Performance Max has been live in Indian ad accounts for a while now, but most D2C brands I audit are still running it the way they ran old Smart Shopping campaigns — and leaving 20-30% of efficiency on the table because of it. After running PMax across nine D2C accounts over the last two quarters, here’s what’s actually moving the needle in 2026, and where I see brands waste budget every single time.
What Performance Max Actually Is (And Why Indian Brands Misuse It)
PMax is a single campaign that bids across Search, Display, YouTube, Discover, Gmail, and Maps using Google’s own signal-matching instead of manual placement control. The mistake I see most often: brands treat it like “set it and forget it” because Google’s onboarding flow makes it feel that way. In reality, the campaigns that perform — averaging 4.2X ROAS in my accounts versus 2.6X for the ones running on autopilot — are the ones where the advertiser feeds the algorithm constantly, not the ones that walk away after launch.
The Asset Group Mistake Almost Every Brand Makes
Most Indian D2C brands launch PMax with a single asset group covering their entire catalog. I split by margin tier and customer intent instead — typically 3 to 4 asset groups per account. For a Mumbai-based home fragrance brand, splitting from 1 asset group into 4 (hero SKUs, mid-margin bundles, low-margin entry products, and a seasonal/gifting group) took blended ROAS from 2.9X to 4.6X within 6 weeks, simply because the algorithm could optimize toward distinct intent signals instead of averaging across a mixed catalog.
Budget Allocation: How Much Should Go to PMax vs Search
For accounts spending ₹1.5L–₹4L/month, I generally cap PMax at 50-60% of total Google Ads budget, keeping the rest in Branded Search and a thin non-branded Search layer. Going above 70% into PMax for newer accounts (under 6 months of conversion data) consistently produces noisier results — I’ve seen ROAS swing as much as 35% week-to-week when an account leans too heavily on PMax before it has at least 30-50 conversions per asset group to stabilize on.
Audience Signals — The Most Underused Lever
Audience signals don’t restrict targeting in PMax, but they massively speed up how fast the algorithm finds the right buyers. Feeding in a first-party customer list (even a modest one — 2,000-3,000 purchasers) alongside category-relevant in-market segments cut the typical “learning phase drag” from around 18 days down to 9-10 days in three separate accounts I manage. For Indian brands with thin first-party data, even a Google Analytics 4 high-intent audience (cart abandoners, multi-page-view sessions) works as a usable substitute signal.
Real Results: A 45-Day PMax Test for a Skincare Brand
One D2C skincare client came in spending ₹80,000/month on a single-asset-group PMax campaign at 2.1X ROAS. Over 45 days we restructured into 3 asset groups by product category, added a 4,500-contact first-party audience signal, and excluded low-margin SKUs from the feed entirely. Results: spend grew to ₹1.35L/month, ROAS climbed to 3.8X, and cost per purchase dropped from ₹612 to ₹398. The single biggest driver wasn’t a creative change — it was the asset group restructure combined with feed-level margin exclusions.
When PMax Doesn’t Work (Be Honest)
PMax struggles for brands with fewer than 15-20 monthly conversions — there simply isn’t enough signal for the algorithm to learn from, and I’ve seen these accounts burn through budget with ROAS under 1.5X for months. It also underperforms for high-ticket items (₹15,000+) where the buying cycle is long and assisted conversions matter more than last-click PMax reporting shows. In both cases, I move budget back toward Search and Meta retargeting until volume catches up.
PMax isn’t magic, and it isn’t useless either — it rewards the accounts willing to actively manage asset groups, feed quality, and audience signals, and punishes the ones that set a budget and check back a month later.
Running ads and not getting results? Book a free 30-minute strategy call — I’ll audit your account for free.
