Start with a disclosure
Feastale is a sister company founded by our own team. We ran its Meta ads, so this is our own experiment, not a client result presented from the outside. All numbers below are a snapshot of July and August 2026, taken from live pulls of the Meta Graph API by the Feastale team. They are not promises about what your ads will do.
One small budget, two audiences
Two Meta campaigns shared a combined cap of ₹200 a day, ₹100 each. "Kerala Lead Gen" targeted consumers in Kerala only. "India - Partner Recruitment B2B" looked for resellers and distributors. A small cap forces a useful discipline: you cannot afford to run weak ads for long.
Test against the best ad, not against a guess
We paused the original ad for about 68 hours so that two challengers got uncontested delivery. The pause was meant to give the new ads a fair look.
We read the result after about 47 hours. The "Onam Unboxing" ad had an 11 to 12% click-through rate and a cost per lead of ₹8.72, against ₹15.49 lifetime for the original. That rests on a real but small sample of 7 leads. The bilingual "Gift Box Reel" got only 26 impressions and no leads, so we paused it. The original was switched back on.
An early signal is not a result
On the first day of the B2B campaign (29 to 30 July 2026) we spent ₹42.73, saw a 7.59% click-through rate and ₹7.12 per lead. Six distinct leads came from Kerala, Tamil Nadu, Karnataka, Gujarat and Rajasthan. We described it as an early signal that might be inflated, and treated it that way.
When leads come in but are the wrong ones
By 6 August 2026 we had diagnosed off-target B2B leads. Both campaigns shared one two-field form (name and phone) with no qualifying question, and Advantage+ audience expansion was switched on. The fixes were new forms with Meta's quality-review and non-targeted-viewer blocking flags, a qualifying question on the B2B form, Advantage+ audience off for B2B, and a fresh creative for the starved Kerala ad. A cheap lead is only useful if it is the right person.
A brake that works without us
We also set a standing automatic stop-loss: every ad pauses when lifetime account spend reaches a cap. It fired correctly at ₹2,000 on 8 August 2026, and the cap was later raised to ₹3,000.
What to take from it
- Pause the old ad when you test, so a challenger gets a fair run.
- Read small samples as hints, and say how small they are.
- Add a qualifying question to any form where wrong-fit leads cost you time.
- Check audience-expansion settings before blaming the creative.
- Set a spending cap that stops ads on its own.
- Treat a lead as a lead, not a sale.
Polaroid Dosa is a creative and digital marketing agency in Palakkad, Kerala, working with businesses worldwide on social media, ads, branding, films and websites.