| Title |
Modelling return on advertising spend. subscription-based b2c business case study |
| Translation of Title |
Reklamos išlaidų grąžos modeliavimas. Prenumeratomis prekiaujančio verslo atvejo analizė. |
| Authors |
Budvytis, Armandas |
| Full Text |
|
| Pages |
48 |
| Keywords [eng] |
media mix modelling, bayesian methods, ordinary least squares, marketing |
| Abstract [eng] |
Applying the model form proposed by Google[3] a media mix model containing the carryover and saturation effects of marketing spend was estimated at the daily level on a dataset containing marketing spend and revenue for a period of two years. The models fit the data well and a strong carryover effect was found in the influencer channel, in particular, integrations with YouTube content creators, and a smaller one in the affiliates channel. The saturation effect estimation failed due to the poor identifiability of the function proposed for this task. When comparing the ROAS estimations, the last-touch attribution was found to be better suited to model the influencer channel. |
| Dissertation Institution |
Vilniaus universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2024 |
