Blog: How data can enrich your marketing strategy

Timing your customer contact strategy

Author: Rinke Klein Entink
Rinke Klein Entink

Published on September 05, 2016

The athletes who participated the Olympic Games in Rio know this all too well. For a top performance, every movement has to be in precise flow. Milliseconds matter when you want to win gold. The same is true for your customer contact strategy.

You want to contact your customer with the right message, at the right time, because only then will you truly be of service. But how do you know when that right moment has arrived? In this post, I will give an example and some steps to follow to improve your timing towards your customers. This may help you in your customer retention programme, to maintain share of wallet and to improve customer experience.

Suppose your company offers three products and you serve three types of industry. Should your timing in contacting these industries be different? That will depend to a great extent on the life cycle of the products. When the three products are consumables with similar life cycles of approximately a month, it’s easy: You know you want to contact your customer after three weeks so you can re-supply them on time. But what happens when the product life cycles are very different? You contact your customer and the response you get is ‘Yes, I do need office supplies, but you keep bothering me about those machines I don’t need now’. That’s annoying for your customer and he may question if you really understand his needs. Things can get even more complicated when you add the diversity of industries. For example, a bank and a car repair firm may both need office supplies, but use them at very different rates.

Time to call

This is when your data comes to the rescue. Your transaction database contains a lot of information on who bought what and when they bought it. Therefore, it is possible, using statistical methods (I won’t bore you with the details now), to estimate the life cycles of different products across multiple types of industry. For our toy example, I’ve made the graph below that shows you how to make the results of such an analysis actionable. After doing the analyses, I was able to estimate the optimal time to re-supply each industry with the three specific products. As you can see, for the commercial industry your sales people can call for all three products at approximately the same time. On the other hand, the financial industry needs office machines every month while only needing office supplies every second month. However, when you start calling governmental organisations about office supplies every two months, you are not doing them a favour.

ActionPlotBlog

Time to analyse

This little example, inspired by analysis I did for one of our customers, shows how you can turn historical sales data into actionable insights for your sales and marketing team. The example features three products and three industries, but scales up pretty nicely to more products and customer segments. This is how to make this work from an analytical perspective:

  1. Determine the level of analysis: Is this going to be at the product level or, as above, at the product group level? A first estimate of frequency of purchase may help you determine what level is most relevant, when considered from your customer contact strategy.
  2. Collect your order data and aggregate the data at the right product (group) level. Organise your data in such a way that you can calculate the time differences between orders, split by customer and by product. Rough guideline: two years of data should suffice, but of course this depends slightly on your business.
  3. Analyse those time differences. An ad hoc analysis of the averages is the first step, but I recommend using advanced analytical techniques based on survival analysis. Such techniques allow you to also make seasonal adjustments and are much better suited to identify relevant differences and similarities amongst your customers. An important step in this exercise.
  4. From the analysis, you can subsequently derive benchmarks, indicating when you expect a customer from a specific industry to reorder one of your products. This provides you with the insights to help you time your message appropriately.

The use of advanced analytical tools becomes especially relevant when your number of customers and products is large. Having a good oversight on a computer screen is then practically impossible. Moreover, the timing of your customer contact forms the frequency dimension of the traditional RFM model. It is therefore a good basis for sharpening your customer segmentations.

Putting this into practice

I have now given an outline of the analytical side, but since this constitutes only the F from the RFM model, it is just one dimension for making this work. Indeed, it comes with requirements for data management and the IT environment, and involvement from sales and marketing is needed in order to ponder the right message and how this message fits within your overall sales and marketing activities.

But I’ll leave that for another post. I hope this information has inspired you to find ways to learn from your customer data and to serve your customers in a timely fashion, with the right offer. Hopefully it helps you win gold for your customer service!

Do you have any questions about this subject? Please contact us.