Are Multiple Channels 'Multiplicative'?
For the better part of a decade, marketers have been taught that it's a good thing to be multichannel.
In other words, it's a positive for catalog or e-commerce brands to branch out into new sales channels.
This may be true. It would've been foolish for a catalog brand not to start a website back in 1997. But would it have been smart or foolish for a catalog brand to open a chain of retail stores in 1997?
To determine if multiple channels are a good thing, you first have to determine if multiple channels are multiplicative, additive or generate diminishing returns.
An easy way to do this is to structure a query that measures customer behavior over time. For a catalog brand, select all customers who purchased two times from your brand in 2008. Segment those customers based on telephone and online orders (both orders via the phone, both orders online or one order from each channel).
Then calculate average demand in 2009 based on your 2008 segmentation strategy. If the segment of customers buying once from the phone and once from online outperforms the other two segments, then multiple channels have the potential to be multiplicative.
Another interesting query involves looking at purchase frequency in one channel (say telephone), then overlaying whether purchase activity exists in another channel (online). Many times you’ll find that the overlay of the online attribute results in a decrease in future telephone demand, suggesting diminishing returns. If spending is flat, then the channels are additive. If spending increases along the online dimension, then the channels are multiplicative.
It's wise to dive in to multiple channels when sales are multiplicative or additive; it's very costly to dive in to multiple channels when diminishing returns are suggested by your analytical results.
Give this analysis a try, and learn whether multiple channels are multiplicative, additive or exhibit diminishing returns.