How are my prediction scores calculated?

To explain how Junbi scores work, it's best to start broad and work our way down in detail. We can start with the overall effectiveness score.

How is the overall prediction score calculated?

The overall prediction score is the normalised average score of the three 'circle' (effectiveness drivers) scores. This means it is not just the average score, but it is a normalised percentile score - so for example, if your overall score is 70, you should be quite pleased, because you received a higher overall score than 70% of the ads in our benchmark.

However, it's important to note that this score is dependent on the Ad Type and Video Player Mode that you select for your prediction.

You can use this as a general indicator of how effective your ad will be at drawing the attention of your audience on YouTube. The overall score is visualised in quintiles, with the top quintile being ‘Excellent’ and the bottom quintile being ‘Bad'.

Now, we will take a closer look at the three effectiveness drivers, and how they are calculated.

How are the effectiveness drivers calculated?

Each driver is calculated slightly differently, however, it's important to note that there is always a raw score and a normalised score. The scores that you see in your Junbi predictions are normalised scores, meaning they have been compared to our YouTube ad benchmark database, and given a percentile value based on how well their raw score performs relative to the rest of the database.

Let's take the ad breakthrough score as an example:

Ad breakthrough is calculated as the percentage of attention on the YouTube ad itself (the area inside the red dotted line in the example above), relative to its surroundings (Thumbnails, comments etc). However, keep in mind that would be the raw score. The ad breakthrough score you see in the circle on your prediction is of course the normalised score. If you just saw the raw score, it would be very difficult to know if your score is any good! Unless you happened to have a database of thousands of ads which you had already benchmarked.

So we did the hard work for you. We take that raw ad breakthrough score and compared it against all the other ad breakthrough scores (tens of thousands) to create a normalised ad breakthrough score. So it will be much easier for you to know how you are doing compared to all those other YouTube ads out there.