Cutting blended acquisition cost by fixing the product page, not the ad account
Aurelia Home was scaling paid acquisition into a product page that converted poorly, and reduced blended customer acquisition cost by rebuilding the page, the delivery-promise messaging and the repeat-purchase flow rather than by changing media allocation.
Measured outcomes
- Blended customer acquisition cost
- −38%
- Month 10 vs month 1, at comparable spend
- Product page conversion rate
- +2.4x
- Rebuilt template vs control, 6-week split test
- Checkout abandonment at delivery step
- −44%
- After pincode-accurate estimates shipped
- 12-month repeat purchase rate
- 11% → 29%
- Cohort measured at month 10
The situation
Aurelia arrived asking for a paid media audit. Spend had tripled year on year while revenue had grown by under half, and the assumption was that the media buying had degraded.
The media buying was competent. The problem was downstream: the product detail page converted at roughly a third of the category norm, and the checkout abandoned heavily at the delivery step, where a national shipping estimate appeared that was wrong for most of the country.
Repeat purchase was almost nonexistent for a homeware brand, at eleven percent within twelve months. Every rupee of growth was being bought again from scratch, which is what made the acquisition cost unsustainable rather than merely high.
What we did
Redirect the engagement
We reported at the end of the diagnostic that a media audit would not fix this, and proposed a different scope than the one we had been approached for. That is a smaller engagement than the one on the table and it was the correct recommendation.
Rebuild the product page around the actual objection
Session recordings and exit surveys converged on the same two questions: what does this look like in a real room, and when will it arrive. We rebuilt around scale context, material detail and an honest delivery estimate computed from the visitor's pincode rather than a national average.
Fix the delivery promise properly
The estimate was wrong because it was optimistic. Correcting it made the displayed date worse for most of the country and improved both conversion and post-purchase sentiment, because the previous number generated support contacts and cancellations when it was missed.
Build the repeat-purchase motion
Segmented lifecycle flows by room and by purchase stage, replacing an undifferentiated monthly newsletter. Homeware has a natural repeat rhythm and Aurelia was not using it.
Only then, revisit media
With conversion and repeat behaviour fixed, we rebalanced spend toward the channels that had looked weak under the old economics. Several were fine all along; they had been judged on a broken denominator.
What we would do differently
We should have pushed harder, earlier, on the delivery estimate. It was identified in week two and took until month four to ship because it needed logistics data nobody owned. In hindsight the sequencing argument we made was right but we made it too politely.
“We asked for an ad audit and were told the ads were not the problem. They talked us into a smaller project than the one we were ready to sign.”