Plush Reduced New Customer Acquisition Cost by 17% While Scaling Ad Spend by 15% — Here's the Optimisation Change That Made It Possible
17.39%
Drop in nCAC
New customer acquisition cost fell from ₹328.96 to ₹271.77 — a saving of ₹57.19 per new customer
15.68%
Increase In Ad Spend
Spend scaled up simultaneously — lower CAC achieved while spending more, not less
₹250
Kill nCAS Threshold Set
Any prospecting ad exceeding ₹250 nCAC was paused — training the algorithm to focus on new customer acquisition
2 Weeks
Time To Results
Meaningful nCAC improvement visible within 7 days; full results confirmed within 2 weeks
Context
Plush for Her is a women’s wellness brand offering sustainable period care, body hygiene, and skincare products. Founded in 2018 to disrupt a fem-care industry full of over-promises, the brand had built strong product credibility — but acquisition costs on Meta were climbing without a clear lever to pull them back down.
Adbuffs had been running Plush’s Meta account for 2 months with a custom server-side event infrastructure already in place — tracking new customers, nCAC, and new customer conversion value separately from blended returns. The data was there. The opportunity was in how we used it.
Objectives
- Reduce new customer acquisition cost (nCAC) on Meta without cutting spend — by changing what the algorithm was optimising for, rather than rebuilding the campaign structure from scratch.
What we did
- Identified that the account was optimising on blended CAC — which meant Meta’s algorithm was counting re-purchases from existing customers as successful conversions, diluting the new customer signal
- Switched the optimisation signal to nCAC, a custom conversion event that fires only when a genuinely new customer purchases — redirecting the algorithm’s learning toward new customer acquisition exclusively
- Set a hard kill threshold of ₹250 nCAC across all prospecting campaigns — any ad set exceeding this was paused immediately, removing manual lag from the feedback loop
- Gave the model 7 days to recalibrate before reviewing results, then monitored week-on-week nCAC as the algorithm stabilised around the new objective
₹328.96
nCAC before implementing the updated CAC signal
AFTER
₹271.77
nCAC after 2 weeks — ₹57.19 saved per new customer,
with 15.68% more spend running at the same time
Key Learnings
Blended CAC hides a growth problem — nCAC reveals it
Plush’s blended CAC looked healthy because returning customers were being counted as conversions. Once we isolated nCAC as the metric and the optimisation signal, the real cost of acquiring new customers became visible — and actionable. Brands that optimise on blended CAC are often unknowingly paying to re-acquire customers they already have. Switching to nCAC optimisation reduced Plush’s new customer acquisition cost from ₹328.96 to ₹271.77 in two weeks, without touching the creative or audience structure.
A hard kill threshold is more reliable than human judgment at scale
The ₹250 nCAC kill rule meant underperforming ad sets were paused the moment they crossed the threshold — not when someone got around to reviewing the account. That speed matters. Every hour an inefficient ad set runs, it pulls the algorithm’s learning in the wrong direction. Automating the cut kept the model under constant pressure to find new-customer audiences that actually performed, and it did.
Scaling spend and lowering acquisition cost can happen simultaneously — when the signal is precise
The standard assumption is that nCAC rises as you scale, because you exhaust efficient audiences first. For Plush, the opposite happened: spend increased by 15.68% while nCAC fell by 17.39%. The reason was signal quality — the additional budget was directed at high-intent new customers, not cheap retargeting. Fixing the optimisation objective before scaling is what made the difference.
Creatives used for our Clients
Want results like these for your brand?