Bot or not? 5 ways to spot a chatbot
Posted: 6 November 2017 | By Caroline Klatt
In the business and technology world, chatbots are becoming an invaluable tool for customer service, sales conversion, and brand engagement when done right. Gone are the days where chatbots sputter nonsense to a customer’s queries– the new wave of chatbots don’t resemble the typical awkward, unresponsive bots of the past. According to a new report by Aspen Software Research, 44% of respondents said that if a company could get the experience right, they would prefer to use a chatbot or automated experience for CRM.
If you receive an instant response at 4 in the morning, you’re likely speaking to a bot
By creating constraints and goals for artificial intelligence-powered chatbots, human chatbot programmers allow the bot to perform more efficiently at the tasks at hand, instead of being a catch-all bot that fails to be helpful at anything. It’s difficult to realize whether you are speaking to a human or a bot as they develop their own colorful personalities and quirks. Here are a few methods that reveal whether you are interacting with a human or bot.
You see preset menus.
When you message a brand on social media, the easiest way to spot a bot is if, upon messaging, an automatic menu pops up with options like ‘Customer service,’ ‘See products,’ or ‘Book appointment.’ Or if, upon messaging, you receive answer choices that you can select, you’re speaking to a bot. These choices help the bot categorize your query, making it easier for the bot to redirect you to the right answer. If fulfilling a customer service role, bots are programmed with calculated, accurate responses to the most common customer questions.
You get instant responses, 24/7.
Unlike humans, bots don’t sleep. If you receive an instant response at 4 in the morning, you’re likely speaking to a bot. Often customer call centers are only open during business hours, so access to customer service at all times of the day is a value-add on the part of a bot. Even if a customer has a special issue and the bot can’t easily understand their messages, bots will stick by the customer’s side unlike a human who could stop replying or exit the conversation. Bots will tirelessly work to redirect the customer to the right information.
It keeps confirming topics.
Bots draw upon huge amounts of data to answer questions, and generally need to confirm the topic at hand to access the right information. If you find the other party doing significant confirming, especially if you switch from one topic to the next, i.e. “We’re still talking about transferring money, right?” or “Do you still want to browse women’s pants?” It’s a bot trying to keep up with you.
Like speaking to a human customer service representative, you need to cooperate with the bot to give them the right information for it to assist you correctly. Bots confirm topics not to be repetitive, but to ensure that they will search for the correct answer to a customer’s issue. In a human conversation, a person wouldn’t abruptly switch from topic to topic ideally. Just like this would confuse a human service representative, it confuses a bot. Therefore, the bot will often confirm topics to ensure accuracy in finding results.
You don’t see any grammatical errors.
Even the most careful humans can make mistakes if they’re typing quickly. Machines are much less prone to error than humans are; making them a top choice for companies that seek efficient, quick and relatively errorless communication with their customers. With pre-programmed messages and responses, the content of a bot is highly regulated and controlled. Even a human with strong spelling and grammar skills can have the occasional blip of the finger, but a bot will always have perfectly grammatical responses.
Caroline Klatt is the CEO and co-founder of Headliner Labs, NYC-based artificial intelligence and mobile messaging company specializing in ecommerce, retail and entertainment chatbots.