April 23, 2013

Omnichannel: Are #Omnichannel Customers Truly More Valuable?

The most popular meme in retail is "omnichannel".  The theory posits that by aligning all channels around the customer, by creating a consistent experience across all channels, by tearing down silos inside organizations, by integrating a digital experience in-stores and online, by fully integrating mobile into the customer process, retailers get to reap the rewards of a customer demanding a fully holistic purchase experience.

You'll have to look hard to find any researcher or vendor who has proof that this strategy leads to a dramatic increase in profitability.

Here's how the argument begins.  We're told that omnichannel buyers, those who purchase from many, many channels, are much more valuable than single channel buyers.  Researchers cite this table as proof - here's what the data typically looks like:

This is where the research ends.  A simplistic query is run, and of course, there's clear proof that customers who shop many channels spend much more than customers who only shop a small number of channels.  In this query, 0-3 month buyers were chosen.  Then we measure historical channels purchased from, and we measure how much customers spent, historically.

Clearly, omnichannel matters.

Or does it?

Researchers need to go one step further, don't they?  Given that a customer is an omnichannel customer and has the same frequency as a single-channel customer, can we prove that omnichannel leads to increased future spend?

This table measures repurchase rates in the next month, after controlling for historical frequency, and for historical number of channels purchased from.  We're still evaluating 0-3 month buyers.


What do you observe, in this table?

Look at a customer with 6 life-to-date orders.  Read across the table ... these are customers who purchased from 1, 2, 3, 4, and 5 historical channels.  The customer buying from 5 historical channels should be much more valuable than the customer buying from 1 historical channel, correct?

But that's not what we observe.

In fact, you have to get to 10 historical orders, and 5 historical channels, before you start to see an increase in future repurchase rate.

Let's look at future spend, not future repurchase rate.


Average spend (no repurchase = $0, averaged with those who do repurchase) illustrates that there is some incremental value to getting a customer to purchase from multiple channels.  Read across the Life-To-Date Orders = 6 row.  You see that four channels are better than 1-3 channels ... mind you, only 20% better, but it's still better.

But there are quirks, aren't there?  Read across the Life-To-Date Orders = 3 row.  Here, buying from one channel is better than buying from three channels.  Read across the Life-To-Date Orders = 2 row.  Here, buying from one channel is better than buying from two channels.

For this retail brand, when a customer is early in the life cycle, omnichannel behavior is counter-productive.

For this retail brand, when a customer is deep in the life cycle, omnichannel behavior is more productive, but only 15% - 25% more productive, not 7 times more productive as is commonly published.

Run these queries for your business.  Seriously, go run them.  You'll see similar results.  Of course, your mileage will vary.

If your results look something like what I've illustrated above (hint - they will look similar to this), then there are a whole bunch of interesting strategic questions that need to be asked.

Strategic Question:  If most of my customers have 1-3 historical purchases, and the gains promised from omnichannel come from customers with 10+ historical purchases and 3+ historical channels, and the gains are in the 15% - 30% range (not 700% or 800% as promised in omnichannel literature), then of what benefit is an omnichannel strategy?

Strategic Question:  If the gains are in the 15% to 30% range, wouldn't I be better off investing in 1 new customer, so that I have 2 customers, than to re-arrange every process in my company so that I have 1 customer spending 15% to 30% more - only if the customer makes it deep into the life cycle?

Strategic Question:  If the payback of an omnichannel strategy is 15% to 30%, only among the 5% of customers who ever make it deep into the customer life cycle, then why should I tear down all silos in my company and work terribly hard to integrate all of my processes around digital channels?

Strategic Question:  What do Forrester Research, IBM, HP, and any other vendor promoting omnichannel integration have to gain by getting you to reinvent your entire business around a concept that only pays back 15% to 30% among the 5% of your customer audience that is deep into the customer life cycle?

Use the comments section to offer your thoughts.  Provide links to research that complements or refutes the information presented here.