HomeCognitive Enterprise

Customer Elimination Management for the ages

Like Tweet Pin it Share Share Email

Customer Elimination Management … CEM … is CRM’s evil twin.

We all have memories of companies doing their utmost to drive us away. If you’re like me, my family offers its sympathies to your family.

No, wait, that wasn’t it. If you’re like me you might have wondered just when the first instance of CEM took place.

Wonder no more. While it might not have been first, science has pushed the date of the earliest known gripe back to 1782 BCE. That’s the approximate date of a clay tablet found in the ruins of the Sumerian city-state of UR …

In the clay tablet, a man named Nanni whined to merchant Ea-nasir about how he was delivered the wrong grade of copper ore. “How have you treated me for that copper?” he wrote. “You have withheld my money bag from me in enemy territory; it is now up to you to restore [my money] to me in full.” (“World’s Oldest Customer Complaint Goes Viral,” Christina Zhao, Newsweek, 8/24/2018.)

Even with the best efforts of digital technology, I doubt your calls to customer service, recorded as they are for training and improvement purposes, will be discovered for translation by even the most diligent of 5918’s archeologists.

In the meantime we’re left to wonder if Nanni received a response that began, “Your clay tablet is important to us …”

We’re also left to wonder, with a bit more relevance to the world of modern commerce, if Digital technologies and practices (no no no no no, not “best practices!”) can, as promised, transform customer service.

But we aren’t left to wonder very long, because the answer is obvious. For companies already dedicated to providing outstanding customer service, Digital technologies won’t transform it, but they will undoubtedly improve it.

For companies that didn’t give an infinitestimal damn before Digital strategies and technologies became the Next Big Thing, Digitization will make their already awful customer service even worse.

In theory, business intelligence technologies, applied to masses of data gleaned from social media, might make a persuasive executive suite case that current service is putrid and customers are defecting in droves because of it while blackening the offending company’s reputation among those who, without the benefit of Yelp, might have given it a shot.

In theory, these same technologies, combined with the near-future capability to interpret telephone conversations for both substance and emotional content, might give that same company’s decision-makers, who couldn’t enter the Clue Store with a plutonium American Express card and leave with any merchandise, the clues they need to figure out why their cost of sales is so much higher than that of their competitors while their customer retention and walletshare continue to plummet.

But in the wise words of 1882 Yale University student Benjamin Brewster, in theory there’s no difference between theory and practice, while in practice there is.

The service a company provides its customers is an inextricable component of the overall value they receive when they buy its products and services. Digitize a business whose leaders don’t personally and intrinsically care about it … who care only about the impact bad customer service has on their annual bonuses and options awards … and the result will be the same bad service, available through more channels.

We’re entering a post-Turing world of chat ‘bots, email autoresponders, and, very soon, AIs with synthetic voices, all poised to correctly interpret what we’re saying or writing so as to accurately diagnose their product’s defects and scour our databases of successful resolutions so as to find the one that precisely fits our situation.

More often than not, though, what these capabilities will give customers are the same useless non-solutions to the problems they contacted the service channel to complain about, delivered a wider variety of more convenient channels but not providing more useful information.

Only now, the IT organization’s name will be on whatever complaints do filter through to top management. Which in turn suggests it isn’t too early to think about the brave new world of software quality assurance. Because in addition to the litany of tests IT already applies to its software … unit, integration, regression, stress, and end-user acceptance being the most prominent … we’ll need to add another.

Call it AIIQ testing. Its purpose will be to determine if the artificial intelligences we’re deploying to support buyers of the company’s products and services are just too stupid to expose to the outside world.

Maybe we can figure out how to use artificial intelligence technology to automate the testing.

Comments (9)

  • >grade of cooper ore


  • I’ve yet to meet a fake customer service agent or conversational interactive voice response (IVR ) system that didn’t make me want to scream AGENT! AGENT! AGENT!. Or actually scream it for that matter. I know this isn’t true for everyone, but by the time I resort to calling customer support, I’ve exhausted self-service channels, or know I have a complex enough problem that it isn’t in the playbook.

