Speaking of the role Marketing should play in Operations, here’s the first article I have ever seen in a Marketing context about optimizing a VRU / IVR. This challenge is really very similar to optimizing a web site. You have path, and traffic down branches of path. You have bounce rates and exits. You have the same “choices with correct context” issue that is the heart of designing good navigation and inline link text. You have usability.
Fact is, a VRU / IVR is a technical interface to humans, just like a web site is. And just like many web sites, it was probably built and programmed by some engineers without a lot of direction from Marketing or Customer Service. There’s really no reason at all why the folks optimizing the web site should not also optimize the phone system too – especially if they do a good job with the web site!
I optimized my first VRU in 1991 at HSN. This was very new technology back then, and our customers were kind of shocked by it. Nobody used it. When I finally found an engineer who could print out a “path map” and I went through all the branches, I understood why nobody used it – confusing choices, unclear language. Sound familiar? Only 2% of customer orders were being processed through the VRU. And besides, customers like talking to live reps.
A real Marketing through Operations problem.
First we worked with the engineers to redo the branching and change the language so the VRU was smooth and easy to use. We pushed high frequency paths to the top of the stucture and sunk the low frequency stuff, eliminating steps for most transactions. Sound familar? I also felt the close on the transaction was ambiguous, so we built in a clear “confirmation” the order had been placed correctly. Sound familiar? How long did it take online carts to include a confirmation e-mail?
But then the Marketing problem. Lots of people had used the VRU and thought it sucked. How do we get them to give the new unit a chance? How do we get them to actually like using it?
Next, we gave the VRU more personality. We named it Tootie and had it toot a horn at the end of the order placing process – just like the hosts did (back then) on the live TV show. An “audio confirmation” if you will. So now the VRU does something the live reps can’t do – give customers a “Toot”. That helps address the “liking to use” issue.
But we still have the problem of trial – how do we get people to try the revamped interface out?
Fashion programming is what the core customer eventually migrated to; we knew this from previous hard analysis (not surveys or gut feel). If we could get these high order frequency customers to use the VRU, we’d get a significant jump in usage. Fashion shows were very high velocity and because sizes and colors are frequently involved, certain SKU’s can sell out quickly. So we had the hosts in those shows talk on air about how if customers used the VRU they would beat out everybody else ordering through a live rep.
In other words, “If you really want this product, you better use Tootie” – the “persona” we created for the VRU. Hard customer benefit.
Within a very short time, we had 20% of customer orders coming through the VRU, on it’s way to 80%. Saved the company an absolute boatload of money – and made customers happy in the process!
How did we know they were happy? Hard analysis (% best customers using VRU, % of their orders placed by VRU) and surveys of course, but we had a better indicator than these metrics – the number of Christmas Cards Tootie received each year.
That’s right, customers sent holiday greetings to the VRU.
How many greeting cards did your VRU / IVR receive last year?
Seriously though – what other customer-facing, technology-driven business processes can you optimize? You already know how to do this from the web site experience. Let’s create a list.
And here’s another link to that article – Listen Up!
David, thanks for expanding this idea on the technical side, worth the read for the historical perspective alone!
Both the Marketing and Technical skill sets together really are required to optimize these systems. Here’s an example from Tootie:
The original VRU opened up with “Enter the item number”. That’s a pretty logical thought on the engineering side; after all, it’s an order-taking interface. But as a Marketing person, I understood the behavior of the customer. First, I thought that most all orders would be placed for the item on screen. Second, I thought it was possible many folks would not have a phone in the same room as the TV (this was 1991 and cordless phones were not a big deal yet). So they would have trouble accessing the item number and this could supress buying impulse.
So I asked the engineers, “Do we know what percentage of orders are for items not currently on the TV screen?” Sure. Answer: 1%
Knowing many of our systems were keyed to current item being sold, I asked, “Can we have the VRU look for current item and just speak it to the customer?” And of course, the answer was yes, that would be easy. I also wanted the VRU to be a bit friendlier, which was also easy enough.
So the new open became: “Hi, this is Tootie! I’m currently taking orders for the (product name). Press One to order this item or press two to order a different item”.
That simple change reduced the “bounce rate” at the initial open by 90%. You see the same kinds of dramatic changes in web site optimization, where a very simple change in the logic flow creates huge improvements in conversion rate.
Engineers and Marketers should get together more often…
For those interested in reading further about Marketing and IVR, David offers these 3 blog posts:
Viral IVR Campaign
Email + IVR Integrated Campaign
IVR / CRM integration
This is a neat story! It seems likely that your demographic was uniquely open to the humanizing of the IVR. Most people are much more concerned with the amount of time it takes to get what the need and the initial assurance that they will get what they need quickly from the IVR.
http://www.charleskingconsulting.com (IVR for Microsoft Visual Studio)