Driving customer retention using predictive analytics
0 likes946 views
Check out this presentation from Qubit's Predict Analytics Webinar where you will learn how to use predictive analytics to solve for customer retention. Learn what predictive analytics is, and how you can utilize the data you already collect about your visitors to predict intent, and optimize the experience. Essential is being able to identify the risk factor of different users and personalize their experience to make them stay.
Ad
More Related Content
What's hot (20)
Viewers also liked (9)
Ad
Similar to Driving customer retention using predictive analytics (20)
Ad
Recently uploaded (20)
Driving customer retention using predictive analytics
- 1. Predictive analytics (Webinar 1) Customer retention May 19th 2015
- 2. we look at simple ‣ What are we trying to solve? ‣ How can we use predictive analytics? ‣ What data are we talking about? ‣ Demo ‣ Q&A David Hitt Director, Strategic Accounts Qubit Stephen Pavlovich CEO & Founder conversion.com Agenda
- 3. Qubit is a global leader in digital optimisation Global • Segmentation • Analytics • Personalisation • Testing
- 4. we look at simple What is the retention problem? 80% of future revenue will come from as little as 20% of existing customer (Gartner). Yet most companies don’t have effective retention programs. Consider conversion rates online which can be as low as 3% for new visitors. Existing customers on the other hand can renew at volumes greater than 80% depending upon the brand and industry. If we look at a wireline provider who tends to have a churn rate of between 2 - 2.5% per month even with a modest customer base of 5 million, that means an estimated 1.3m customers or $2b in revenue is lost every year. Frightening numbers. The revenue opportunities associated with an increase in retention can be massive. EXAMPLE
- 5. What are the causes of churn?
- 6. What is predictive analytics? The practice of audience profiling by utilising existing data sets to determine patterns and predict future outcomes and visitor intent ‣ Predictive ‣ Descriptive
- 8. Stage 1: Information gathering Journey begins on a comparison website accessed on a tablet.
- 9. Stage 2: Purchase process User selects the right product, and completes the purchase online with the selected provider.
- 10. Stage 3: In-life customer support Customer will have a number of experiences during the period of service that will impact their propensity to churn or renew.
- 11. What data can be collected and analysed? ‣ First party digital: ‣ Device ‣ Location ‣ Channel ‣ Products ‣ Price ‣ Quotes ‣ Current providers ‣ Subscription length ‣ Browsing history ‣ Keywords ‣ Demographics ‣ First party: ‣ Contact history ‣ Claim history ‣ Fault history ‣ Billing issues ‣ Address changes ‣ Third party: ‣ DMP ‣ Credit score ‣ Geographic
- 12. EXAMPLE
- 13. Qubit Decipher Dashboard using Tableau for visualisation using dummy data. Using predictive analytics we can select a channel (highlighted) and determine propensity to churn (risk) Red is churn, green is renewal. Here Direct has a very low chance of churn, but vertical search (highlighted) is high. Highest risk channel Lowest risk channel
- 14. Qubit Decipher Dashboard using Tableau for visualisation using dummy data. In the Vertical Search channel, we can click on a specific affiliate and drill down a list of contacts who entered the site through that channel, and determine their risk of churning
- 15. Qubit Decipher Dashboard using Tableau for visualisation using dummy data. Here we can click on Paid Search, and see the risk associated with certain key words. From there, we could also drill down the visitor information to determine which visitors are at risk. ‘Cheap Car Insurance” in this example has a 30% chance of churning, so we would adapt our marketing spend accordingly.
- 16. Qubit Decipher Dashboard using Tableau for visualisation using dummy data. Next we can go to the Visitor Page Dashboard to analyse how a user’s actions on the site 30 days after signup can signal their intent to churn. In this example, we see that homepage visits have a low risk of churn, where as FAQ has a high risk. So, users visiting the FAQ page within 30 days of signing up represent a great opportunity to reduce churn. Highest risk behaviour
- 17. Qubit Decipher Dashboard using Tableau for visualisation using dummy data. Finally, we can have a look at all the data to determine the size of our problem. From there, we can drill down on entry points to determine which present the biggest risk, and exactly what that risk is. This analysis will inform what brands, affiliates, keywords etc we should be spending our marketing budget on.
- 18. we look at simple Personalising the experience The next 2 slides we look at how to personalise the experience for at risk users. 1) Personalising by cross-selling a product that reduces propensity to churn 2) Targeting users on the FAQ page with a rang of personalised and relevant discounts to lock them in and reduce churn.
- 19. Personalised experience to drive retention Standard Personalised
- 20. Personalised experience to drive retention Standard Personalised
- 21. Example recap Predictive Analytics Onsite optimisation Visitor Cloud
- 22. The business value ‣ Customer retention ‣ Stop existing customers from moving to the competition ‣ Improved targeting of offers ‣ Know who to target with which offer based on their individual score ‣ Improved loyalty and willingness to recommend ‣ Positive word of mouth driving more conversions ‣ Increased customer lifetime value ‣ Increase average subscription period and value ‣ Optimise the partner/affiliate programme ‣ Manage partners based on true conversion value