Product Title Engineering: How to Structure Titles That Actually Get Clicks
A Shopping title is doing three jobs at once. It is the primary keyword signal that tells Google which search queries your product is relevant for. It is the first line of copy a shopper reads when deciding whether to click. And it is a targeting input that directly influences which auction segments your product enters. One field. Three simultaneous functions. Most product titles fail at all three.
The failure mode is almost always the same: titles written by whoever set up the product catalogue, structured for internal logic rather than for search matching, and never revisited after the initial upload. A furniture retailer with 4,000 SKUs whose titles read like warehouse codes is not making a strategic choice. They are operating with a default that was convenient at setup and is now costing them impressions, clicks, and ROAS every day the feed runs unchanged.
This article covers the structural principles behind effective Shopping titles, the formula that works across most product categories, category-specific adaptations with direct examples, and a testing method that identifies title improvements without requiring a full catalogue rebuild. The goal is titles that match the right queries, earn the click when they appear, and do both consistently at scale.
How Google Reads and Matches Product Titles
Google processes Shopping titles as a keyword signal in the same way its search algorithm processes page titles. The words in the title are indexed and matched against incoming search queries. The match does not need to be exact; Google applies semantic matching that connects query intent to product attributes. But the quality of that matching is directly proportional to the specificity and relevance of the information in the title.
Two title mechanics are particularly important to understand. The first is front-loading. Google weights the beginning of a title more heavily than the end, in the same way that search snippets give more prominence to early keyword placement. The most commercially important attribute of a product should appear in the first three to five words of the title. For most product categories, that attribute is the product type. For branded products in categories where brand name is a primary purchase signal, brand comes first.
The second is character limits and truncation. Google Shopping titles support up to 150 characters, but the displayed title in a standard Shopping ad unit is typically truncated at 70 characters on desktop and fewer on mobile. This means the first 60-70 characters of a title must contain the information that drives both relevance matching and click-through. Attributes that appear after character 70 still contribute to keyword matching, but they will not be seen by the shopper unless they expand the listing.
The Title Formula That Works Across Categories
A consistent title structure applied across a catalogue serves two purposes: it ensures that the most important attributes appear in the correct positions for every product, and it makes systematic title optimisation feasible at scale. A catalogue without a title formula has to address each product individually. A catalogue with a defined formula can be updated by changing the template.
Standard Title Formula:
[Brand] + [Product Type] + [Key Differentiating Attribute] + [Variant: Size / Color / Material]
Example (Apparel):
Patagonia Fleece Jacket Lightweight Recycled Navy Blue Medium
Example (Electronics):
Sony Wireless Headphones Noise Cancelling Over-Ear Black
Example (Home Goods):
Le Creuset Cast Iron Dutch Oven 5.5 Quart Flame Orange
Brand first for products where brand recognition influences purchase intent. If a shopper is specifically looking for your brand, placing it first ensures the brand name appears in the visible title before truncation. If brand is not a primary purchase driver in your category, product type takes the first position.
Product type second as the primary keyword anchor. This is the term that matches the broadest relevant search queries for the product. ‘Fleece Jacket’ is a product type. ‘Outdoor Layering Piece’ is a marketing phrase that matches nothing a shopper types. Use the term shoppers use, not the term the brand team prefers.
Key differentiating attribute third. This is the attribute that separates this product from other products of the same type in your catalogue or in the broader market. Lightweight, waterproof, noise cancelling, organic, handmade: whatever makes this specific product meaningfully different from the generic category.
Variant attributes last. Size, colour, material, and model number sit at the end of the title. They contribute to long-tail query matching and help Google surface the right variant to the right shopper. They do not need to be in the visible window because a shopper who clicks on a ‘Fleece Jacket Lightweight Recycled’ listing expects to see variant options on the product page.
