I make jokes regularly in the office about how I’m going to use Aspect’s interactive text response and natural language understanding platform to build an AlyxBot that will attend meetings for me and report back. (If any of my coworkers are reading this and feel like they’ve been in a meeting with such a bot… well, I plead the Fifth).
After reading that, I really wanted an AlyxBot. A simple chatbot with some pre-built preferences and a couple of API integrations that can tell my friends I hate asparagus and am listening to a lot of Run the Jewels? Oh, that’s fun. It’s like a preference center! For ME!
Most of the pop-culture data you’d get from my social and music service APIs is admittedly not going to be of much help when it comes to customer service. But by extension, a similar bot with a more general education could also take on the task of interacting with companies. When it comes down to it, companies might say they want to personalize service, yet often don’t ask a lot about me. Most aren’t working with much more than my age, zip code, net worth and income. At the most basic level that’s what they know how to market to, as they assuredly have some general personas they can combine with any account history they might already have to determine if I’m in the market for a car, or house, or pants, or their mixtape.
The same sort of preference map that might be used to control your own personal me-bot isn’t really a new concept in the customer service world. There are strong parallels to the idea in the mostly theoretical world of customer service applications of VRM (vendor relationship management) – the idea that customers should be able to set global preferences for how they wish to interact with businesses. My hatred of asparagus would be a little micro for VRM (or would it?), but my distaste for voice calls has broader applications.
Yes, the “omnichannel” revolution has made customer service much better, especially for those of us who aren’t into voice as a channel. Still, there are two glaring issues:
The first issue is that I need customer service in the first place. Much like “no UI” is the new UI via blissfully interface-free technologies like Amazon’s Echo, why can’t “no service” be the new customer service? If I’ve laid out enough of my preference map, I should be able to automate most issues out of existence. If I tell my bank that I’d like to keep my checking account balance between $500 and $1000, can’t they just handle that in the background without any confirmations from me?
Second is that despite myriad improvements to the journey, it’s still a journey. I appreciate being able to use the channel of my choice. I breathe a sigh of relief when I don’t have to rehash the context of my call to a live agent. But for as much as the variety of choices for handling automated customer service have improved the process beyond the linear IVR -> Agent format, I still occasionally liken the process to the “Choose Your Own Adventure” books I read as a kid, or the “Zork” video game. Every choice I make, whether wading through IVR menus or having text-based dialogs with ITR and mobile web apps, should get me closer to resolution… but there’s always that nagging fear I’ll end up somewhere pitch black where I’m likely to be eaten by a grue.
The Internet of Things, though nascent, is only going to exacerbate our demands that companies be able to predict our demands. If my lights are smart enough to learn my schedule and turn on and off as needed after I’ve given them some basic direction, I’m also going to expect them to be smart enough to understand when one has stopped working, is under warranty and a replacement needs to be expedited to my house. I don’t want to call in; I want to authorize my bot to collect a RMA from the vendor for replacement for all my in-warranty devices.
In that respect, the bot would be your concierge, your bouncer. The you that businesses go through before they get to you.
If I were to forecast one of the biggest potential hurdles for adoption, it would be to get people comfortable with this kind of delegation. When you manage people, one of the hardest things to learn is what you can and should delegate. Same goes when you manage systems. In either case, you have to be able to articulate what you want, and the person/system to whom you are delegating the task has to be competent to complete it, and possess enough critical thinking skills and autonomy to ensure the order is the right order for the situation.
How much are you comfortable delegating? Does the amount go up if you feel like you are truly delegating it to “you”?
I trust myself, so I’d trust AlyxBot VRM to screen my calls, texts and emails, make my doctor and dentist appointments, shuffle my funds, and probably even to make some purchases on my behalf – if I’m convinced of the power of “conversational commerce” and using bots as my shopping assistants, it’s the next logical step.
Implementing something like this would be unlikely to be a painless process, but nothing ever is for early adopters, and somehow we always get over it, chuck our deprecated devices in a closet somewhere and go for it again. Just like one IoT light bulb managed to DDoS an entire house, the world of “no service” would probably have its share of funds being transferred in infinite loops due to errant decimal floats, and other newsworthy-but-not-particularly-earth-shattering issues as it grew with fits and starts. But it could lead us to a far more streamlined existence.
- Chatbots, Continuity, and a Helping Hand - May 26, 2016
- No Service Is The New Service: Bots, VRM and Delegating To Myself - May 4, 2016
- Facebook, The Commerce Engine? - April 8, 2016