    I’m not sure if or when the day will come when a conversational bot can handle:
    “My kid’s going out of state to college, and I need to understand how to get care for here there, given that we’re in your HMO. In addition to whatever enrollment is needed, I want to understand the benefit levels, and most importantly I need to find a specialist local to that area to take care of her while she’s there. The specialist must be covered by your plan, taking new patients, provide services within a short distance from the university, and sub-specialize in her very specific health concerns, which I’ll be happy to explain to you. This is extremely time sensitive because she is going next week.”

    Maybe a fake agent can reset a password or tell me my account balance or any number of simple things. But my experience so far tells me that being really useful on even moderately complex matters is a long, long way off.

    I’m sure digitization, AI, and machine learning are all delivering value today for a narrow set of problems. But as with any new technology, there sure is a lot of snake oil out there.

    • Jim – You nailed it! I love one of my banks, whose call in voice tree starts with ‘press zero at any time to connect to a service rep.’ The ones with nested trees whose options bear no resemblance to my question are at the other end of the spectrum.

  • As far as Jim green’s issue with agents is concerned, would you rather have a live human reading from a script?

    As far as companies are concerned, Microsoft is a good example. The Edge browser won’t print without borders. I went looking for an expected date that will be implemented and all I found was 4 years of requests for it. Requests from lots of users.

    Not sure what century they are living in but the paperless office ain’t here. Looks like a lot of the same old same old to me about Internet Explorer and why users didn’t need things M$ didn’t want to build and didn’t need things M$ wanted to shove in there regardless.

    Microsoft is listening to users on all sorts of different channels but with exactly the same tone deafness.

    • Well sure, I’d rather have a human reading from a script. I can stop them, interrupt them, and possibly even reason with them.

      For example, I can say (and have said) to my cable company: “I have already unplugged the cable modem from power, unplugged the coax, unplugged the splitter, plugged it all in again, and powered up. The results on my modem have been…”

      I think we can agree that it’s the rare company that really does customer service with kindness and caring. My main point is the automation today hurts more than helps.

      And a wise man once said, “The service a company provides its customers is an inextricable component of the overall value they receive when they buy its products and services. Digitize a business whose leaders don’t personally and intrinsically care about it …”

      Great thought-provoking and conversation-provoking article, Bob.

  • Bob – many thanks for the attribution for the theory/practice quote!

  • The remarkable weakness of some current AI models was revealed in all its glory when Apple, in typical Apple hubris-induced myopia, shipped that $400 ‘competitor’ to Amazon’s $99 Echo.

    Turns out, Apple’s Siri is remarkably AU (Artificially Unintelligent). She’s kind of good enough for simple requests on your iPhone, but put her into a competing position with Alexa, and it turns out Siri is easily stumped. And pretty worthless as a result.

    You would imagine that Apple would have the non-artificial intelligence to spend $99 and see what the market leader is capable of doing. And yet, if that happened, there is no evidence of it in their poorly reviewed (mostly by users) shipping product.

    So even multi-billion companies in the tech sector with piles of cash to burn on R&D can’t come even close to getting it right currently.

  • “I will not accept here any copper from you that is not of fine quality. I shall (from now on) select and take the ingots individually in my own yard and I shall exercise against you my right of rejection because you have treated me with contempt.”

    This demonstrates that in 3800 years, things have not changed. Today we are tasked with 100 percent incoming part inspection. It is a pain in the posterior, but in some cases essential to quality control in today’s automotive parts manufacturing.

    As far as customer elimination management is concerned, AT&T drove me away years ago with their customer non-service practices.

    About the only time the automated system works without human intervention is for power outage reporting or bill paying. And the automated bill pay systems sometimes still dump me to a human agent, with an added five dollar fee.

Comments are closed.