Category-Specific Adaptations With Examples
The base formula holds across most verticals, but the weight and position of specific attributes shift depending on how shoppers search within a category. The before-and-after examples below illustrate how the formula adapts across three distinct product types.
| Category | Before | After | Why It Works |
| Apparel | WMS-BLK-L-032 | Nike Women’s Running Shorts Dri-FIT Black Large | Brand first (strong purchase signal), product type and gender context, material technology as differentiator, then variant. Moves from zero keyword signal to full match coverage. |
| Consumer Electronics | Bluetooth Speaker Portable | JBL Portable Bluetooth Speaker Waterproof 20W Bass Boost Black | Brand anchors the listing. Product type stays. Waterproof and 20W are specific decision attributes shoppers filter by. Colour variant closes the title. |
| Home Goods | Large Wooden Cutting Board Kitchen | John Boos Maple Wood Cutting Board Large 20×15 Inch End Grain | Brand first. Material (Maple Wood) is a primary quality signal in this category. Dimensions are a purchase decision factor. End Grain construction is a differentiator for buyers who know the category. |
The pattern across all three examples is the same: replace generic or internal language with the specific attributes a shopper types when they know what they want. The improvement is not about adding words. It is about replacing the wrong words with the right ones in the right order.
What Not to Put in a Product Title
Knowing what to exclude is as important as knowing what to include. Several common title elements actively harm performance by consuming character space, triggering policy flags, or diluting the keyword signal.
| What to Avoid | Why It Hurts |
| Promotional language | Words like ‘Sale’, ‘Best Price’, ‘Free Shipping’, or ‘Limited Offer’ in titles violate Google’s Shopping policies and trigger disapproval. These belong in the promotions attribute, not the title. |
| ALL CAPS words or phrases | All-caps formatting is a policy violation and a CTR signal killer. It reads as shouting, reduces perceived listing quality, and increases disapproval risk. |
| Excessive punctuation | Multiple exclamation marks, ellipses, or decorative symbols consume characters and degrade relevance scoring. A title is a data field, not a marketing headline. |
| Vague category placeholders | Terms like ‘Product’, ‘Item’, ‘Goods’, or internal category names that mean nothing to a shopper contribute zero keyword signal and reduce match quality. |
| Redundant repeat of the same word | Repeating the same attribute twice (‘Black Jacket Dark Black’) does not improve match quality and wastes character space that should carry distinct keyword signals. |
| Marketing taglines | Brand slogans, positioning statements, or descriptive marketing copy belong in the description field, not the title. ‘Built for the Outdoors’ is not a product type. |
Testing Titles at Scale Without Rebuilding the Catalogue
For large catalogues, testing title changes across all products simultaneously carries risk: if the new title structure reduces relevance scores across a product group, ROAS can drop across thousands of SKUs before the problem is identified. A more controlled approach uses custom labels and campaign segmentation to test title changes on a defined subset before rolling out broadly.
Step 1: Select a test segment. Identify a product group that is underperforming relative to category averages. Fifty to two hundred products is enough for statistically meaningful results. Use a custom label to tag these products as the test group.
Step 2: Apply the revised title structure to the test group. Create a feed rule or update the test products directly with titles structured according to the formula. Do not change bids, campaign structure, or any other variables during the test period.
Step 3: Run the test for three to four weeks. Shopping performance data takes time to stabilise after feed changes. Google re-indexes the feed and adjusts relevance scores over several days following a title update. Measuring results within the first week produces unreliable signals. Three weeks provides enough data for meaningful comparison against the pre-test baseline.
Step 4: Measure impression share, CTR, and conversion rate. ROAS improvement from title changes flows through three metrics in sequence: more relevant queries match the product (impression share on target queries increases), more shoppers click because the title is descriptive and specific (CTR improves), and more clicks convert because the query-to-landing-page match is tighter (conversion rate improves). If impression share and CTR improve but conversion rate does not, the landing page is the next variable to investigate.
Platforms like EnhanceFeed automate this testing workflow by applying title transformation rules to defined product segments and tracking performance changes at the feed level, removing the manual coordination between feed updates and campaign reporting that makes title testing operationally intensive for large catalogues.
Title Engineering Is Conversion Engineering
The product title is the single highest-leverage field in a Shopping feed. It determines query eligibility, influences ad copy quality, and affects purchase confidence before the click. Teams that treat it as a data entry task rather than a performance engineering decision are leaving a measurable gap between their current ROAS and what the same catalogue and budget could generate with better data.
The formula is not complicated. Brand, product type, differentiating attribute, variant. Front-load the commercially important terms. Stay within 70 characters for the visible window. Exclude everything that wastes space or triggers a policy flag. Apply the structure consistently across the catalogue using feed rules so that new products enter the system correctly from the start.
The catalogues that consistently outperform in Shopping auctions are not the ones with the highest bids. They are the ones where someone, at some point, decided that the title field was worth taking seriously. That decision is available to every team managing a feed today.
