Marketing Mix Modeling for
E-commerce
%
Increase in ROI*
Increase in Sales*
Download the Industry's Toolkit
Toolkit includes: Data Request File, Marketing Mix Modeling Workflow, Industry Use Case & the Model Validation White Paper
Industry Overview
Challenges in Marketing
- Small data size
- Digital focused
- Lots of competition
KPIs
- Sales (by customer segment OR by SKU)
- Sales volume
- Quantity of orders
- Sign ups
- User value
Customized Solutions
Light Marketing Mix Model, Sales forecasting, Channel ROI Measurement, Marketing Budget Optimization, Marketing Data Management, triangulation with MTA and experiments results, CPA and ROI
Client Review
Frequently Asked Questions
Q: When building Marketing Mix Models, do you take into account seasonality and external factors?
A: Yes, seasonality and external factors such as weather, economic conditions, financial conditions, are part of our analysis. They are included into the marketing mix models as variables to allow us to identify how they contribute to sales (positively or negatively)
Q:Can you take into account the sample sent with products?
A: Yes, samples are represented as non-revenue orders. They are modeled in addition to the sales model and can be used to asses if samples influence sales. (I don’t think we dealt with samples before)
Q: What’s the difference between using MMM and MTA for measuring marketing performance?
A: Besides the fact that it requires aggregate data that is privacy compliant, Marketing Mix Modeling allows you to uncover the long-term effect of your media investments.
Q: I only have 1 year of data, can I still run an MMM analysis?
A: Yes, if you have weekly data for 1 year, we can use that as an initial point to start building your first Marketing Mix Model and then deploy a rescoring approach where we input the new data available on a quarterly basis.
Q: Do you have a tool I can use solely for optimization purposes?
A: Since we know that optimization is a major need for e-commerce companies, we have developed a dedicated Marketing Optimization Software called “Insights” to be shared among teams. It allows for revenue and budget optimization while giving you the liberty to adjust periods for seasonality and build your own constraints.
Q: How can the impact of organic search and branded paid search activity be accurately assessed in E-commerce models without overestimation?
A: Given that these channels are endogenous and susceptible to influence from other factors such as media activities, Mass Analytics has implemented a modeling technique known as nested modeling. This approach enables analysts to construct a separate model for these key performance indicators (KPIs) and nest it within the main model. Consequently, organic search and branded paid search activity won’t be overestimated, and due credit is given to other contributing channels.
Our Clients
