Oh no! Sorry to hear that you entered a subject line into Predict but couldn't generate a prediction. Below is a list of a couple of reasons why and what you can do to change the subject line so we can estimate performance.

Reason One: Your Subject line Is Too Short.

If your subject line is too short, then we won't have enough information to make a prediction. If that's the case, you can simply lengthen the subject line by adding more emotional or descriptive language, OR you can try selecting one of the suggestions within the Prediction. See the example below:


"This cannot wait" contains fascination, which actually performed well for this dataset, though a fascination phrase alone may not be enough to generate an estimate. Thankfully, if you can't generate a prediction or get a negative uplift prediction, Predict will suggest adding one or more high performing phrases.

In the example above, Predict suggests the following three phrases: "report", "hot deal", and "it's all about you". I chose "it's all about you" and generated a new subject line, "This cannot wait - it's all about you" I re-Predicted and successfully generated a predicted uplift of 6.62%.


Reason Two: Your Subject Line Contains No Emotion.

If your subject line contains no emotion, or if the description is too unique like in the example below, we may not be able to generate a prediction. Again, if you try using some of the phrases Predict suggests, you may be able to generate a positive prediction.


I simply added "Because We Appreciate You", which is fascination, to the subject line, and was successfully able to make a prediction.


Reason Three: We Don't Have Enough Data.

Let's continue with the example above except let's try an emotion which we haven't tested in that dataset. If I used a Challenge phrase as opposed to a Fascination phrase, I can't make a prediction (see screenshot below).

No prediction with Challenge language

Why? Because challenge was barely tested in this dataset. Not only that, but "goatastic yogamats" is a completely unique product. We don't have enough historical data on those emotional and descriptive phrases.

You'll need to try another emotion. If you need some inspiration, you can find some in your dataset!

Go To Your Dataset

Click on any emotion in the Emotional Insights Graph

This will open up a menu beneath the graph containing all of the subject lines in the dataset which contain that emotion. In this example, we chose exclusivity. Within each subject line, the underlined phrases are the exclusivity phrases.

Add an emotional phrase to Predict

Great! Now we have a positive prediction for this subject line. Keep trying these methods above to create a variety of subject lines. If you run into any issues, please email prosupport@persado.com!

Did this answer your question?