If you want to create an effective Shopping Ads on Google campaign, then it’s essential to know what makes a profitable, optimised Product Feed.
The information and images you supply to Google Merchant Centre are the foundations that your campaign is built around. And they can have a big impact on whether you attract and convert customers.
It’s easy to assume you can just upload your existing product information into Google Merchant Centre. But if this hasn’t been optimised to appear in search results, or the Shopping tab, then you’re likely to miss out on sales. You could even be turning shoppers away.
Our guide to what makes a great feed for Shopping Ads on Google will apply equally to in-house campaigns, or if you’re working with Premium CSS partners. Mistakes with the Product Feed are some of the most common mistakes eCommerce brands make with Shopping Ads on Google.
And our tips and insight are based on practical experience. We’ve worked with RedBrain clients to optimise their product feeds and deliver around £5 million in sales every day.
In fixing specific issues with client data feeds, we’ve seen an improvement in traffic by a factor of more than 100 times. So we know how important it can be. But it’s not all about just increasing traffic numbers. For many clients, especially smaller retailers, the improvements in targeting and product performance are key to driving incremental revenue.
There are six traits we’ve identified as part of any effective product data feed:
It may seem like a short list, but there’s a lot of work and detail involved in each step. Especially if you need to apply your learnings to a massive product inventory. Or your business needs to operate multiple feeds to reach the full potential for incremental growth.
Individual areas covered include:
- Unique Product Identifiers
- Product Descriptions
- Stock Availability
- Product Type and Categorisation
It’s why the RedBrain solution is to combine machine learning and AI with human expertise to craft fully optimised product feeds as part of our managed Shopping Service.
If you work in eCommerce, you’re probably familiar with competitor research and analysis. Pretty much every retailer will want to check what other businesses are offering. But have you looked specifically at the competitors in Shopping on Google and Bing?
You might find that there are shops and retailers competing for promotion that you might not have expected. Or that specific products will underperform, due to the different tactics being employed by your competitors.
For example, if you’re displaying a product with a 500ml volume against competitors selling smaller amounts, then your item will seem more expensive. And many customers might go elsewhere without even realising any price difference is down to different sizes.
There are a few ways to look at competitive research for Shopping Ads on Google, and truly understand who you’re up against.
Google offers Shopping Insights for initial research, although this is currently limited to United States data at the moment. But there are European and UK insights regularly published by Google, and on our own RedBrain articles.
There are also third party research tools, such as the PLA Research by SEMRush, which lets you search for adverts by competitors simply by entering their website address. Obviously, each of these services will depend on how much information is collected by the provider.
If you have existing campaigns up and running, then you can enable the Market Insights programme within Google Merchant Center. This will give you access to benchmarks and data including the Price Competitiveness and Best Sellers reports. You can also use the Auction Insights report available Google Ads.
One thing your competitive research shouldn’t lead you into doing is duplicating product descriptions from rival brands. While the product specifications and dimensions will be the same for multiple retailers, you shouldn’t copy descriptions en masse, as this is deemed abuse of the Shopping Ads on Google policies.
It’s obviously important to ensure your products show up at the right time and place to capture relevant shoppers. And having a well organised and categorised product data feed will help this to happen reliably.
Although the Merchant Centre will use specific details such as titles and descriptions to display relevant products against search queries, it’s important to ensure the Product Type and Categorisation are optimised.
Google recommends that you use a ‘google_product_category’ value which is at least 2-3 levels deep to ensure your adverts are delivered to the best possible audience. You should start by using the supported Google product taxonomy of around 6,000 product categories (you can download the full list, here) to move from the broadest information to the most granular.
Individual products are limited to just one full category path. The syntax for your feed can also use the corresponding numerical category ID.
- Clothing & Accessories > Clothing > Outerwear > Coats & Jackets > Denim Jackets
- Arts & Entertainment > Party & Celebration > Gift Giving > Gift Cards & Certificates (53)
The ‘google_product_type’ value offers a similar function but will use your own taxonomy rather than the designed Google values. The Product Type isn’t a mandatory requirement, but along with the Product Category it will help to ensure relevance, and more effective bidding by grouping products.
It can be tempting to try and include promotional information in the Product Type, or insert keywords repeatedly. But this can potentially risk your account being suspended. What is important is that you make the most of both the Product Category and Product Type.
Although the same values can be used for both the category and type data, this misses a big chance to optimise your product feed.
- Home > Photography > Lenses > SLR Lenses
Making the most of your type and categorisation opportunity will either require a significant amount of manual time and effort. Or you can save your resources by working with a Google Shopping service which can take care of the optimisation on your behalf.
For successful campaigns with Shopping Ads on Google, you may want to include multiple feeds. This is recommended for running discounts and sales which avoids the need to manually alter your primary feed (and can lead to errors when sales end, for example).
Or you may want to start separating feeds for specific product categories and types, particularly if you want to run different bidding strategies across product segments.
There are a number of product attributes which need to be supplied accurately to Google. Not only does this ensure that items are displayed in the right way, but inaccurate information can lead to products or even entire accounts, being disapproved or suspended.
