By: Eric Siegel, Founder, Predictive Analytics WorldIn anticipation of his upcoming conference presentation, Model Based Fiber Network Expansion at Predictive Analytics World for Business Las Vegas, June 3-7, 2018, we asked Nishant Sharma, Director, Predictive Analytics at Charter Communications, a few questions about his work in predictive analytics.
Q: In your work with predictive analytics, what behavior or outcome do your models predict?
A: Our models are mainly focused toward:
- Identifying the right prospect/candidate to approach for sale for a specific networking product, which could be a cross-sell or a new customer. The prospect identification is also helpful in sizing up the market, resource allocation and territory assignment for sales teams.
- Determining the cost of expanding the fiber network in the footprint.
- Calculate the probability of customer churn to take proactive measures.
Q: How does predictive analytics deliver value at your organization – what is one specific way in which it actively drives decisions or operations?
A: Models based on predictive analytics help in determining which buildings would be the right candidate for fiber expansion, what is the right size of the market and how many sales reps should be allocated for a region, optimizing the marketing campaigns.
Q: Can you describe a quantitative result, such as the predictive lift of your model or the ROI of an analytics initiative?
A: One of the models that was created for selling a specific networking product showed that in just 20% of prospect universe, the model was able to acquire 75% of the total target. This result was the result of back-testing and not validation testing. That means, the list was created using the model in August 2017, and checked for performance in Dec 2017.
Q: What surprising discovery or insight have you unearthed in your data?
A: Using the model, we could show that our previous marketing campaign was reaching to only half the prospects. The model was able to identify the prospect based on their behavior and not vendor supplied filters, thus it was more impartial. This led to the data provider revisiting their database to find solutions and fix the data.
Q: Sneak preview: Please tell us a take-away that you will provide during your talk at Predictive Analytics World.
A: How to go over and beyond than just creating a statistical model and getting into the hands of ‘C Suite folks’- by creating a full story and adding more data elements to bring the model into production, and making the results more usable by merging it with existing platforms like SalesForce.
Don’t miss Nishant’s conference presentation, Model Based Fiber Network Expansion, on Tuesday, June 5, 2018 from 10:30 to 10:50 am at Predictive Analytics World Las Vegas, 2018. Click here to register to attend.
By: Eric Siegel, Founder, Predictive Analytics World