Spandex | Predictive Analytics for Revenue Growth
How Spandex increased sales margins and customer retention using predictive analytics and machine learning
Spandex is one of the world’s leading suppliers in the printing and graphics industries, with a huge client base and wide product range.
In a highly competitive industry, Spandex wanted to protect and grow its market share against competitors’ pricing tactics. The Spandex sales team were effective at managing their biggest accounts to prevent customer churn and identify opportunities to cross-sell and up-sell but were unable to give all accounts this level of attention. This left Spandex vulnerable to customer churn and missed opportunities for profitability.
Syntagium worked with Spandex’s marketing team to develop a model that analysed millions of rows of customer transaction data, looking for patterns and anomalies in customer purchase behaviour. This included what they purchased, how often they purchased, what their average spend was, and the time since their last order.
We identified how product choices had evolved over time, and the shared attributes of loyal customers – the key to increasing customer retention.
These insights were fed into a machine learning algorithm to produce daily predictions of the future behaviour of every customer. The model predicted the likelihood for each individual customer’s next purchase date and product selections, and these predicted behaviours were then assessed against actual behaviour so that the model improved over time – AI in action.
An automated email campaign was implemented around the model, designed to:
- Nudge customers who missed their usual ordering pattern
- Deliver tailored, product-specific discounts to customers close to buying
- Cross-sell and up-sell by offering customers complimentary offers after a purchase
- Deliver customer insights and action items to sales reps’ inboxes
- Improved sales margin
- Increased customer retention and customer engagement
- Deeper insights into customer segmentation