The purpose of the thesis was to create advertisements on the social networks Facebook and Instagram for the company Maristella, which sells handmade products online, to analyze the results of these advertisements after the advertising time has finished, to determine which were successful and which were not, and to define the reasons for this.
The goals were to create content for the advertisements (photos and videos), correctly set up campaigns by synthesizing the information of the theoretical part, and determine which forms and styles of advertisements produced the best results using the experimental A/B testing method with four pairs of advertisements.
The results of the diploma thesis are created advertisements and confirmed or refuted hypotheses. We set link clicks as the ad performance metric. Two of the four hypotheses were refuted, one was confirmed, and one was advised to be tested again.
We found out
- that the ad with a longer text generated more clicks than the one with a shorter one,
- that the shorter video showcasing just one product performed better than the longer one featuring several products,
- that an ad with a photo of the product on the model generated slightly more clicks than a photo of the product itself, and
- that the results of the two ads where we tested video and photo as formats were similar. In this case, we neither confirmed nor refuted the hypothesis. Instead, we suggest retesting.
- We also recommend testing different creatives and ad settings repeatedly with a larger budget and a longer duration of ads to ensure the most realistic analysis and confirm or disprove our hypotheses definitively.
|
