Complexity on the backend, variability on the front

Great article about Average Handle Time (a call center metric, often linked to customer service representative incentives) here, from Tripp Babbitt, a bright guy posting on the Customer Management IQ site (a must-visit). (In the ever-spaghettifying nature of the Internet, Tripp is commenting on a blog posting from Blake Landau. And now, my comment on that comment on a comment.)

I have two sayings. (Well, I have more. But for right now ...) They are really rules of the road for any and every company, consulting firm, pundit, non-profit. It doesn't matter who you are, these are the rules.

First, a company's job is to create greater perceived value than their competitors. 

Second, a company should do this by managing complexity on the backend while handling variability on the front.

The AHT metric is a perfect storm where using it as an incentive violates both rules.

rule one: greater perceived value than competitors

Perceived value is essentially a scorecard. In the old days, it was a utility function (Bradley Gale's original approach, since modified by him, his firm and many others to be broader). My approach is really to break apart the number of things you measure into three groups: product, service, and experience. These are fairly arbitrary at times (is a McDonald's hamburger a product, a service, or an experience? Yes.) But with some discipline and an up-front clarity about who "owns" which key performance indicators, you can create a scorecard. It may be articulated into your systems by hooking it up into Balanced Scorecard (BSC) - or not. The key thing is to measure how customers perceive your value along those three dimensions, and to give people responsibility for fixing what's broken. 

TIPS:

• Make sure the scorecard is relevant to the customer, and revealing about their true wants and needs. 
• Then, compare your performance with the performance of your competition.
• Subtract your score from your competitors' score. (I'm oversimplifying - but stay with me.)
• Pray that the difference puts you on top.

And if you need to improve your scores, look at the next section. 

But look here first at average hold time (AHT). If your CSRs are handling customer problems - moments of truth where total lifetime customer value can be created or destroyed - the last thing you want is to disrespect them by having CSRs hurry the call along, or even (gasp) hang up on the customer. Yes, that shameful activity happens all the time. 

Instead, you want to make sure you create greater perceived value in your call centers than your customers have come to expect. Ideally, you'd do this "on brand" as well, so that you don't overspend trying to make every little thing about that customer call perfect. Just make it effective and memorable - and better than they would expect from competitors. 

AHT at worst works against this goal by not giving the CSR the time needed to deal with the customer's moment of truth in an emotionally effective, memorable way.

Even at its best, AHT reduction misses the point. The problem is that it includes the word average. The real gold in data mining your call centers is to determine the causes of call time variability. Averages squash out any of that corporate value. 

Why do you care about the causes of variance in call times? See the next section.

rule two: manage complexity on the backend and variability on the front

So, let's focus on variability on the front end. What causes such variability? 

1. You cannot execute consistently on the same challenges.
2. Challenges you thought you needed to address are wrong, and so tools and training are non-existent.
3. There exists at least one challenge you never thought of.
4. The same customer has a variety of needs at different times.
5. You have a variety in your customers and stakeholders you have not yet modeled or dealt with.

You are looking at the following "stable objects" in your management world in these circumstances: process variability, challenge variability, and customer variability. 

What you want is to be able to handle these variances from the "norm" well. And Six Sigma/Lean is not necessarily the only tool - nor even the best tool, particularly when you really want to address customer challenges and variations. After all, a customer changes in their needs over their lifetime; if you want customers for life, you must provide a combination of variable service and branded experience. You want variation in your output!

So, what kind of variations should you address? How complex are your customers' challenges? Where are they within their lifecycle as they define the relationship with you? 

If you only had that information.

But you can get it, can't you? From your call center. Because where also are you going to look first for the hard realities facing your own customers at moments of truth ... where your longterm profitability is at stake? 

The call center should not necessarily be a customer therapy opportunity, although empathy does help. Instead, it should be an opportunity for managers (call center management and others) to determine the root causes of the variability in phone call durations, and without judgment ask what the consequence of these insights are to product, service and customer experience. 

And then, go back to rule one.

Update on Thursday, October 22, 2009 at 8:18AM by Paul K. Ward

Blake at Customer Management IQ just posted a comment on my comment on a comment about her comment. In the interests of post-modern self-referential content, I am obligated to comment on what she said... 

Seriously, the distance among Seth Godin (mentioned in Blake's entry here), Tripp and me is very small. Empathy matters (per Seth), not just in human terms but in terms of developing customer engagement (see Gallup and Carlson); Tripp might say, "Don't manage to the metric unless it actually helps you do your job", and average handling time doesn't help CSRs actually do their job; and I say that AHT artificially pushes your stats to the "norm" and thus destroys information and indicators that are critical to improving everything at your company. You should seek out variability where it happens - always - and instead of squashing it right away (as you might with Lean and Six Sigma), ask about what's causing it. 

Another one of my sayings: "Make regular what you can, but don't make regular what you must not." Making things "regular" (turning them into consistent systems) lowers costs and helps make a scalable business model. But at some point, the variety of interactions you have with customers is exactly right. You're handling their situations, at a reasonable cost to you, in ways that reflect their wants and needs, moments of truth, and the values of your brand.

Now the challenge is not to let this variability creep too high again as you desperately seek out more and more customers who want a wider variety of products and services from you which you cannot possibly execute well. Then you're pushing your overall customer satisfaction stats down (probably to the norm of your sector) and losing potential differentiation. And it's differentiation that both companies and customers ultimately want. It makes for profits, and a clear brand that people can love. 

Remember when Target was indistinguishable from Wal-Mart? Talk about a disciplined approach to variability ... just look long and hard at Target.