Humans built Artificial Intelligence to reduce mental work. Khoros built A.I. to reduce customer and agent mental stress.
Walking through the halls of history, humans' ideas of what the future might look like is often laughable at best. Most examples show the brightest mind of our past missed the mark by a mile. Once in a blue moon, history's forward-thinking minds make shockingly accurate predictions about how the future will look.
Take a look at the painting series ‘En L’an 2000’ (In the year 2000).
Courtesy of Wikimedia Commons
In the late 1800s, a group of French painters was asked to envision what the year 2000 might look like. While we might not have flying firefighters or underwater lawn games, they absolutely nailed one key aspect of our modern world, automation.
To be fair, there was no possible way that those old french painters could have imagined how small technology would be or the invention of the computer. Credit where credit is due, their mechanical cleaning robot is oddly similar to what currently sits in roughly 17 million households. A simple artificially intelligent cleaning hockey puck called the Roomba.
What exactly is A.I.?
Shouted in the news as a frightening new buzzword and strewn throughout pop culture, from Skynet in the Terminator series or HAL 3000 in 2001: A Space Odyssey, artificial intelligence has developed an unsettling otherness.
This narrative does make for a great science fiction horror story, but it’s a bit off course from the truth. In the simplest terms, A.I. is nothing more than a computer program that completes tasks commonly suited for humans.
Take one of the newest models of the Roomba, for example. To those who own one, it can seem like magic that the way it can map out a room and keep your floors clean enough to eat off them. Behind the scenes is an impressive piece of A.I. born out of a simple problem. In previous models of the Roomba, the tiny cleaning robot would spread pet accidents around the house because it couldn’t detect said pet accident.
To any human tasked with cleaning a floor, avoiding a messy pet would be child's play. For the designers at Roomba’s parent company, iRobot, this problem was a computer engineering nightmare. How do you teach a robot to see, understand, and avoid pet accidents? A.I. was the most straightforward answer to solve this question.
Photo Owner | QualComm
How exactly the iRobot team programmed their A.I. is a closely guarded secret. They were willing to tell CNN that it has to do a lot with image recognition algorithms. The scope of how these algorithms work is sadly beyond the breadth of a single article; however, the layman’s explanation would follow along the lines of feeding an algorithm millions of photos, and testing which version of the algorithm can identify your photos correctly and then keeping that version.
But A.I. is brand new. Right?
The version of A.I. that we use today is brand new in a manner of speaking. Some versions of A.I. exist in everything from vacuum cleaners to phones, computers, cars, and even your thermostat. Surprisingly, the idea for A.I. goes back to the early 1900s when computers first came to fruition.
In his 1950 paper, Computing Machinery and Intelligence, mathematician and early computer scientist Alan Turing developed an 'imitation game' for computers. A test to see if a computer could trick a human into believing the computer was a fellow human. In Turing's mind, if humans could store information and then apply that information to solve new problems, why couldn’t machines. The main reason that A.I. is seemingly popping out of nowhere today was the sheer cost of computing.
According to the Economist, In 1956, the cost of a single megabyte of data (roughly one minute of MP3 song data) was $99,250 when adjusted for inflation. In 2019 that same one minute of song data cost approximately $0.00002.
There’s a lot more to A.I. than fancy robots.
As mentioned above, the technologically infused world of today is full of A.I.: Algrothims, deep learning programs, and A.I. running every social channel, streaming website, or service. Airline A.I. programs can detect at which max price point a customer will buy a ticket, and even IBM's Watson doctor program is helping to detect lung cancer in humans.
While this is all well and good, one A.I. system has gotten quite the bad rap. The customer support chatbot or automated voice. It’s stuck into an endless cycle of miscommunication jokes in pop culture. This does make for great content but it is far from reality. Today's chatbots are so powerful that Vice News reported in early 2022 that people fell in love with Repilka’s A.I.-powered digital friend chatbot.
These chatbots are so prevalent that everyone from banking care systems to yoga mat companies is using chatbots and A.I. in several ways to reduce the time from customer question to answer. At least that’s the theory. Some brands use these chatbots exceptionally well. For example, after six questions in Instagram DMs, you’ll be sent to either the brand's community page, receive your answer via chatbot, or connect directly to a live agent. Others use chatbots to make you answer the same six questions and leave you no closer to a solution.
Go with Flow.ai to improve your care system.
For many, the use of more advanced technology, especially those built with A.I., is a change not worth making. It costs money to retrain staff, and have several cross-organizational meetings internally, which takes time and money, and the saying “if it's not broken, don’t fix it” is a saying for a reason. The question is, is it still profitable to lose customers repeatedly for poor customer care? Is it the best practice to lag behind and spend money on powerful tools only to not use them to their full potential?
One customer here at Khoros cited a 200%+ increase in ROI after implementing custom Khoros’ bots into their Care suite. The most likely reason for this is simple; happy customers. Happy customers will spend more with brands and recommend those brands to people in their social circles.
Take a fictional but all too common situation of a customer needing a simple exchange form for a pair of sneakers they bought on a brand’s Instagram page. It’s most likely that this customer is in a younger, more tech-savvy demographic. All they need is a page link to send their sneakers back for an exchange. They are not going to care if they get that link from a chatbot or a self-service page.
An important question to ask is, what is the likelihood of that customer continuing to use and recommend that sneaker brand if they can get an answer to their question in under 1 minute? How are the brand’s agents going to perform when they can focus on more complex customer calls and not a customer just looking for a single form?
Powerful doesn’t mean one size fits all.
Now, the idea of using the full suite of Khoros Care with Flow.ai for every brand is wildly shortsighted. Every brand is different in size, needs, and customer demographics. Some brands have younger, tech-savvy customers who would much rather use self-service, and some brands will have customers still paying their bills with paper checks and would love to just be connected to a live agent.
No matter which customer demographic is using a brand’s Care system, it is wise to use Chatbots and Flow.ai to get your customers routed as quickly as possible. If a customer can get their answer in 3-4 questions with the bot, that customer will be pleased.
Being mindful of your chatbot routing will be vital in utilizing their abilities. While increasing your call deflection could be appropriate for simple questions, relying on chatbots to answer every customer question could leave your customers frustrated, without an answer, and possibly looking at other options.
Before implementing any care system, take a moment to discover exactly who your customers are, and how to best route your chatbots to provide the best possible service.
Now that you have a broad understanding of A.I. and Khoros Flow, expand your knowledge and stay up to date on the latest in innovation by heading over to the Khoros resources below.
- Preview Flow.ai and the onboarding experience by taking a look at our crash course video series
- Get acquainted with our Flow.ai getting started series
- Subscribe to the Care Blog for more best practices, success stories, and product roll-outs.
- Khoros' Introduction to Agent Assist webinar
