Writing Google Shopping and Performance Max headlines that convert
How to write PMax headlines within Google's exact character limits so your Shopify products actually get clicked and bought.
Written by Mantas Jurgutis — Founder, Adsify — builds the Google & Meta automation merchants use daily
Editorially reviewed by Adsify Editorial on February 6, 2026 — Reviewed against Shopify, Google Ads and Meta official documentation.
Headlines carry more weight than merchants expect
Performance Max assembles ads dynamically across Search, Display, YouTube, Gmail and Discover, and Google's machine learning decides which headline, description and image combination to show in each placement. That means a weak headline doesn't just underperform in one spot — it drags down the whole asset group's Ad Strength score and, with it, delivery. For a Shopify store, headlines are often the only text a shopper reads before they see your product image, so treating them as an afterthought is one of the most common reasons PMax campaigns stall at 'Learning' or deliver mediocre ROAS despite good product-level performance elsewhere.
The character limits, exactly
Google Ads requires short headlines up to 30 characters, at least one long headline up to 90 characters, and descriptions up to 90 characters, per Google's Performance Max asset requirements. You can supply up to 15 short headlines, 5 long headlines and 5 descriptions per asset group. Google explicitly recommends variety over repetition — five near-identical short headlines give the algorithm nothing new to test. Count characters including spaces and punctuation; a headline that's 31 characters will simply be rejected at upload, not truncated, so check before you paste a batch into the asset group editor.
Short headline (30 chars): lead with the concrete detail
At 30 characters you have room for one idea, not a pitch. Weak: 'Amazing Quality You Will Love' (uses all 30 characters saying nothing). Better: 'Waterproof Hiking Boots' (24 chars) or 'Ships Free, Arrives in 2 Days' (30 chars). Specific nouns and numbers outperform adjectives because they match what a shopper actually typed into the search bar. If your product has a defining spec — weight, size, material, battery life — that spec is almost always a stronger 30-character headline than a superlative claim, and it's easier for Google to match to a relevant query.
Long headline (90 chars): this is your value proposition
The 90-character long headline is where you can actually make a case. Structure it as [what it is] + [key benefit] + [proof point or offer]. Example: 'Merino Wool Base Layer — Regulates Temperature From 20°F to 70°F, Machine Washable' (86 chars). Avoid burying the benefit at the end where it may get cut off in narrower placements like Gmail. Front-load the most persuasive clause. If you sell a bundle or subscription, this is also where '30-Day Money-Back Guarantee' or 'Free Returns, No Questions Asked' earns its place — trust signals convert better in the longer format than in the clipped short one.
Descriptions (90 chars): answer the objection, not the pitch
Descriptions run alongside headlines in Search results and should do different work than either headline — namely, handle the objection that stops a shopper from clicking. If your product is priced above category average, the description is where you justify it: 'Handmade in Portugal, 2-year warranty included.' If shipping cost is a friction point, put 'Free shipping over $50, orders ship same day' here. Don't repeat the short headline verbatim in the description; Google's combination testing gets less useful information when two assets say the same thing, and Ad Strength feedback will often flag it directly.
Write for combinations, not for a single ad
Because Performance Max mixes and matches your assets automatically, each headline and description needs to make sense in isolation, paired with any of your other assets. Avoid headlines that only work as a pair, like Short Headline 1: 'Buy One' and Short Headline 2: 'Get One Free' — if Google shows Headline 1 with an unrelated long headline, the offer reads as broken. Instead write '50% Off Second Item' as one complete headline. This single-asset-completeness rule is the most overlooked cause of confusing PMax ads that technically pass Google's review but read oddly to shoppers.
Use Ad Strength as a checklist, not a target
Google's Ad Strength indicator (Poor, Average, Good, Excellent) rewards headline count and keyword diversity, and it's a useful pre-launch checklist: are you using close to the maximum 15 short headlines, do they cover different angles (price, material, use-case, urgency, social proof), and do none of them duplicate each other. But Ad Strength is not a performance guarantee — an 'Excellent' asset group with weak underlying offers will still underperform a 'Good' one with a genuinely strong product and price. Use it to catch obvious gaps, then let real conversion data override it after two to three weeks of delivery.
