Tag Archives: Multi-Channel Marketing

eMetrics 08 (SF)

As opposed to eMetrics 08 Toronto, don’t you know…

A really big shew, for sure. 

With the tons of WAA EdCom stuff going on, and the tremendous opportunities to just run into people in the halls and have hour long spontaneous “shootouts” (thanks for your help with “The Cluelessness of Crowds), it can be difficult to get to all the sessions I want to see.  

Still, I always try to catch sessions outside of the mainstream that look interesting.  Often nobody comments on these overlooked sessions, so I like to bring them to the surface.

The presentation by Egan van Doorn of the Dutch Automobile Association (ANWB) called Connecting Web Analytics with Decades of Marketing Metrics was such a session. 

Here, the beauty was in the simplicity and purity of the approach.  Classic Database Marketing – the targeting, the pacing.  No breathless monthly or weekly blasting of the same message to every customer.  No, to each customer the right message at the right time.

ANWB works with the understanding the calendar doesn’t matter nearly as much as the customer’s individual behavior.  When the customer is ready, they say so.  It’s all about Pull – gently bringing them to you, not beating them over the head.  Context, relevance; what they want, when they want it, while they are interested in it.  Like Search, right?

Web analytics folks often view multi-channel ideas as too complicated, and they’re really not – if you are using the right methodology and if you have some discipline.  Apparently, ANWB has both.

From a Marketing perspective, ANWB pays close attention primarily to high value online events.  Forget page views, visits, etc.  What they want to know is this: what action was taken for which we have a related product?  They store these events in the customer record, and then play out the online / offline Marketing stream accordingly.  If they can reach them online, that’s obviously cheaper.  If they decide to go offline (in the mail) they have their timing issues down and they make it happen. 

Very efficient, highly productive.  Huge increases in response rates, even offline when using online behavior to trigger the Marketing event.  Classic Database Marketing.  And there’s a reason they are so good at this – they’ve been doing the same thing offline forever.

If you can make money doing this offline, you can make an absolute pile of money doing it online because the Marketing is so much cheaper.  The problem is, most online folks don’t have access to that Database Marketing background, the understanding of how to optimize remote relationships.  So instead of playing it as Database Marketing, they play it like Media (Push) Marketing.  And they get unremarkable results.

How simple is it to do multi-channel right? 

Here’s an example, courtesy of ANWB.  Customer comes to the web site.  Customer searches for and finds info for “bike and hike” trails.  When this happens, customer is shown banners offering a “Bike and Hike Trails of the Netherlands” book during the rest of the visit.  Customer maybe buys the book.

Or not.  If they don’t, and waiting a reasonable amount of time for the sale to occur online, ANWB goes in the mail with an offer on the book, and then later on, a modified offer in the mail if there is no response.  Customers buy scads of these books.  Enormous ROI, both online and offline. 

Then repeat this scenario with every product line – what is the trigger event, what is the timing?  Man, that’s a beautiful business they’ve got going there.  Just printing money.

They do have one advantage – as a membership org, each customer has a unique ID, offline and online.  This was raised as an “unfair advantage” in terms of their success.  Disagree. 

Megan Burns of Forrester said as much in the 2nd half of the presentation.  The reason people don’t usually factorize to do this kind of stuff is they can’t project the ROI, they don’t know what they would do with a unified view of the customer to generate incremental profit.  So they can’t justify spending the money to make it happen.

This is really the same Push versus Pull issue I mentioned before – as long as you batch and blast, as long as you keep using the offline Push model, there’s no point in understanding any of this multi-channel stuff.  When you get ready to accept that the behavior of the customer is your key to relevance, and test through a couple of scenarios (as all offline DB Marketers have done), the ROI of the offline / online join becomes self-evident and justifies the spend to set up for it.

Wait a minute, you say – there’s no reason anyone would want to log into our web site.  Oh.  But now you are into Marketing Strategy.  Not the same issue.

Why won’t they log in?  Let’s say you don’t think you can get people to “log in” so you can create a database match.  Here’s the real question – have you conceived an experience for your web site that is worth a log in?  If the experience is worth it, people will log in, and you will have a database match.

That’s Marketing, my friends.  It’s not just about the “Push” MarCom stuff.  It’s the Strategy, the whole picture that creates the Pull that is so incredibly powerful. 

This is what makes interactive different.

Let’s assume you think I am wrong, that what I’m saying couldn’t possibly be true.  After all, how could so many people get interactive wrong?  Wisdom of Crowds, right?  After all, look at all the folks who got it so right in 2000 (not).  THAT was a Crowd.  This Jim Novo, he’s just pushing the ideas in his book.

OK, fair enough.  Here’s another voice that has been added to mine.  More on Akin’s book in the future – I’m only 1/2 through!

Share:  twitterredditlinkedintumblrmail


Follow:  twitterlinkedinrss