Truth be told, you don’t need a Ph.D. in data mining or statistical analysis to reduce churn. Which is good news.

Truth be also told: investing tons of money in data mining tools does not guarantee success – more often than not, it guarantees failure.

Churn reduction is a complex process.  There are many reasons why it can go wrong, and from our discussions with telecom service providers, it seems to go wrong most of the time. Why is that?

Reason #1: the data that you are using are of low quality

Reason #2: the data mining techniques that you are using are not working

Reason #3: the retention campaigns that you are deploying are not working

Reason #4: the retention campaigns are too expensive i.e. the costs of retention outweigh the benefits

Reason #5: there are too many things to fix in your business before you can seriously think about reducing churn.

Fortunatly, there are solutions to your problems:

Fix Problem #1: buy a good data warehouse; or use a software solution for churn reduction that does not rely on your ‘crap’ (excuse my French) data warehouse but extracts the data that itneeds directly from the billing and CRM system, and store these data in its own database.

Fix Problem #2: stop using regression models that do not work; use self-learning models that remain up to date all the time and keep improving themselves, and use the complete history of each individual customer in your business.

Fix Problem #3: stop using data that are one week or one month old, use real-time data than are at most one day old.

Fix Problem #4: stop addressing the wrong customers, those that did not plan to churn in the first place, and focus on the customers that are most likely to react to your retention measures.

Fix Problem #5: stop encouraging customers to throw their SIM cards away after 2 weeks of usage; improve the quality of everything that your do in your business, especially all the customer touch points.

Fortunatly, there are solutions on the market that can help you.  One of them is Lumata. It does not need a dataware house. It’s got self-learning predictive models. It works in real-time, and mostly automatically. It provides daily performance reports of your retention campaign. And it has proven itself in the field by generating massive churn reduction and revenue increase for the operators where it has been deployed.

It is time to ACT. Read this white paper

Download “Self-learning Live Predictions” ACT750-SelfLearning-Live-Predictions-DataSheet.pdf – Downloaded 438 times – 500 kB

if you are committed to reduce churn today. And contact us at Investaura or contact our partner Lumata to learn how we have helped our clients reduce their churn level and increase their ARPU.