The situation
A software company for the events industry ran a customer advocacy programme. Happy customers were invited to take part by sharing their experience, which is some of the most persuasive marketing a company can have.
The problem
Not every customer says yes. The programme was using a marketing model to decide who to approach, but only about 2.5% of those approached went on to take part. The aim was to spend less time on customers who were unlikely to respond.
What we did
The team built a new model to estimate, for each customer, how likely they were to take part. It used what was already known about each customer and learned from the customers who had said yes before.
The programme could then focus its effort on the customers most likely to say yes.
Alongside the model, the team built an in-house tool so that the model could be refreshed, and new ideas tried, without long manual work.
The outcome
Conversion on the programme rose from 2.5% to 4%, which is about a 60% improvement in how well it performed. The tool also reduced the effort needed to build and update models by around 75%.
What it meant for the business
The advocacy team got more results from the same list of customers and spent less of their time on long shots. Because models could be refreshed quickly, they stayed current as the customer base changed.
Could this work for you?
If you contact customers or prospects in volume, whether to renew, upsell or ask for a favour, a simple model can help you put your effort where it's most likely to pay off.
Related services: Predictive modelling and forecasting, Marketing analytics
This project was led by our founder, Achyutanand Jha, in an earlier role. The client is not named, and figures are those reported at the time.