In a previous post we discussed the results of an AB test of the subject line carried out in different countries. On that occasion we tested whether the personalization of the subject line with the user's name affected the ratio of openers and the click to open. We saw how the increase in openers ranged from +44% in the best case, to +4% in the lowest case. We also saw how in Mexico was where personalization generated a greater increase in both openers and click to open.
In this post we present the results of another A/B test also carried out in Germany, France, UK, Mexico, Portugal, USA and Spain. On this occasion, we tried to find out whether the inclusion of a symbol in the saudi arabia mobile phone number list subject line affected the same metrics . In the following image we show the results of two different campaigns in which the test was carried out. As we can see, the inclusion of a symbol did not cause a significant variation in any of the cases. Why does the personalization of the subject line with the user's name generate higher interaction rates and not the inclusion of a symbol?
The interpretation that we at Digital Response make is that elements that appeal to the “person” have persuasive capacity precisely because they produce a climate of direct “dialogue” or “conversation”, 1 to 1. Emoticons are an ornament, something that can “attract attention” aesthetically speaking, but that do not trigger, from a psychological point of view, higher levels of persuasion.
The main lesson that we can draw from this exercise is that “connecting” with the user, either by establishing personal recognition codes or by including content adapted to their interests, is the best way to obtain better responses. In any case, we must recognize the role of the “context” that surrounds the entire action: the notoriety of the brand, the general engagement of the users of the database with the brand, the frequency of the campaigns, the seasonality, etc. In this sense, there are some similar exercises published that yield different data. We believe that it is precisely the different contexts that can explain the different results.