Every image added to a benchmark is automatically assigned an audience score for every audience defined for the projects that contain the benchmark. This score reflects predicted audience engagement for the image.
Note: This scoring process also occurs with project images.
Once the range of benchmark image scores is established, the scores of each benchmark image are placed on a 0-100 distribution scale. The images with the lowest engagement scores are at or near the zero end, the images with the highest score are at or near the 100 end.
The benchmark scoring scale is presented in a red-to-green color format, using the following scoring “battery bar” system:
Higher scoring benchmark images are those more likely to visually engage your audience, when compared to the entire set of benchmark images. These are the assets you should look to as a reference for improving your own images.
The median value is the middle point: half the benchmarks images have lower scores and half have higher scores.
The median is a good indicator of the overall audience success of benchmark images. In the example above, there are more benchmark images with low and very low scores than those with high scores. So the median score - the overall audience engagement with these benchmark images - is relatively low.
The benchmark scoring is used for predictive analysis, as a measuring stick for audience engagement of a project image. Like with benchmark images, each project image is also assigned an engagement score for each audience. Where this project image score falls along the benchmark 0-100 range is the image’s Vizit score.
In this example, the project image was more visually engaging than 38% of the benchmark images (or less visually engaging than 62%, depending on your perspective).