Analytics cannot tell you how people in general will behave in the future, but it does tell you how your customers have behaved in the past. Businesses spend a great deal of money on market research. Even so, many mistakes are made, not least because of the risks of sampling error, of asking the wrong questions, or because there will be factors in real life that may not be anticipated. In contrast, the current customer based will often provide a larger population than any that could possibly be assembled through another kind of research exercise, and will also permit insights to be drawn over a much longer period time. Current customers are an authentic representation of real life. To effectively exploit this valuable resource of information, you need two things. First you need the ability to collect, store, mine and analyze the data. Then you need to know what questions to ask.
Collecting, storing, mining and analyzing data. Does that sound familiar? Sounds like the kind of thing you need to do for revenue assurance. There will be strong feelings about whether revenue assurance people should concern themselves with helping sales and marketing. I myself made the argument that revenue assurance works “without influencing demand” in the TM Forum’s definition. The point of that statement was to exclude the promotion of increased customer sales as an element of revenue assurance proper. However, whether or not it is revenue assurance, it will offer an opportunity to revenue assurance people to get increased value from the tools, skills and data they already possess. If they can do it, and if they are finding that returns from addressing leakage are diminishing as a result of their hard work over time, is that a bad thing?
A business may possess a lot of data, but data is not the same as information. For data to become information, there has to be a use. Knowing that families have 2.6 children on average is not useful, because no family has 2.6 children. By the same token, inferences about “most” people, the “average” and all that are probably just convenient fictions. They make it easier to think of how large numbers behave, without really understanding how each and everyone in that large number is actually behaving. Successful analytics involves drawing conclusions from actual individuals, by cutting the data based on theories of how to group individuals into coherent groups that tend to behave the same way. This makes marketing both personal and relevant.
One of the most effective uses of analytics is to tailor prices and promotions to drive increased sales and improved margins. Provide a selection of customers with a better rate than normal, and see if the increase in sales justifies the lower price. The connection between data and customer is already established for postpay customers. The rise of mixed postpay/prepay plans, or offering enhanced services and special offers via the internet can be ways to increase the span of knowledge to prepay customers. Future promotions can be targetted at the customers that are most responsive. This will enable the business to segment its customer base according to the different utility curves of different customers, and hence refine its pricing strategy and offerings accordingly.
Analytics can be used to assess the relative profitability of competitor’s pricing schemes and incentive programs, or to perform hypothetical what-if analyses of proposed new prices and incentives. Particularly where there is highly stratified pricing, with rates decreasing as consumption increases, comparing new prices and pricing points to current customer behaviour will help to understand the potential for revenue growth or cannibalization depending on customers opt to buy more or less. Overlaying the volumes of sales per individual customers with competitor’s pricing will identify which customers would be better off if they switched to competitors, and which customers are benefitting from the best available deal at present. This may help with identifying price reductions which would draw price sensitive customers away from competitors, and also price rises which would still better the offers made by rivals.
Adding data on cost of sales changes the focus from revenues to margins and profits. In a similar way, it is possible to include data on the timing of cashflows, say from bulk purchases of ‘bolt-ons’, to gauge how these can be improved through understanding the segmentation of the customer base. Any numerical data on costs and revenues that can be associated with individual products/services and with individual customers can provide a rich basis for comparison to competitors and pre-assessing the impact of proposed changes to prices and offers.
Customers that superficially seem profitable may be viewed differently once all costs are taken into consideration. Time spent handling customer complaints, or a track record of returning goods, may indicate the customer is more costly to serve and less desirable than originally thought. When identifying which customers to offer loyalty benefits to, it is worth directing these benefits to customers that are cheapest to serve by virtue of the smaller demands they place on the business. For example, there may be incentives for customers to purchase on-line because of the lower cost of taking the order compared to processing a sales order through a call center or in a store. It makes sense to extend that logic by and prioritizing customers that place a smaller burden on the business, for example because they submit orders in a way that consumes less staff time or because they raise fewer customer service queries.
Revenues and costs are readily susceptible to analysis because they are numeric and because it is relatively simple to gather the necessary data and associate it with products and customers. In practice there are many ways to measure and segment customer behavior, and hence look for trends and groupings within the customer base. Geography may be a factor in sales and costs, thus enabling different strategies for different locales. Geography can be analyzed by either the customer’s home address or, in the case of mobile phones, where the customer is when the phone is switched on. Demographic factors like age or cultural leanings may also provide a viable basis for analysis and segmentation. This kind of data might be obtained via credit checks or by asking customers to submit to a survey.
What data will be relevant depends entirely on the product on offer and the nature of human behavior. Generalization from past experience may be useful, but will be misleading if nobody has considered some options for altering the offering or segmenting the customer base. No amount of data will assist in drawing useful conclusions if the wrong questions are asked. That takes the skill of the imaginative marketeer. But harness that insight to data and the power to perform analysis, and, like a scientist, the skilled marketeer can progress from forming a theory to being able to corroborate it in practice. This will lead to better decisions made with greater confidence and improved understanding of the results that are subsequently achieved. Imagination combined with data makes for good business sense.