May 24, 2010

Mobile Marketing: Measuring Channel Shift

Recall our table from last week:




April







March Visit Use Website No

Website App + App Activity Totals
Visit Website 10,000 50 100 20,000 30,150
Use App 100 300 200 1,000 1,600
Website + App 50 250 300 1,000 1,600
No Activity 20,000 400 400 0 20,800
Totals 30,150 1,000 1,000 22,000 54,150

When measuring channel shift, always remember the "multi-channel mantra" that is posited by those who manage the incumbent channel: "Multi-channel customers are the best customers".

They aren't the best customers because they are multi-channel, of course. They are already the best customers, and as a consequence, they are often the first to try new channels.

What matter isn't whether they try multiple channels or not, what matters is what they do after trying a new channel. It's what they do after that is dangerous to the incumbent channel.

Look at the table from last week. Let's look at the customers who experience both the website and the app.
  • "Re-Visit Rate" = (50 + 250 + 300) / (50 + 250 + 300 + 1,000) = 37.5%.
Now, we'll calculate the "Re-Visit Index" for each channel:
  • Re-Visit Index, Website = 50 / (50 + 250 + 300) = 8.3%.
  • Re-Visit Index, App = 250 / (50 + 250 + 300) = 41.7%.
  • Re-Visit Index, Website + App = 300 / (50 + 250 + 300) = 50.0%
Any index with a value > 20% suggests that the customer wants to change or migrate to other channel combinations. In this case, the customer is not going to be a web-only customer ... the customer is either app-only, or app+website. This is a common trend early in the evolution of a new channel .... customers either switch back to the old channel in the early days, or the customer goes back and forth between channels.

Pay close attention to the fact that the app index is five times greater than the website index (41.7% to 8.3%). This is a strong indication that the customer prefers the app over the website.

1 comment:

  1. Oh the times they are achangin'. Who know two years ago we would be relying heavily on the usage of "apps"

    ReplyDelete