Performance Max Asset Reporting in 2026 — What Indian Brands Should Pull, Pause and Promote

Operator take: Asset-level reporting in PMax is finally giving you something to act on. Run the 14-day audit cadence. Don’t pause “Low” assets in week one — let the model learn. By week three, the data is clean enough to be ruthless.

Performance Max was a black box for two years. Marketers couldn’t see which assets the model was promoting, which were sitting unused, and which were quietly burning budget.

That changed in early 2025 when Google rolled out asset-group-level reporting with rating signals. By the start of 2026, with the latest interface update, Indian advertisers finally have the data needed to manage PMax campaigns the way they’d manage a regular ad set: deliberately.

This is the audit framework our team in HSR Layout runs across the 30+ Indian PMax accounts we manage. It’s tested across e-commerce (Meesho-grade D2C, premium fashion), real estate, financial services, and travel. The cadence is the same; the thresholds shift slightly by category.

The four signals worth reading every week

Inside Asset Group → Insights → Asset Performance, you have four indicators per asset:

• Performance rating — Best, Good, Low, Pending
• Combinations served — how many ad combinations actually used this asset
• Impressions — visible reach
• Conversions attributed (when available)

“Pending” is the most misunderstood. A Pending rating means the asset hasn’t yet served enough volume for the model to evaluate it — not that it’s weak. Premature judgement here costs accounts thousands of rupees.

The 14-day audit cadence

Most agencies audit PMax monthly or react when a number drops. Both approaches are wrong. Monthly is too slow to course-correct; reactive is too late.

We run a 14-day cycle. Here’s exactly what happens at each touchpoint:

The pull / pause / promote decision tree

This is the ruleset we apply at each audit:

Pull (delete from asset group):

• Disapproved assets that can’t be re-edited (policy issues, brand-name conflicts)
• “Low” rated assets after 14+ days with >1,000 impressions
• Assets generating zero combinations served after 21 days

Pause (keep in library, exclude from active group):

• “Low” rated assets that have served <500 impressions — give them another 7 days first
• Seasonal creative outside its window
• Long-form video that’s underperforming short cuts of the same content

Promote (replicate, expand):

• “Best” rated headlines — add 2–3 close variations within 48 hours
• “Best” rated images — re-use the visual concept across square + landscape + portrait crops
• “Best” rated videos — produce a 6-second cut + a 30-second cut from the same asset

The mistake every Indian PMax account makes in week one

Pruning too early.

An asset rated “Low” on Day 7 isn’t actually low — it’s under-tested. The model needs roughly 5,000 impressions per asset to reach confident rating, and that’s 10–14 days of normal pacing for most Indian brands at ₹40k–80k/month per asset group budgets.

If you pause “Low” rated assets on Day 7, you’re not improving the campaign — you’re depriving the model of the variety it needs to find combinations that work. We’ve seen accounts plateau at 1.8× ROAS for months because of impatient asset-management.

The cleanest principle: more variety, not less, in the first 14 days. If anything, add 5 more headlines and 5 more images on Day 7 instead of pausing weak ones.

What changed in 2026: the audience signal

The bigger story this year is that asset-level reporting now shows you which audience signals are firing for which assets. That changes the playbook in two ways:

1. You can now align creative with audience. If your “lifestyle imagery” assets perform on lookalike audiences but flop on demographic-only signals, you have evidence to brief the next creative round around the audience that actually responds.
2. You can identify dead audience signals. An audience that drives zero impressions across all your “Best” assets is a signal not to use that audience in upcoming campaigns.

Account-level mistakes that overshadow asset-level work

Even with perfect asset hygiene, three account-level decisions undo your gains. Watch for them:

• Single asset group running too many products. 30+ SKUs in one asset group means the algorithm can’t learn which creative pairs with which product. Split into 4–6 product-themed asset groups.
• Conversion goal too narrow. If you only optimise for “Purchase,” PMax can’t lean on early signals (Add to Cart, View Item) to learn faster. Add intermediate goals as secondary conversions.
• Search themes mis-set. Adding broad search themes (“real estate”, “online courses”) confuses the model in India. Tighten to specific intent-based themes (“3 BHK in Whitefield”, “GMAT prep online India”).

What the 2026 update opens up next

Asset-level reporting is the foundation. The next layer Google is rolling out (in beta in select Indian accounts) is combination-level reporting — which exact headline + image + video bundles drove conversions. Once that’s available widely, the audit cycle will shift again.

Until then: 14-day cadence, pull/pause/promote discipline, more variety in week one, and ruthlessness only after data is ready.

If you want a working PMax audit run on your account by an operator who’s done it for 30+ Indian brands — we run free 30-minute walkthroughs from our HSR Layout studio.

Webfluence is a Bangalore-based performance marketing studio running paid, SEO and creative for 30+ Indian brands. If you’d like a working session on what any of this means for your brand, our team takes free 30-minute calls from our HSR Layout office.

Want more like this? Subscribe to Pulse — daily intelligence from the Indian marketing front lines.