Today’s post is from Matt Lindsay, President of Mather Economics. Mather is a data analytics consulting firm serving clients in many industries around the world. Mr. Lindsay provides concrete steps for any video company looking to leverage customer data to improve the performance of their business.
Many companies launch into a customer analytics journey without a clear objective in mind. The goals are simply to better understand their customers or to find ways to work smarter. The truth is that there are two parts to any successful application of customer analytics: the process of capturing, processing and analyzing data to arrive at a result, and the implementation of the findings in an operational process.
In our experience, the latter of these two activities is the most challenging due to operational and system limitations throughout the organization. For this reason, we recommend companies that seek to apply customer analytics start with the desired end in mind. In other words, identify a point of leverage in the business where the application of analytical insights can be implemented within an operational process and the supporting IT systems to affect customer outcomes, and then work backwards from there.
Identifying a potential application of analytics possible within an organization will determine which analytical methods should be used, what data is necessary to support the models and what type of insight can be implemented.
For instance, can the billing system handle different product bundles with the associated messaging? Can we get data on how customers have reacted to bundle offers in the past to estimate a logit model? Can our customer service center handle the complexity of the new offers? And most importantly, does a successful initiative provide incremental revenue sufficient to justify the investment in time and resources?
Start with the end in mind
In many cases, starting with the end in mind enables quick victories and defeats. These initial efforts should require relatively minimal investments in data infrastructure and resources. Many of the required resources can be outsourced, such as using cloud-based servers for data storage and analytics, which will help make a more informed purchase decision in the long run.
Significantly, these small-scale initial projects will provide invaluable insights about the viability of larger-scale efforts before big ticket investments are made in people and technology. Even if these first attempts at customer analytics do not provide a large ROI, they will pay for themselves in the long run if they help you avoid mistakes down the road.
Move from one-size-fits-all
For many OTT video companies, customer relationships are ongoing, meaning customers are acquired, retained, invoiced, serviced and up-sold during their lifecycle. This type of repeated interaction enables an accumulation of data on a particular customer, which will allow several types of analytics to be completed.
Companies can test how various acquisition offers perform not only with initial acceptance rates, but also with retention. We have found that relatively simple adjustments to retention processes can save many customers that otherwise would stop their service. In many cases, moving from a one-size-fits-all process to one that is tailored to the preferences or particular needs of certain customer segment can have material improvements in customer outcomes and profitability.
For instance, we find that most companies have a single late payment messaging process that is pretty unpleasant for the recipient. In many cases, a customer with a current late payment has had many prior payments made on time. There may be a credit card renewal issue or a family event that has changed their typical payment patterns. Sending a harshly worded request for payment is not helpful for a long-term relationship.
In a similar case, what if a customer always pays a few days behind their due date? Should they be sent a late payment notice every month? Maybe they are paid on the 15th of the month, but their bill is due on the 10th. Identifying these patterns and designing business processes around them is often an easy opportunity to show the power of customer analytics to improve business performance.
Applying the target and control tests
Another tool available to customer analysts are target and control tests of alternative business tactics. These are particularly helpful when historical data is not relevant to a new project. In one case, we worked with a customer to test various retention campaigns.
We divided customers into three churn risk groups, and we tested three offers of varying cost. The results of the test are summarized in the graph within this article. Interestingly, the least cost retention offer, the $1 greeting card, had the greatest effect on churn for two of the churn groups. This type of test can be accomplished with relatively little in the way of capital investment and data integration.
The good news is that the first few steps in applying customer analytics do not need to be overly complicated or require significant investments. The better news is that once a few quick wins are in the books, greater demands and bigger budgets for customer analytics will follow.