Understanding multi-channel dynamics – Part 2

This article, written by Neil Mason, was originally published on Clickz.com and is republished here with permission.

In my last article I looked at the measurement aspects of understanding multi-channel dynamics. This week I’m going to look at some of the analytical approaches. Having put in place the mechanisms to track cross-channel behaviour, it’s important to explore the observed dynamics of the interaction between the online and offline channels and to understand why some of these behaviours are happening and whether they are desirable or not.

A point of focus for an organisation might be to understand why a customer who starts a transaction online then ends up completing it offline. Many organisations would usually prefer those transaction to be completed online as its cheaper to process the transaction. The reasons why this channel shift occurs could be down to the way that an organisation does business, down to traits of consumer behaviour or because of the complexities of the product.

A while back we worked with a bank to map out the channel dynamics and to try and measure the channel shift from online to offline. This was complicated by the fact that the bank had installed internet terminals into its branches to allow prospective customers to fill in applications for some of the simpler products online but in the branch. The idea was that it would reduce the need for customers to wait until branch personnel were available and that one branch person could help many customers at the same time. Branch personnel would also be freed up to sell more complex and higher value products. However, what the bank found was that the branch personnel would often lure people away from the branch terminals to do the transaction on their own systems. The reason was simple and that was the branch personnel didn’t get commissioned on sales that were made on the terminals in their branches. In order to get the desired behaviour the bank needed to capture the IP addresses of the terminals in the branches, link them to the sales made on the terminals and then allocate those sales back to the branches. In that way the branch personnel were much happier about allowing people to “self serve” in the branch.

Last time I talked about a holiday company serving an older target market. Having set up the measurement tracking capability to look at cross channel behaviour, we set about analysing why channel shift as happening. We looked at the bookings that had been made on the website and compared them against bookings where someone started the process online and then had completed the process in the call centre. Across of the things that we looked at the gender of the person making the booking was the biggest factor. Men were more likely to do their research on the internet and book online. Even if they had ordered a brochure they were more likely to go back online to make the actual booking rather than call the call centre. Women on the other hand we4re far more likely to use the site for research only and to order the brochure but would then call the call centre to make the actual booking. Focus groups confirmed that this was the preferred apporach for women and so in this case channel shift was down to gender differences.

In some cases channel shift might be down to website issues. We conducted a similar piece of analysis for an insurance company looking at channel dynamics on their car insurance products. Once again we assembled the data to look at the bookings that were made online and compared them to those bookings that end up in the call centre. We looked at a number of different characteristics including the type of insurance cover, the car being insured as well as the demographics of the policy holder. In this instance given the breadth of the data we used Chaid analysis to identify those characteristics which were the most important in predicting channel shift. The results were somewhat surprising. Rather than demographics being the most influential factor as I had had suspected, it was actually whether someone had bought a particular optional extra on the policy. If they had, they were far more likely to have completed the transaction in the call centre. Armed with this information, the company went back and reviewed the site processes for buying this particular optional extra on the policy and could see where the process could be improved to help reduce the need for people to call the call centre.

Channel shift may be down to organisational issues or site issues. These issues can be addressed. Other factors may be more ingrained in the way that customers want to do business and so in these cases channel shifting should be embraced as long as it’s recognised accordingly.