Looker Studio blending lets you create charts, tables, and controls based on multiple data sources. But there is much more to blends than the definition implies. In this series of blog posts, we are going to go over the fundamentals of blending as well as cover some advanced use cases.
Why blend data?
The blending feature opens a world of possibilities for effective analysis of data. You can enhance your reports with additional data from other sources, for example, add customer information from a CRM database to your ecommerce transactions reports or analyze effectiveness of your ad spend by adding product sales details to your ad impression and click reports. But, in addition to enriching your data with more fields, you can also do things otherwise impossible with a single data source. One example is blending the same data source with itself. This one simple trick allows you to create calculated metrics based on a subset of the same original metric. This is most useful when working with custom events and can also be helpful in other use cases. Say you have a pageview report filtered to a specific section of your site, for example only product detail pages. You want to show product pageviews by category as well as what percent of all site pageviews it represents. You could use a “Percent of Total” comparison calculation field, but because this report is filtered to only include product pages, this would only tell you percentage of total pageviews of product pages, and not the entire site. With blending, this becomes possible.
When should you NOT blend?
Blending is not possible when you have more than five data sources. Currently, Looker Studio only allows you to blend up to five data sources. If you have more complex requirements, you will need to do some prep work of combining some of your data sources before you add them to your blend or find a different reporting and visualization tool that does not have this limitation.
Blends can also result in a very large dataset if the data sources you are blending are large and you are not retrieving a subset of that data. This can either slow down your reports or break your dashboards completely if the resulting dataset is simply too large for the tool to process. Google strongly advises that blending should be used primarily to create subsets of original data.
You should also think twice about blending when you need to compare your metrics to a different time range, such as previous year or previous month. In cases when there is a “null” value in one of the blended table rows, no comparison values will be displayed for any of the rows, even for those where the deltas do exist. Google is aware of this limitation and recommends that you do not show any comparison metrics when your data is expected to include null values. There is a potential workaround for this that we will get into in the next blog post.
Blending Pros and Cons
Google Data Studio has long been a go-to tool for data analysts and marketers alike, and now Looker Studio is positioned to continue to serve as one of the best free tools out there for easy data visualization. It’s not without its issues, but like with any complex tool, there are possible solutions for any problem. In upcoming blog posts, we’ll learn about different types of blending as well as dive deeper into some of the creative tricks for more complex reporting scenarios. If you are looking for help with creating new Looker Studio reports or maintaining your existing dashboards, please reach out to email@example.com.