“I want to identify our customer profiles to do targeted marketing campaigns.”


That was the business problem posed by an online company which lets thousands of their customers send personalized gift boxes. For years, we have been helping this company with its data warehouse. Millions of records have been aggregated for churn calculation, average spend, inventory turnover, etc. to provide excellent insight for management to make data-driven decisions on stock purchasing, sales incentives, bundle promotions. But to compete in today’s retail industry, the company realizes how important it is to tailor their campaigns for individual customer preference and spend their marketing dollars wiser. Gone are the days when the department compiles a few general campaigns and see which sticks. Presence Consulting was asked to develop a model to show where the customer segments are based on their purchase histories for the Chief Marketing Officer (CMO) to design targeted campaigns.


Data Analysis


We collected as comprehensively as possible all the behavioral (mostly transactions, click-path) and promotional (Ad words, email campaigns, etc.) data. We cleansed it, took out irrelevant pieces (we call this feature selection in DS), ran a number of algorithms in R, and built a clustering model. We located where the customer clusters are and formed different personas. The marketing group was able to design campaigns with much greater effectiveness. After a couple of launches, we collected more relevant data and fine tuned the model. Not to mention the company also identified missed opportunities and built strategies to capture them.

We are able to achieve a 16% increase of conversion rate in targeted email campaigns.




Do you have a boat load of sales, inventory, customer, or operational data you wish to get more insights out of?

Give us a call, perhaps we can offer a few ideas.