Unique Product Identifiers:
It’s not just categories which are used to identify individual products. One key area is the Unique Product Identifier, which can be Global Trade Item Numbers (GTINs) (which are most commonly seen on barcodes), or a Manufacturer Part Numbers (MPNs).
Google holds a database of GTINs, so it’s important that any information is supplied correctly. Invalid numbers will be flagged by Google within the Merchant Centre. Products assigned a GTIN but submitted without one can be disapproved and won’t display.
Products which often don’t have an assigned GTIN include antiques, replacement parts or store-brand products. If your product doesn’t have a GTIN, you can use a Manufacturer Part Number or Brand attributes to help identify your product.
You can find more information on GTIN details for your product feed via the Merchant Centre Help Centre.
Price is an important required element for any Google Shopping product listing. It needs to be kept accurate, and displayed prominently on your landing page, in the currency for your country of sale.
The amount shouldn’t vary on your landing page due to the location of the user. And it shouldn’t include the delivery cost. There’s a separate shipping attribute for that information.
Common issues with product prices include the time it can take for changes to be updated by Google, supplying incorrect information, or having to make frequent changes if you’re running sales and promotions.
The price and availability information will always be set to automatically update and overwrite the previous data by default. But if you’re relying on Google to use Schema markup and crawl your website, then it may take a little time for the newest changes to be reflected in your campaigns.
Discounts and sales shouldn’t be run using the main pricing information. The ‘sale_price’, and ‘sale_price_effective_date’ attributes should be used either within your primary feed, or a supplemental feed, to display the lower price alongside the original retail cost as a sales price annotation. It’s also possible to display sales pricing for individual product variations such as colours or sizes, or to set it as low as 0 for items which come with a contract such as mobile phones or tablets.
Ultimately, pricing errors including differences between product data feed and the landing page or displaying the wrong currency can lead to accounts being disapproved and suspended.
The reason for disapproval can be checked via the Diagnostics tab in Merchant Centre and will be emailed to the listed technical contact for the account. When the problem has been corrected, it’s possible to request a manual review. But this can take up to three working days to be completed, meaning that you could lose a substantial amount of sales while your campaign, or product, is offline.
Availability is a required field for your product feed if you want your adverts to display. This can be done via Schema markup for ‘Offer.availability’.
Along with pricing, Google Merchant Centre will automatically update the latest information, and you can list items as ‘in stock’, ‘out of stock’ (which will stop ads appearing), or ‘preorder’ for new items which are not released for sale.
Out of Stock listings should be used for products which are temporarily unavailable due to inventory, your website being offline for a time, or to temporarily stop adverts for displaying even if a product is actually listed as available via your website.
Incorrect information doesn’t just affect your campaign standing with Google. It will also impact on performance and sales. And it can confuse and annoy disappointed shoppers if they visit your website only to find their intended purchase isn’t actually available. Which will hurt your branding and reputation in the future.
It will also confuse and potentially annoy shoppers, which can damage your branding and reputation.
Each product title is limited to 150 characters, but only around 45 will typically be visible on your adverts. If you’re using Schema markup for Google to automatically pull information from your site, this will be under the product.name property.
Your titles need to be grammatically correct. It’s important to avoid having products disapproved and blocked from displaying because you’ve used symbols, HTML tags or block capitals. Or if you’ve used promotional, rather than descriptive, text.
The key is to be specific, relevant and accurate to ensure your product is shown to the right customers. As a general rule, the order of importance should be Brand, Product Type and then Attributes.
The rules around product titles don’t restrict you from optimisation. Keyword research is important to ensure your titles are using the words and phrases potential consumers are using for searches. Using keywords effectively doesn’t mean repeating them within your title, known as ‘keyword stuffing’.
Although Shopping Ads on Google are displayed based on product data, rather than an advertiser choosing particular keywords to appear against, it can still be worth referring to your competitive analysis and keyword research.
This can be used to ensure that your product titles and description are relevant for the names and attributes customers are searching for when they look for a specific item. Sources for new campaigns can include Google Search Console, which displays queries and clicks for your website, Google Keyword Planner for broad forecasts of paid search volumes and competition. And various third party research tools exist, including SEMRush, Ahrefs, Moz and more.
When you have Shopping Ads on Google campaigns running, one of the key pieces of optimisation is adding negative keywords to your settings. This will prevent your adverts running against irrelevant customer search queries. If you’re operating on a CPC basis, this can cost you money, as well as impacting the performance of those adverts.
You can use the Search Terms Report in Google Ads to see which queries triggered adverts to appear by product groups, and then identify those resulting in traffic which was high cost but low value. It can also reveal information that might need to be added to your product descriptions to match customer intent more effectively.
The product description field gives you up to 5,000 characters to describe what you’re selling in much more detail.
The most important unique selling points should be listed early in the description, as only the first 160-500 characters are displayed by default. To see more of the information, users will need to click to expand the text.
To encourage customers to either buy the product or read more information, a description should be written in a natural, readable style which will engage and hopefully interest anyone searching for that product.
