Goal
For many e-commerce companies, there is one very important factor that drives both costs and sales: marketing. But where do you as a company have to spend your marketing euros? Adwords? Facebook? Or a TV campaign? How do you find out in the complex landscape of mixed marketing campaigns how much effect every euro really has? And at what point investing in a particular channel has no effect anymore?
Motivation
Many marketing managers can measure the impact of individual marketing campaigns, such as conversion of an Adwords campaign. The indirect effect of a TV commercial on turnover is already a bit more difficult to measure. But how can you reliably mesure the effect of numerious campaigns? The different campaigns also affect each other. It is therefore essential to know that the TV commercial does not contribute enough, so that the budget can go to an extra Facebook campaign next time!
Result
The marketing manager chooses a performance indicator: revenue, or perhaps even brand awareness. She gives a budget and with the push of a button an ideal budgeting follows the different channels that the team can use.
Approach
We use a unique Bayesian model that can take into account all the different influences that come with marketing. First of all, the influence of different campaigns is analyzed. But the model also corrects for any delayed effects of campaigns on performance indicators and the ceiling on the impact of campaigns. These are elements that are very difficult to analyze with other techniques.
Solution
Roadmap
Campaign optimization
By making an in-depth analysis after the optimization over channels also specific campaigns can be optimized. This allows the marketing manager to analyze which brands or products have the most effect on her performance indicators.
Result
Result
Research online purchase offline
Customers do not always purchase online. The online channel is also used for research, after which customers complete their purchase offline. By identifying these purchases, the effect of online marketing activities on offline sales can be analyzed.
Result
Result