We built a propensity to buy model that identified and prioritised who in Neptune’s customer base was most likely to make a new purchase.


Neptune, a home interiors retailer, wanted to increase the conversion rate of its website visitors by understanding the root causes underlying their motivation to make new purchases. It wanted to improve overall customer experience and determine the optimal customer journey.


  • We advised clients on which customer touchpoints are most effective for onsite conversion
  • Developed a bespoke algorithm to harmonise and dedupe the dataset
  • Merged customer and transactional data to determine the correlation between different products purchased
  • Designed a random forest model to identify which key customers characteristics correspond to repeat purchase(s).


Spotted and cleaned over 50,000 duplicates in the customer base (35% of the total)

Raised client’s awareness of which points and patterns of contact are most effective

Built a product correlation matrix to display in a visual way how well different products sell together

Identified several variables most predictive of ‘second purchase’ customer behaviours

More client stories

Connect with us

Fire us a message and we’ll be in touch

Contact us