Google recommends using rich descriptions to help with readability. These allow HTML formatting including line breaks, italics and bullet lists to help make your text accessible to customers. Any issues with tags will need to be checked, as unclosed tags or other problems will cause your advert to be disapproved and not appear.
You should include relevant information including product specifications, materials, physical sizes, and other important attributes such as colour and gender. This is where prior keyword research and reports can help to identify important terms to include.
Rather than repeating the same information again, and again, try to use related terminology or synonyms to increase your chances of appearing for multiple relevant queries. If you want to target ‘trainers’ as your main keyword, you could also use ‘sneakers’ and ‘shoes’, alongside the colour, style and any specific sports and activities they are designed for.
Visual appeal is a key selling point for many consumers. It’s an obvious point for luxury and aspirational items, but it also applies to more mundane purchases.
Ensuring your inventory is shown in the best way is an important part of optimising your product data feed.
Images for Merchant Centre will be taken from a submitted url. Not only do the pictures need to meet specific file format and size requirements, but promotional text and frames are not permitted.
If you don't follow these requirements, your product will be disapproved and you will be notified in the Diagnostics section of your Merchant Centre account.
- Supported formats are: GIF (.gif), JPEG (.jpg/.jpeg), PNG (.png), BMP (.bmp) and TIFF (.tif/.tiff)
- Non-clothing images: At least 100 x 100 pixels
- Clothing images: At least 250 x 250 pixels
- No image larger than 64 megapixels
- No image file larger than 16 MB
Google recommends that you display the entire product with minimal staging and a simple plain white or transparent background. Your products should take up between 75-90% of the image you submit. Within those guidelines, it’s still possible to make your listings stand out. The way to add products in use, including graphics and illustrations, is with up to 10 additional product images using the ‘additional_image_link’ attribute.
Additional images have the same format and size limits. But you can use highly visual selling techniques by showing specific details or angles of the product. It’s also the place for images of people using your product, graphics or illustrations, or focusing on products within a bundle of items.
Google does provide an automatic image improvement service within the Merchant Center, although it may not always be successful. This will look at any images disapproved due to a promotional overlay, and apply an automatic removal tool. If successful, the image will be replaced and the offer reapproved. The automated service doesn’t always work, in which case the image may stay disapproved, or a non-promotional part may be affected, in which case you can always replace it with a manually edited version.
You can see what has been automatically improved, or has stayed disapproved within the Diagnostics page of your Merchant Center account by looking for “Improved image quality [image link] within the Issue column.
If you’re changing a large number of images it can take a little time for them to be fetched and included in your advertising. Creating a new image url will generally appear more quickly than using an existing url, and it’s recommended that you submit a percentage of a large collection of images on a daily basis.
Monitored and Updated:
It’s important to regularly monitor, manage and update your Product Data Feed for Shopping Ads on Google as required. It’s not something that can be set up and left indefinitely. All products expire in Google Merchant Center after 30 days, so a new feed has to be available within that period as a minimum. Which means new issues can be popping up on a frequent basis.
Google Merchant Centre provides a Diagnostics report which will highlight any new or existing issues with your data quality which should be improved. There are three types of problems which will be brought to your attention.
- Account Issues - these are problems which affect everything, such as no returns and funds policy on your website.
- Feed Issues - problems related to fetching, uploading and processing your feeds
- Item Issues - product specific problems including unavailable landing pages or invalid images.
Individual issues are assigned 2 levels of importance, to help you prioritise.
- Errors: Major problems which need to be resolved quickly. Identified by a red circle icon
- Warnings: Important fixes that will make a big difference to performance. Indicated by a yellow triangle icon.
It’s important to respond to any Errors or Warnings which can lead to items, feeds and even accounts being suspended. Notifications are also available in Common feed errors include:
- Product Crawl issues – Page not found (404) errors, invalid urls or crawl access disallowed by the website robots.txt file.
- Images not displaying – recently uploaded or changed, sizes too large, incorrect urls, access blocked by robots.txt, image redirects, or dynamically generated images without http headers. Placeholder images are the most common issue we encounter.
- Data quality – robots.txt exclusion, or Unclear value on your website (Google is unable to extract price, availability or condition of your offer)
Management of your product data feeds is particularly important during big events such as Black Friday, or when a new launch is occurring. And especially if you’re making large scale structural changes to your website. Google has a range of tips to help your products stay approved during those times.
How your feed can drive sales
Every improvement and enhancement you can make to the product data supplied to Google Merchant Centre will increase the chances of your adverts appearing at the right time and place to relevant customers.
The result may be that you increase traffic viewing your campaigns and then visiting your website. But it can also increase relevance and lower costs by reducing wastage if your adverts currently appear when they shouldn’t.
And optimising your product title, pricing and imagery will help to attract the attention of potential customers.
The impact of optimising your product feed for Shopping Ads on Google is difficult to quantify in exact amounts. Minor tweaks to individual products may deliver a small uplift. But fixing a more serious issue for one new RedBrain partner increased their traffic by a factor of 100. Typically a retailer will see the best results from constant improvements and a cumulative effect, which is why the RedBrain platform uses AI and machine learning to scale optimisation across massive product inventories.