Response Analysis – easy to do, so why is it ignored?

Generally, a business will put in a huge amount of work to win a client – pitches and proposals adding up a significant portion of dedicated resource. But once they start to use a businesses service, sometimes the dedication placed in winning the business seems to dwindle. Admittedly, this isn’t all the time, but the potential for it to happen is there.

In terms of data, data purchases / sales and predictive modelling, one of the easiest traps to fall into is to fail to fully “close the loop”. If a data services business wins a client with the aim of providing targeted prospect data or implementing a Customer Insight solution, invariably there would be a scoring process for the client’s existing data. This could be using a client’s own variables / values, an appended set of variables and flags, or a mixture of the two.

With the data scored, the next stage of the project begins. If its prospect generation, then the scores should reflect a prioritisation within the service providers own data source. If it’s an insight service, then the initial aim should be a schema for promoting and prioritising customers based on current & future campaigns and incentives. Either approach will provide a set of data that is intended for marketing purposes. These processes are fundamental and the core of what is proposed by the service provider, and bought into by the client.

But where to go from here? A fire-and-forget approach (as discussed on this blog previously) is easy, and hassle-free. It’ll result in some success and some failure. Because of the thought put into the process from the outset there’s also a chance that a client would come back to the service provider and ask for a repeat of the data file. And because of a shift in the data due to time lapsed, new prospects may appear, and old prospects may disappear.

However, this isn’t using effectual test-and-learn techniques; the process isn’t benefiting from valuable response analysis. If you can begin to understand why some responders agreed with the predictions, and why other responders seemed to appear out of nowhere, then a contact strategy can be finessed. In fact, there’s a strong argument to suggest that because a client has taken positive action in order to source or score their data prior to marketing, that very activity changes the make-up of the data over time – in effect, what you knew before may now be irrelevant, because an insight has driven a fundamental shift in strategy. Admittedly, a dramatic swing in a target market is quite unlikely! But the increased use of insight must be constant and self-training, rather than piecemeal and self-fulfilling.