A curious incident occurred with our analyst. From a well-known Ukrainian company that sells clothes, he received a newsletter with an offer to buy a dress. The work of the entire analytical Department was suspended for at least an hour, as the staff laughed uncontrollably, imagining 2m tall Alexei in some light summer dress.
We wondered how much money “leaked” online store, saving on verification of its customer base in just one parameter, not to mention the analysis of the history of previous purchases?
Forecasts regularly performed by our analysts, allowing you to quickly and easily find out: what to offer to the buyer, when and how. And most importantly – what not to offer. After all, the behavior of people in the past is the basis of actions in the future. Since the contact center is a place where much data about consumer behavior, about real purchases, orders, questions, and complaints are accumulated, it is possible to determine how consumers will act in the future.
Let us explain with an example. The largest global seller of dietary Supplements reactivated its distributors through telemarketing in the contact center. The analysis showed that distributors in Western Ukraine with Kyivstar mobile numbers are reactivated by 30% more than distributors in this region with Vodafone numbers.
Conversely, in the Eastern part of the country distributors with Vodafone numbers are reactivated by 40% more than distributors with Kyivstar numbers. This analysis helped to predict conversion by region and mobile numbers and to discard those who are less likely to be reactivated. Thus, expensive telemarketing removed the bulk of the costs-calls to those who do not need it. As a result, the conversion of reactivation was +15% compared to last year, and project costs were reduced by two times.
It is possible to predict purchases and responses to advertising messages, reactions to change the properties of the product. You can use data to make management decisions across the company, and you can define a segment for each individual consumer and build personal communication. Arsenal of tools – from the simplest RFM-segmentation and RF-grid to cluster analysis and other data-mining tools. Then the consumer will receive an interesting message via Viber, SMS or e-mail.
The most interesting conclusions appear when data of different nature from different sources are combined. The purchase affects the time of year, time of day, weather, human mood, the situation of the child at school, success at work and even politics. We combine all available data in the analytical process.
How does data analysis help the contact center work?
Let us look at the example of one of our projects. We were approached by an online clothing store with the task of reactivation to stimulate the receipt of orders from customers who have not bought anything for a long time. The customer’s internal CC also worked on this project, but the management was not quite satisfied with the result – rather high costs for the return of one client.
To reduce costs, DirectCall’s primary goal was to increase conversion – the number of reactivated clients relative to the original sample.
The implementation of the project began not at the operating site of CS but in the Analytics Department. We segmented the database to determine the most efficient time to make a call. For the analysis, we used statistics on previously made calls and identified a set of factors in which the response was maximum.
For segmentation, we used gender, age, a region of residence, operator, number of purchases and expiration date of the last, the results of previous calls.
At the same time, the most effective objections of consumers were worked out. We segmented the database and determined the optimal set of objections for each segment. As a result, the operator was provided with dynamic prompts with recommended answers for each consumer. In addition, the operator made a call to the consumer at the optimal time and with convincing arguments.
As a result, the cost of returning one consumer decreased by 2 times. We would not be able to achieve this result by making a standard base call, applying standard dial schedules, and using a common “set of phrases” to overcome objections.
When we find a useful pattern in the data set, such as sales growth in the rain, and no one can explain why it happens, our customers call it magic. We believe that we can and should use the existing pattern not only for the wow-effect but also to increase conversion.