Context Library & Benchmarking

This guide will walk you through the brand new Context Library and Benchmarks features.

Before you get started with the Context Library and Benchmark features, you will need to collect some images. These images are: your bundle of context background images and an array of previous content to benchmark against.

Once collected, upload your context images to the Context Library, and bundle these into Context Packs. Then, create your first benchmark using your collection of previous content. Publish your Pack and Benchmark to have them available across your account. These can then be pulled into any Study via the Context report side-panel.

Please read below for a more detailed guide of these features!

The Context Library is a brand new feature that acts as a repository for your Context background imagery. These images can be used account-wide for quick and consistent collaboration.

Click on the new 'Library' tab at the top of the page to access the Context Library. By default, the 'Contexts' tab should be selected in the panel on the left of the page. This section holds individual context images - these are tagged with Use Cases for easy organisation and selection.

To add a new Context to the library, click the 'Upload' button in the top right of the page.

First, assign your use case using the drop down in the panel on the left of the page. In this example, we'll choose Hero Image. Then, choose your file or drag and drop into the page to upload. Finally, edit your dropzone to highlight where your testing content will sit, and select whether to fill your background using the square at the bottom or leave transparent.

You should end up with an upload page similar to this:

Once this is ready, click 'Upload' in the top right of the page.

Your new Context should now appear in the Context Library with the correct Use Case attached. This can now be pulled into any Study, whenever you select the Context report type.

Context Packs act as a collection of context backgrounds for a Use Case. For example, an eCommerce Hero Image Context Pack would contain context images of the digital shelf for all relevant eCommerce sites for a brand; such as Amazon, Walmart, Tesco etc.

Again, these packs are account-wide and can be used to instantly and consistently test new imagery in a multitude of contexts. Context Packs are built using context images that have already been uploaded to the library. Ensure your chosen context images are already uploaded before you start to build your Context Pack.

To start, select the Contexts Pack tab in the bar on the left of the page. Then, to create a new Context Pack, click the 'New pack' button in the top-right of the page.

Name your pack, and select the relevant Use Case from the drop down list.

A new page will open and you will be able to select any Context Images that match your given Use Case from the library to add to the pack.

Once you've selected your chosen context images, click 'Save' in the corner to save the pack, this option leaves the pack in a draft state, allowing you to make further changes. The 'Save & Publish' option, will publish the pack, meaning the content inside cannot be changed. This keeps your data integrity secure, ensuring any analysis done using the Context Pack isn't affected. Only published packs will be available for use in Studies.

The final tab on the left is for Benchmarks. This tab allows you to create a context benchmark in the platform, allowing you to test in context stand-out for any new content, against high-performing previous versions, or competitors.

To create your benchmark, select the 'New benchmark' button in the top right of the page. Name your benchmark and select the relevant Use Case. Once created, select the Benchmark from the list to start setting it up.

Click the 'Upload' button in the top right of the page to add the assets to create the Benchmark. For this example, we'll use an array of Finish dishwasher tablet product thumbnails. Once your images are uploaded, select 'Publish'. Once published, you will no longer be able add or remove assets from the benchmark.

To test against your benchmark, create a Study, select the same Use Case as your created Benchmark, add your content, and select the Contexts report. Select a Context Pack to instantly test your new media in the relevant background environments. Click the trophy icon in the panel on the right of the page, and select the benchmarks drop-down. From this list, select your relevant benchmark. In this example, we'll be testing Finish dishwasher tablets, against our dishwasher tablet Benchmark, using a pack of eCommerce background images.

Your average chosen metric scores will be displayed against each version, along with the percentage difference to the Benchmark score.

Once a context pack or benchmark is published it triggers a process that analyses each asset from each published benchmark in each of the contexts across all published packs. The results of this are some pre-processed scores stored and ready to use in a study.

When you choose a benchmark in a study, you will have already selected a context pack. So now we have a unique combination of pack and benchmark to lookup the scores for from the pre-processing.

For each context in your chosen pack, the benchmark score is the average score of all assets from your chosen benchmark. This is done for all metrics so switching between them is seamless, as is changing your pack or benchmark.

You can then see the difference between your test asset scores and your benchmark scores for the given benchmark. Your test asset scores are calculated in real time.