Halving cost per result sounds like a creative breakthrough. It usually is not.
On the campaign below, cost per result fell 51% — ₹175.61 ≈ $1.85to ₹85.48≈ $0.9. Here is what each factor actually contributed.
Baseline campaign against the scaled Advantage+ campaign, after Conversion API was added alongside the browser Pixel. A 51% reduction across the 60-day period.
The four things that could have caused it
When cost per result drops sharply, there are only four real explanations: better creative, better audience, better landing page, or better data. Everything else is a variation on these.
Most agencies assume creative. In this case, creative mattered least.
Data quality: the largest single factor
The account had Meta Pixel installed and no Conversion API.
Browser-side tracking alone loses a substantial share of conversions to iOS privacy restrictions and ad blockers. The platform then optimises toward the people it can see converting, which is a biased sample of the people who actually convert.
Once server-side tracking ran alongside the Pixel — with events on donation initiated, completed, and thank-you view — the algorithm had a materially more complete picture. It stopped optimising toward a distorted version of the buyer.
This is unglamorous, invisible in any report, and produced more improvement than anything else we did.
Audience: identified through paid tests, not guessing
Three campaigns ran before the winner: a baseline at ₹175.61≈ $1.85, an audience test at ₹193.07≈ $2.04, and a creative test at ₹242.34≈ $2.56.
Read that sequence carefully. Costs went up before they went down. Those middle campaigns were not failures; they were purchased information. The ₹242.34 ≈ $2.56campaign told us which creative direction did not resonate, which narrowed what the scaled campaign should use.
Structured testing produces compounding efficiency. Random testing produces random results and a confused algorithm.
Landing page: removing steps, not adding persuasion
The original page was informational — built to explain, not to convert. Every additional step between intent and completion costs conversion rate, and paid traffic makes that cost recurring.
We shortened the form, moved credibility signals above the ask, and made the call to action unmissable. No new copywriting genius. Just fewer places to lose people.
Creative: specificity over sentiment
The creative change was directional rather than dramatic: generic appeals replaced with specific stories, numbers and outcomes.
It helped. It was not the primary cause, and claiming otherwise would be the flattering version of the story rather than the accurate one.
Budget structure
Campaign budget optimisation let the platform allocate dynamically instead of us guessing. One ad ended up absorbing roughly 90% of spend because it genuinely outperformed the others on every metric.
Manual budget shifting would have reached the same conclusion slower, and by then the auction conditions would have changed.
How to apply this to your account
In order of expected impact:
- Check whether Conversion API is live. Not "the Pixel is installed" — server-side, with mapped events. This is the highest-leverage fix in most Indian ad accounts right now.
- Look at what event you are optimising for. If it is link clicks or landing page views, you are asking the platform to find clickers, and it will.
- Time your landing page on mobile data. Not on your office wifi on a laptop.
- Design your next three tests as a sequence, where each one narrows the next decision.
- Only then rewrite creative.
Most accounts we audit have a data problem wearing a creative problem's clothing. The team keeps producing new ads to fix something that new ads cannot fix.
If you want the tracking layer checked properly, that is what our free growth audit starts with.