So many still don't get it...

Despite a 10 year struggle with technology being flung in users faces, some of us still think that if we can just get to the right technology, all problems get better. It just ain't so!

So, here's the latest entry Service with a Smile. Besides being just a tad self-contradictory, it just simply misses the fact that there is a right approach to create the right experience BEFORE you select a technology. To think that the solution to customer problems caused by technology is more technology and an interesting business model is just more insanity.

The difficulty is that while all three companies take needed approaches, none solve the problem directly. Upfront good design is what’s needed. Having that would have meant that ClickFox didn’t have to find out for the client (who spent a lot of money) that callers were confused between which ZIP code. It would show that not every caller or company needs rotating, customized menus from Angel; many just need the right menu first. It would show what to do with all the wonderful data that Austin Logic collects and correlates.

The right design approach addresses all these things at the right place: The start! In fact, Forrester, while handing out all those awful grades, singled out one design for a special case study and referenced that in the white paper as an example of what should be done. And it wasn't the promise of a new technology, rather the promise of the right approach. One that’s been proven multiple times in the past few years for large and small clients.

There really is no paradox involved in automation. Studies show that many people like automation. They’re just picky about liking automation that works! Privacy, convenience, and speed all drive people to automation, but bad design drives them right back out. It doesn’t have to be that way. The days of bad design should have been over years ago, but too many times IT departments, bottom-line-only focused customer service managers, and the “we’ve always done it this way” crowd refuse to consider that all the goals can be balanced to make a greater portion of all stakeholders happy.

See, we can please the folks that want good automation. We can please those who know that their reason for calling isn’t handled in automation. We can clarify for the confused. We can achieve savings goals. We can help cross-sell. But only if we focus on all those things in design. It all comes down to creating the right experience for the callers. That’s a focus that does NOT start with technology. It starts with goals. Individuals are treated as such with the right process of finding out who they are. The failure of turning to technology as a starting point is to sell great school busses to people who want to ride bicycles. If we find out first that people want low-impact, low-density, quiet transportation, then we might go find a really good bicycle maker. The right technology selected by the right design.

Now, I’m not slamming any of these companies. Actually, I think all three offer critical components of successful business solutions: usage tracking, scalability, and knowledge manipulation. But, they still cannot overcome bad design. Even in the ClickFox example, how many callers were turned off by doing it wrong the first time? Finding out later is too late. Designing right, from the beginning, is the missing leg of the stool, and really the most important.

Comments

[Peter Leppik]

Hear, Hear!

The question is: how do you sell good design to a customer who sees testing and designers as optional extras?

I linked to this article from our blog here.

[Rick Rappe]

Awareness of this dialog came to me thru participation in the VUIDs chat group, and I was not terribly surprised to find a comment from my boss and VocaLabs founder Peter Leppik. He very diplomatically makes some of the same points I would have, but I would like to add a couple of comments that come from the perspective of sales and marketing. In an almost whiny tone of voice I have repreatedly heard from speech technology sales people that their hands were tied. That the customer had a primary focus on cost control, and were they to try and modify the focus to good service as a priority over saving money, the potential customer would simply seek another vendor. Well, maybe and maybe not. And while I could go off on a tangent about establishing rapport with the customer in order to have them listen to what might be contrary suggestions; it is true that that level of trust is often difficult to establish. But is IS possible to disuade the customer from incorrect priorities. Its just harder to do than to simply write off the experience as "Thats what the customer wanted and if we didn't give it to them, they'd just go elsewhere." An example that comes to mind is with one of our more significant clients who uses our services to rate caller acceptance of various persona/voices. When the client directs that they 'like" the voice of Suzy more than Jane for their speech application, our client takes the scientific approach by recording various voices and asking real callers to listen and rate them as to favorites. Armed with statistical proof that Jane is the bettter choice, the customer readily bows to the expertise of the vendor and not only does the better voice get used, but the trusting relationship between client and vendor is solidified. That brings me to a more in-your-face comment as to why so many speech applications suck. There is no alternative but to lay the problem at the feet of the VUI designer. Said plainly, the demand for VUI expertise exceeds the supply. Many came to VUI design with a background in graphic/web page user interface design (GUI) and quite naturally brought the methods that work in GUI design to the VUI space. Trouble is GUI is not VUI and adherence to methods that work in designing a computer program do not adapt well to a voice-computer interface. I well remember an early encounter with a well known VUI designer who's PhD. came from the GUI field. She was adamant that with a focus group of 6 individuals, she could fully design a high quality speech application. Well, I'm sorry, but she is wrong. It may be true that she creates quality designs, but that is more because she is good at her job; NOT because a GUI inspired design methodology is the right solution. In GUI, watching how a focus group individual interfaces with a computer screen, includes the observation that with GUI the subject has plenty of time to think where to place the cursor and click through a process. But with VUI, there are no visual clues, just a voice command. It is much harder to step through a VUI design where hitting the "back" arrow to change course isn't an option. And with over 60 ways in English just to say "yes", a 1/2 dozen testers simply won't cover the possible ways things can go awry. Lastly, there is a tendency for the VUI designer to want to show off technical expertise with the result that confusing call path dialog sneaks into applications simply because the designer was being clever. The K.I.S.S. principle applies to VUI in spades. An example that comes to mind was a VUI team convinced that a persona with an "attitude" was clever and memorable. The persona was given a rural "NASCAR" type accent and attitude the VUI team thought appropriate for the application. At one prompt there was the phrase "...and just say No if you don't give a darn." Well feedback from test callers revealed a high number of callers that thought they heard "give a damn", not "give a darn" and were so VERY put off by such language, they panned the entire application. Clever? Sure. Correct and appropriate? Hardly. And yet another "sucky" application was born.

[one-eyed king]

Rick,

there are definitely a lot of bad designers out there, but the problem really comes down to ignorance on both sides. Sort of like the dance of death between people who have high cholesterol and McDonalds. "It's inexpensive!". "We just deliver what the market wants!".

Neither side pursues truth and knowledge. Neither side considers real consequences. Neither side really cares.

And in our case its worse because a third party suffers the consequence.