Problem solve Get help with specific problems with your technologies, process and projects.

Predicting churn - worth it?

I question the effectiveness of churn modeling for the telcom industry. Subscribers churn for reasons such as their contract is up, they want a new phone, they have bad coverage in their area, they are on the wrong rate plan, or they move. Predicting churn in these instances is of little help in preventing it. In fact, telling someone their contract is up may only remind them to switch to another carrier.
I couldn't disagree more strongly, so I'm going to be very direct in my response.

The 'conventional wisdom' issues you raise are typical of what the majority of marketers in the wireless telecom industry believe; and it explains, at least in part, why the industry rates of churn are so high.

You may challenge the usefulness of predictive churn models in your industry, but I'd invite you to closely evaluate the impressive results achieved by many companies. Predictive models work - if they're rational, comprehensive, current and actionable.

So far this year, I've seen effective prediction churn models used by telecoms in France, Switzerland, U.K., Italy, and other countries. Some models are simple, some are more complex; but companies use them for action. They apply the results of these models for early intervention - correspondence/communication, product, plan, service - resulting in significantly reduced churn.

Invariably, these models depend upon:
- A well-developed customer business strategy
- An end-to-end management process of customer knowledge/relationship
- As detailed and in-depth information about customers, and customer profiles, as possible, leading to tight segmentation
- Frequent updating of customer knowledge and predictive segmentation
- Flexible marketing actions with customers
- Ongoing evaluation of action results, i.e. impact on the business

Let me share some insights presented by a colleague at Bell World in Canada, at wireless telecom customer loyalty conferences held this year in London (February) and Vienna (June). With 1.5% monthly churn, his organization is the North American leader, and they are continually striving to lower this level. They understand the factors you've pointed out in your question; however, they view them as opportunities to reduce churn rather than an admission of defeat.

For example, Customer Service is encouraged and empowered to be proactive with customers. They dialogue with at-risk customers to isolate problem issues. They review each customer's profile to determine how the rate plan might be modified to make it more attractive. They compare, in real time, Bell World's offers with what the competition provides. Information from these customer interactions goes directly to a 'Churn Desk' and the Marketing Department for joint review and action.

Over 700 CS reps have been trained in churn handling. Bell World has implemented, as well, an aggressive churn reduction program which includes hardware upgrade, application of a high percentage of annual per customer revenue for retention offers, an attractive incentive plan for the CS reps, and a comprehensive service process improvement program. Finally, Bell World has implemented a CS 'workbench', in which reps are provided with both the current predictive churn model and feature recommendations for them to consider.

Bell World also does a great deal of brand identity and value building. They see the brand as a highly critical companyasset, to be enhanced and protected over the long term. Unlike rate plans, coverage areas, contracts, equipment and the like, brand image is one of the intangible elements of value which create, or undermine, true customer loyalty and advocacy.

The loyalty, and prospective churn, impact of both the tangible and intangible elements of value can be determined through formal qualitative and quantitative customer research. In addition to predictive churn models, these research methods can be used to help anticipate customer turnover. They include evaluating the impact of expressed and unexpressed complaints, and setting up targeted loyalty indices.

Dig Deeper on Customer loyalty and retention

Have a question for an expert?

Please add a title for your question

Get answers from a TechTarget expert on whatever's puzzling you.

You will be able to add details on the next page.

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.