What makes a good product title for Google Shopping?

Of all the attributes you submit, the title does the most work in deciding which queries your product can be shown for. Category, product type and identifiers narrow the field, but the title is where the specific wording lives, and specific wording is what a shopper's question contains.

That makes the title a research task rather than a naming task. The right words are the ones shoppers use, which are frequently not the words the manufacturer uses. Your Google Ads search terms report is the best source available, because it shows the real phrasing that led to clicks and conversions on your products. Site search and support messages fill in the rest.

It is also the field a shopper reads before deciding whether to click, which means it has to be honest and legible at a glance. A title assembled to satisfy a matching algorithm at the expense of readability tends to lose on click-through what it gained in impressions.

One more consequence: because the title is doing so much work, small changes to it move performance more than most feed edits. That makes it worth treating as something you test rather than something you write once.

There is no single formula, and applying one across a mixed catalogue produces poor titles in most of it. What works is a pattern per category, built from the attributes that a buyer in that category uses to identify the product.

In apparel and footwear the pattern usually runs brand, then the garment or style name, then the defining attributes: colour, material, and size where size is submitted in the title. Shoppers search by brand and item, and the attributes disambiguate variants.

In electronics and appliances the model number matters as much as the name, because people paste it. Brand, product, model, then the specification that differentiates, such as capacity or screen size.

In parts, consumables and industrial supply the compatibility is the query. Lead with the brand and part number, then state what it fits. A cartridge title that names the printer models is answering the actual question.

In food, supplements and household goods, size and count belong early, because a shopper comparing prices needs them to make sense of the offer.

For unbranded or private label goods, lead with the descriptive product name plus the attribute that identifies it, since there is no brand to anchor on.

Google's product data specification sets the boundaries. The title must describe the product and must match the landing page. Promotional text is not allowed, so no mention of sale, discount, percentage off, free shipping, best seller or limited time. All capitals and gratuitous symbols are not permitted either.

Within those limits, a few conventions consistently help. Front-load the important words, since the visible portion is short in most placements. Keep the length within the 150 character limit and treat roughly the first 70 as the part that will be read. Use the singular product name rather than a category phrase. Include size or colour in the title only where shoppers search with them, which is common in apparel and rare in furniture.

Avoid repeating a keyword to strengthen matching. It reads badly, it is against the specification, and it does not work.

One thing worth checking is duplication across variants. If every size of a product carries an identical title, the listings compete against each other and the shopper cannot tell them apart, so include the differentiating attribute for variants that are submitted separately.

Generate them from attributes rather than editing them by hand. Once brand, product type, colour, size, material and model exist as clean fields, a feed rule or a supplemental feed can assemble a title per category pattern, and any change to the pattern applies across the whole group at once.

Keep the generated title in the feed rather than changing the product name in your store, so the storefront title stays natural for shoppers while the feed title carries the search wording. Both must describe the same product, but they do not have to be identical strings.

Test by splitting a category into two comparable halves using a custom label, apply a new pattern to one half, and run both for long enough to see a difference at your click volumes. Measure impressions first, since a matching change shows there before it shows anywhere else, then click-through rate, then conversion rate. A pattern that gains impressions and loses conversion rate is matching more loosely, not better.

Re-test when the catalogue changes materially or when a season shifts the vocabulary. Title patterns are not a one-off project, and a store that revisits them a couple of times a year keeps finding gains that a store which set them once does not.