

Per ad, not per campaign. Nothing gets judged in its first three days.

After that: an ad that has spent 2x to 2.5x your target CPA with zero purchases is dead. At ₹250 that's roughly ₹600 burned with nothing to show for it. An ad with purchases but sitting at 3x target or worse past that floor is also dead. Everything else gets two weeks, and at two weeks you force the call either way. The ones that trickle along just under break-even are the ones that quietly cost the most.

And if every ad trips the zero-purchase rule while add to carts keep rolling in, stop killing ads and go back to question one. That's the checkout, not the creative.

One last thing. COD being most of the game for unknown sites in India is not an exaggeration. People will not go prepaid at that price point from a brand they've never seen before. If checkout is the bottleneck, no amount of creative testing fixes it.