Seasonal and inventory-aware headline rotation
Headlines referencing 'Limited Stock' or 'Ends Sunday' need to be accurate — Google's shopping ads policies prohibit false urgency, and inaccurate claims risk disapproval or account-level review. If you run flash sales, update or pause those specific headline assets when the sale ends rather than leaving stale urgency copy live; shoppers who click a 'sale ends today' headline three weeks later on a full-price product page will bounce immediately and that bounce signal feeds back into your Quality and conversion data. Build a simple recurring calendar reminder tied to your promo calendar rather than trusting memory.
Worked example: a $68 ceramic knife set
Assume a Shopify store sells a 5-piece ceramic knife set at $68, free shipping over $50, lifetime sharpening included. Short headlines: 'Ceramic Knife Set, 5-Piece' (26), 'Never Needs Sharpening' (23), 'Free Shipping Included' (23), 'Rust-Proof Blades' (18). Long headline: 'Ceramic Knives Stay Sharp 10x Longer Than Steel — Lifetime Sharpening Included' (79 chars). Description: 'Dishwasher safe, BPA-free handles, ships free over $50' (56 chars). Each asset stands alone, none repeat, and together they cover material, durability, price and logistics — the four objections most likely to stop a knife-set purchase.
Testing headlines without breaking Learning
Performance Max re-enters a partial learning period whenever you make significant asset group changes, so batch headline edits rather than tweaking one at a time across a week. A practical cadence: review headline-level asset performance (labeled Low, Good, Best in the asset group report) every two weeks, replace only assets rated Low, and change no more than a third of your headlines in a single edit. This keeps enough consistency for Google's systems to retain learned signals while still letting you cut genuinely weak copy.
Localizing headlines instead of translating them
If you sell into multiple countries, direct translation of a strong English headline often produces an average one — idioms, unit systems and cultural urgency cues don't map cleanly. 'Weighs Less Than a Phone' works in the US; in a market using different phone size norms it may not land. Rewrite the underlying claim per market using local units (grams vs ounces, cm vs inches) and locally common offer language rather than running the same headline set through translation and calling it done.
Common headline mistakes that quietly cap performance
The recurring patterns worth auditing for: headlines that repeat the brand name in every slot instead of product benefits, headlines written entirely in title case with exclamation points (reads as generic and can suppress click-through), missing numeric details when the product has an obvious one (size, count, weight), and descriptions that just restate the product title from the Shopify feed. None of these will get an ad rejected, but each one quietly narrows how well your asset group differentiates itself from a shopper's other five open tabs.
Where AI-generated copy helps and where it still needs a pass
Tools that generate ad copy from your Shopify product data — Adsify included — are genuinely useful for producing volume: pulling specs, materials and price points straight from product descriptions into properly-sized headline drafts across 15 short and 5 long slots in minutes rather than an afternoon of manual character-counting. What still needs a human pass is voice consistency and claim accuracy — confirming a generated 'best-selling' or 'award-winning' headline is actually true for that SKU before it goes live, since Google can disapprove unsubstantiated superlative claims.
Building a headline bank you can reuse
Rather than writing headlines fresh for every campaign, maintain a spreadsheet organized by category: material/spec headlines, price/value headlines, shipping/logistics headlines, urgency headlines, and social proof headlines, each pre-counted for character length. When you launch a new product, you're filling in specifics against a proven structure rather than starting blank. Over several campaigns you'll also start to see which category of headline actually drives the asset-level 'Best' rating most often for your store, and that pattern is more useful than any general best practice.
The takeaway
PMax and Shopping headlines succeed or fail on specificity within a hard character budget. Write each short headline, long headline and description as a complete, standalone thought; cover distinct objections across your asset slots; keep urgency claims accurate; and revisit performance every two weeks rather than daily. Do that consistently and the algorithm has genuinely differentiated material to test, which is the entire mechanism Performance Max relies on to find your best-converting combination.