About automatic item updates
Issues such as latency between updates on your website and updates to your data submissions can result in inaccurate or stale product data. Automatic item updates can help you avoid this by keeping your product data up to date. Automatic item updates allows Google to update your product listings on Shopping ads using the structured data markup we find on your website and advanced data extractors. Automatic item updates is currently offered for the price and availability attributes.
For example, if your most recent data feed contains an item that costs $4 but your product landing page lists it as $3, we will update the item to $3 in your Shopping ad.
This article explains how automatic item updates work.
Automatic item updates lead to an improved user experience, more traffic to your Shopping ads, and to higher conversion rates because users already see the correct price in your Shopping ads and are not sent to out-of-stock items. Since we’re updating your Google Shopping product data with information found directly on your landing pages, this feature also minimizes the number of mismatches between your Shopping ads and your website. This in turn reduces the risk of account suspension for price and availability mismatches. If too many mismatches occur, however, automatic item updates will stop working and your account will be at risk for preemptive item disapproval. Learn more about preemptive item disapproval
Automatic item updates will be turned on by default in your account. If you decide to turn them off, items with mismatched price and availability will be temporarily disapproved rather than updated. To manage automatic item updates in your account, click here.
Automatic item updates isn’t a replacement for your feeds or for regular updates of your product data. It is designed to fix small problems with your price and availability accuracy for a small percent of your items. You should continue to provide accurate product data frequently as we will not be able to cover all of your items with automatic item updates.
About advanced data extractors
Advanced extractors are able to find price and availability information on a product’s landing page. They use a combination of statistical models and machine learning to detect and extract product data from your website independent structured data markup. These extractors work in addition to schema.org structured data markup and using schema.org markup is still recommended for retailers.
Example automatic item update
In the image below, you can see when we update the price and availability of an item, using price information and a .txt data feed for this example. When you upload an item (green icon) with a price of $1, a user (blue icon) will see this price. If we find a price mismatch after your upload, we will update the price of the item automatically. In this example, we find a price of $2 on your website, which doesn't match the price provided in your most recent data submission. From this point on, the user will see $2 as the price of the item. The next feed upload (time B) will overwrite this update and the user will now see a price of $3.
Getting started with automatic item updates
1. Implement valid structured data on your landing pages
The first step to getting started with automatic item updates is ensuring that you have valid structured data on your item’s landing pages. If your landing pages don’t contain any structured data, or the structured data markup is wrong, incomplete, or not valid, we'll use advanced data extractors to update items automatically. If you turn off Automatic item updates, or our extractors are unable to determine any price or availability information, the item will be subject to temporary item disapprovals. Learn more about getting started with structured data.
2. Review items that have been updated
If you have not disabled Automatic item updates, you can view the current status of automatic item updates on the Automatic item updates section of your account if you’ve enabled this feature. You can also find all items that were recently updated automatically in the downloadable report available in the Diagnostics page under Products.
Troubleshooting issues with automatic item updates
In order to ensure that updates are done correctly, we may temporarily stop updating items automatically if we detect an error. You can view these errors in the Diagnostics page under Products in your account. Once all of the errors are fixed, we’ll start updating items automatically.
Be aware that we need to access your product landing pages and extract structured data markup information successfully before the the issues in the report can be resolved. This may take a couple of weeks.
The errors below will result in a temporary stop of automatic item updates:Insufficient match of structured data and product data
We require a minimum match level between the product data you submit to Google and both the schema.org structured data markup on your landing pages and the information available to the advanced data extractors. If this data differs too often, we stop updating items automatically until the match level has reached the required minimum. To improve the match rate we recommend you provide more frequent updates of your Shopping ads product data.
Note: We often see price mismatches due to sales prices. When a product is available at a discounted price this needs to be reflected in the structured data and in your product data. Another common source of mismatches are price ranges and variants. If there are multiple variants shown on the landing page, you can use the
gtin attribute to specify which variant a price corresponds to.
In order to ensure that updates are done correctly, we compare the user-visible information on your product landing pages to the data that we find using either structured data markup or using advanced extractors for a sample of offers. If we find a mismatch, we stop updating items automatically
If your product landing pages have invalid price and availability information in the structured data markup, we might not update items automatically. Review the listed examples and ensure that they contain valid structured data. The Structured Data Testing Tool can help you debug any issues.
The error below will significantly reduce the scale at which we’ll be able to update items automatically:Limited crawling capacity provided
Automatic item updates rely on our ability to access your product landing pages frequently to provide updates in a timely manner. Google Search Console allows you to control the rate at which Google is crawling your web site. These settings will also be used for your landing pages provided in your product data.
On the server side, we recommend treating requests from Google (user agent 'GoogleBot' and 'AdsBot') like requests from regular users and not apply any throttling of requests.
Limitations of automatic item updates
We won’t be able to update items effectively in all instances depending on the frequency of price and availability updates on your web site. As a rule of thumb, automatic item updates won't be very effective if the price or availability of the majority of the items on your website is updated multiple times per day. This includes, for example, scenarios where prices are regularly updated based on currency exchange rates. In these cases we recommend that you use the Content API to update your product data.
Advanced: Accounting for site latency
One of the causes of inaccurate or stale product data is latency, which is the time between updates on your website and updates to the product data you submit. We’ve factored in this latency period to automatic item updates.
The image below shows how latency may impact how we update your items automatically, using price information and a .txt data feed for this example. The yellow bar represents a latency period before a feed upload. Let’s say you have an item that is displaying to the user at $3. We then detect a price mismatch within a latency period (time C). A user will now see a price of $4 even after a feed upload has provided a price of $5 because the detection of the mismatch happened within the latency period of this feed upload. The next feed upload (time D) will then change the price to $6 because the detection of the mismatch didn't occur within the latency period of this feed upload.