Applied Marketing Data Science

The nerds are here to help.


Let us do the complex work.

Once you have all your data consolidated into a single location and have secured the buy-in from your leadership by harvesting enough low-hanging fruits from the integrated data, it is time to take data analysis to the next level. This includes:

marketing lead scoring

1. Conversion modelling & lead scoring

Score leads to either develop campaigns that guide them towards a desired action, or to use these predicted conversions to boost the bidding process with enriched data in various advertising platforms. Harvest significantly more revenue at lower cost with better optimized ad campaigns, made possible by predicting conversions at a much faster pace and in higher volumes than collecting the actual sales data. Lead scores can also be used for sales and marketing attribution.

2. Attribution, Incrementality and Marketing Mix Modeling

Optimise media spending using advanced analytics to determine incremental impact of various channels on conversions and sales. Our team can set up an attribution model using data from your digital channels as well as in-house data from offline channels. We use a wide variety of attribution techniques ranging from business rules (first click, last click, time-lapse etc) and statistical approaches (Shapley, Markov).

We also use Marketing Mix Modelling (sometimes called Media Mix Modelling) to overcome limitations in multi-touch attribution and achieve a comprehensive view of the contributions made by different sales and marketing channels.

marketing data science
Customer Lifetime Value

3. Predicted Customer Lifetime Value

Identify the most valuable customers (and who is likely to become a valuable customer) and focus efforts on revenue growth. Pitch LTV against Cost per Acquisition to get a more realistic view of long-term business growth driven by different advertising channels and approaches.

4. Data-driven segmentation

Use statistical methods to find valuable, actionable segments among your customers, to profile and understand these segments, or to use them directly in advertising audiences. Increasing return – and profit – on ad spend by utilizing your own understanding of customers.

market segmentation
market basket analysis

5. Market basket analysis

Identify upsell opportunities by identifying products that are more likely to be purchased when a certain set of one or more products is already in the basket. This can already be very useful as a once-off or an “exploration” basis, or it can be deployed as a scoring engine, used as a recommendation engine, or a component thereof. It is also called “association rule engine” and it can be used as a low-complexity statistical technique for a very wide array of use cases both in and outside e-commerce.

6. Other modelling / machine learning solutions

There is a wide array of machine learning and statistical solutions for various marketing and sales related business goals. Our arsenal also includes niche methods, e.g. the likes of social network (graph / link) analysis, forecasting, numeric optimisation, recommendation engines (matrix factorisation). We like simple solutions, but sometimes complex algorithms are needed to bring about the desired uplift.

machine learning solution

Making the complex digital landscape simple