If you run an operating store, you have probably hit the Performance Max asset uploader and gotten a wall of "add more assets" nudges. This article gives you the exact spec sheet in one place, then does the thing every other spec list skips: it tells you which of these numbers actually moves money, and which ones are just Google's intake form.
The full Performance Max asset spec sheet
Here is every text, image, and video requirement for a Performance Max asset group, straight from Google's documentation. Treat these as the ceiling — you supply up to this many of each, and Google mixes and matches per auction.
Text assets
Per Google's Performance Max asset requirements and the asset-group best-practices help page:
| Text asset | Max count | Character limit |
|---|---|---|
| Headlines | 15 (minimum 3) | 30 |
| Long headlines | 5 (minimum 1) | 90 |
| Descriptions | 5 (minimum 3) | 90 |
| Business name | 1 | 25 |
| Display URL path | 2 | 15 each |
Google recommends at least one headline of 15 characters or fewer and one description under 60 characters, so the system has short options for tight placements (Google Ads Help). Double-width languages like Chinese and Korean count each character as two.
Image assets
Images upload as JPG or PNG, 5 MB maximum, with the main subject inside the center of the frame. The three orientations and their sizes, per Google's asset requirements:
| Image type | Aspect ratio | Recommended | Minimum | Max count |
|---|---|---|---|---|
| Landscape | 1.91:1 | 1200×628 | 600×314 | 20 |
| Square | 1:1 | 1200×1200 | 300×300 | 20 |
| Portrait | 4:5 | 960×1200 | 480×600 | 20 |
| Logo | 4:1 (and 1:1) | 1200×300 | 512×128 | 5 |
Portrait is the one most accounts skip, and it is the one that fills vertical, mobile-first slots on Discover and YouTube. Upload all three orientations — leaving portrait blank quietly removes you from those placements.
Video assets
Videos live on YouTube, run a minimum of 10 seconds, and come in horizontal (16:9), vertical (9:16), and square (1:1), up to 15 videos total across orientations (Google asset requirements). If you do not supply a video, Google auto-generates one from your images — functional, but rarely as good as something you shot on purpose.
Why hitting spec is the easy part for an operating store
Here is the thing the spec lists do not tell you. For retailers, Performance Max is mostly a Shopping engine wearing a multi-channel coat.
The agency smec, which manages feeds for 350-plus retailers, reports that 74–97% of Performance Max cost goes to feed-based (Shopping-style) ads rather than the image, headline, and video assets you sweated over. The creative assets matter — but the load-bearing input is your Merchant Center product feed: titles, GTINs, prices, images, and availability.
Shopping has no keywords. Your product title is the query-matching surface, which is why front-loading brand, product type, and defining attributes into titles does more for impressions than a fifteenth headline variant ever will (Google Merchant Center guidance). A perfectly spec'd asset group sitting on a thin or mispriced feed is a sports car with no fuel line.
So read the spec sheet above as a floor, not a finish line. Fill every field — all 15 headlines, all three image orientations, a real video — because unfilled assets cap which auctions you can enter. Then spend the rest of your attention on the feed.
What the specs actually buy you — a worked example
Say you run a store doing 340 orders a month at a $31 average order value, and you are moving budget from Meta onto Google. Your hero product is a mug that sells for $24.
Walk the margin before you touch a single asset field. Base cost and fulfillment run $11, shipping share is $3, and processing fees are about $1 — so your contribution is $24 − $11 − $3 − $1 = $9 per mug, a 37.5% margin.
Break-even return on ad spend is one divided by contribution margin: 1 ÷ 0.375 = 2.67x. That means every dollar of Google spend has to return $2.67 in revenue just to break even on that mug. Your target ROAS should be built from that number — break-even times a profit buffer — not from a benchmark you read somewhere.
Now notice what the asset specs did and did not do. A complete asset group gets your mug eligible across every Google surface. It does not change that 2.67x floor. If your feed title is "Mug - White - 11oz" instead of "[Brand] Funny Cat Dad Ceramic Coffee Mug, 11oz," you lose impressions no headline can win back. This is the profit math behind the specs, and it is covered end-to-end in our Google Ads economics guide.
The spec sheet in context: specs are inputs, ROAS is the scoreboard
Meeting spec is necessary and cheap. The expensive mistakes happen one level up, in bidding and measurement.
Target ROAS is where the asset work pays off or doesn't. Set the target too far above what your account has historically delivered and Google simply declines auctions it predicts won't clear the bar — its own documentation warns this "may limit the amount of traffic your ads may get". And remember that ROAS is revenue, not profit; a headline figure that looks great can still lose money on a thin-margin product. We break down exactly what the metric does and doesn't include in our explainer on the official definition of Google Ads ROAS.
Context on the market helps too. Retail-category search clicks averaged $4.14 in LocaliQ's 2026 benchmark data, and WordStream found roughly 87% of industries saw year-over-year CPC increases. Rising click costs make feed quality and margin discipline matter more, not less — because the auction keeps getting more expensive while your break-even point stays fixed.
If managing the feed, the asset groups, and the bidding at once is more than you want to own in-house, it may be worth looking at Google Shopping ads management services or a dedicated Google Shopping ads agency.
Where Victor fits
PodVector AI's Victor is an AI employee that operates your Google Ads and Shopify together, not a dashboard you log into. Victor reads your live store data, computes true per-order profit on every SKU — the $9-on-a-$24-mug math above, done automatically — and surfaces the products where your feed and spend are actually making money.
Every write action Victor takes is approval-gated: he proposes the change, you approve before anything executes. Put Victor to work on your store and see the profit math behind your Performance Max spend.
FAQs
What are the character limits for Performance Max headlines and descriptions?
Headlines are capped at 30 characters (up to 15 of them), long headlines and descriptions at 90 characters (up to 5 of each), and the business name at 25 characters, per Google's asset requirements. Google recommends including at least one short headline of 15 characters or fewer so the system has a compact option for small placements.
What image sizes does Performance Max require?
Three orientations: landscape at 1.91:1 (recommended 1200×628), square at 1:1 (recommended 1200×1200), and portrait at 4:5 (recommended 960×1200), all as JPG or PNG files up to 5 MB, plus a 4:1 logo (Google asset requirements). Upload all three orientations — skipping portrait removes you from vertical mobile slots on Discover and YouTube.
Do I need to upload a video to Performance Max?
No — Google will auto-generate one from your images if you leave video blank, but the results are usually weaker than a video you make yourself. The spec is a minimum of 10 seconds, in horizontal, vertical, or square, up to 15 videos total (Google asset requirements). For an operating store with budget on the line, a real 15-second product video is worth the effort.
If I meet every asset spec, will my Performance Max campaign perform well?
Not on its own. Meeting spec makes you eligible everywhere, but for retailers the majority of Performance Max cost flows through feed-based ads — 74–97% of it, per smec. A fully spec'd asset group on a weak product feed still underperforms. Fix the feed and set your target ROAS from break-even math, and the assets do their job.
Should Performance Max replace my Standard Shopping campaign?
Usually not — run both. An Optmyzr study of 24,702 campaigns found 82% of advertisers run Performance Max alongside Shopping or Search, and the two now compete on equal Ad Rank in the same account. The current best practice is portfolio construction, not replacement.