Apparently, I don’t know my own size when it comes to ordering T-shirts.
Truth be told, I have a bit of a T-shirt problem. I own and wear a lot of them. Roughly 100 graphic tees of different designs currently take up most of my closet space. A few months back, I did one of my occasional shirt culls to get rid of those that had become too small through shrinking in the wash (at least that’s my excuse), or too worn out and ratty to be seen in public (translation: My wife says, “You’re not really going out wearing that, are you?”).
The result is a row of now empty hangers, just waiting to welcome some new shirts.
This provided the perfect opportunity for me to pick up a few new shirts from the graphic designer in the UK whose range of shirts I’ve seen online and liked. So last month I ordered myself six designs, all XL in size. At least I thought I had.
I woke the following morning to find an email from Taylor at the distribution company with the subject line, “Did you mean to do that?” Taylor pointed out that while five of the shirts I’d ordered were “XL,” the sixth was a “S.” They asked if I had ordered that shirt for someone else or, as the design on it followed a similar theme as the others, was it actually for me and I just hadn’t selected the right size?
Talk about customer service.
Taylor-ing My Customer Experience
Taylor’s email got me thinking about the future of customer service as we evolve towards more automated systems, and increasing use of artificial intelligence and machine learning technologies.
Would an AI customer service system catch that mistake? Would it even know that it was a mistake? Or did it take a Taylor, a real live human-in-the-loop, to see the discrepancy in the data (size selection), interpret that as being due to two possible scenarios, and reach out and establish a helpful relationship with the customer?
Most automated fulfillment systems linked to ecommerce sites would have just completed the order and sent me the “S” T-shirt, leaving me to wonder why and:
- Probably complaining that they got my order wrong.
- Possibly eventually realizing my mistake and asking about a return and exchange — costing both myself and the company time and money.
- Gift the shirt.
Related Article: Customer Service Friction: A Double-Edged Sword
Would AI Catch My (Very Human) Error?
Could an AI/ML application act the same way that Taylor did? Quite possibly, by training it to look for discrepancies in data patterns from online orders and subsequently consult a series of models linked to those pattern discrepancies. Each of those models triggers a response reaction, such as an email asking the customer to double-check the order.
But someone would have to work up those scenarios, build the models and train the AI. If you have a sufficient volume of people like me who forget to move the size indicator from the default “S,” and are having to deal with the logistics and costs of returns and exchanges, then maybe. But an easier way to resolve that may be through a better UX design on your ecommerce website so it doesn’t generate a large volume of incorrect orders in the first place.
If the mistake is from a returning customer, we may have collected and analyzed enough data on their normal buying patterns to provide a personalized response along the lines of, “This is different than your usual order — please confirm.”
But what about a first-time customer, as I was in this situation? You don’t have purchasing history data on which to make those sorts of decisions. It took an individual to see the mistake I’d made and extrapolate what that could mean.
Related Article: We Know Chatbots Are Falling Short, But Why?
We’ll Always Need the Taylors of Customer Service
Technology and automation have a large part to play in the future of customer service, especially when it comes to handling large amounts of routine data. But when it comes to handling the exceptions, there will always be the need for the Taylors to be able to apply some human intuition and proactively reach out to help resolve an issue before customers even know they have a problem.
And thanks to Taylor and their intuition, guess which T-shirt company I’ll be going to the next time I have a few empty hangers to fill?
Alan J. Porter is the Chief Content Architect at Hyland software.