Benchmarks are used by Vizit as a measuring stick, against which project images are compared.
A benchmark is a collection of “apples to apples” images that are evaluated and scored with respect to a specific audience lens, to see which are the most engaging, and which are the least.
After all benchmark images are run through our AI engine, each image is assigned a percentile score, based on how it compares with the other benchmark images.
The composition of a benchmark, and the number of benchmarks you create, will vary depending on your business goal. You may only need one benchmark, for a single analysis of your images in a specific category. If you have a wide variety of assets in multiple categories, you’ll want separate benchmarks for each analysis.
There are five key pillars of benchmarks we recommend:
- Category contain images that represent the entire category for the product type you are evaluating, including a broad range of competitor images. Examples could include best-sellers, top organic search results, or a selection of SKUs from a large market-representative set.
- Channel or retailer-specific contain images from a store or website, such as Amazon or Target. Within a retailer site, you can source images from an entire category, best-sellers list, or search engine results.
- Market or Region contains a set of images from one or more retailers, competitors, or categories from a specific market or region, like the US, UK, India, etc.
- Asset Type contains a specific collection of assets, such as hero, carousel, below the fold, or social media content.
- Brand contains a specific collection of assets from targeted brands, whether evaluated from a competitor or owned brand assets.
Within a benchmark of any category, you can further pinpoint your image comparisons by mixing and matching these 5 pillars. These could include benchmarks for:
- Face Wash - Amazon & Walmart - Carousel Content (Category + Retailer + Asset Type)
- Dry Cat Food - UK Amazon - Hero Content (Category + Market + Retailer + Asset Type)