If you’ve ever had a bad customer service experience at your local store, you’ll know how damaging a bad first impression can be.
It will be tainted, and the likelihood is that you’ll tell your friends and family about your negative experience. Their impression of
It’s much the same with chatbots, and how they are used on websites. Chatbots exist to facilitate your online
However, according to Paul Adams, Vice President of Product at Intercom this couldn’t be further from the truth. Here’s a recent statement from Adams: “Today’s bots have no manners. They don’t take no for an answer, they keep spamming and are generally frustrating”.
Strong words. That said, chatbots can serve your purposes quite well. In fact, it’s predicted that by 2035 such AI will boost labour productivity by 40%.
With this in mind, we at Luminate offer a few pointers on brushing up on your bots’ Ps and Qs so they’re optimised for lead generation success. Polite lead generation success, that is.
We think that some parts of Adams’ statement are true, but for the most
First, to get a clearer picture, we should briefly go over just how chatbots work.
How Chatbots Work
There are two kinds of chatbots: rule-based chatbots and artificial intelligence (AI) chatbots. The difference between the two is very much down to their sophistication.
You’ll likely have seen rule-based chatbots on your mobile phone - for example, say you get text alerts from your local cinema advertising their latest deals. When you receive a message that says ‘Text ‘241 MOVIE’ to our number to take advantage’ and you send the text, the application will recognise the ‘241 MOVIE’ as a command and will process your offer. The developer behind all this will have created a ‘rule’ for the bot to recognise your message, most likely sending you a confirmation email in the process.
By contrast, if the same premise were to be applied to AI chatbots then you could have replied with “I’ll take the offer, thanks!” or, “yes please, I want to see a film”, and the chatbot would understand.
It’s all part of their natural language processing. The processing behind AI chatbots
Tone of voice
The tone of voice has to sound human, and it’s this human aspect that you’re aiming for throughout. The last thing you want is for your chatbot to sound like R2D2!
By giving your bot a human voice, you’re mirroring the natural conversations we have on a day-to-day
Ensure your chatbot is as responsive as possible to emotional cues. It should be intelligent, able to spot and respond to a user’s emotions. This can be incredibly
Explore the English language
In moderation, of course. We’ve seen more than a few bots go rogue with potty-mouthed rants and, due to some fault or other be put out of operation.
On the other hand, it’s absolutely important for your bot to communicate with a varied lexicon so that it doesn’t repeat itself. Again, the human element prevails.
Your bot should speak from a ‘bank’ of assorted
Ensure the bot doesn’t break out of character
This is about consistency. It’s important your bot is consistent in its tone of voice and mannerisms, and that it stays a positive representative of your business.
The ‘Tay’ bot rolled out by Microsoft should serve as a reminder of when chatbots can go catastrophically awry - of course, this wasn’t so much a break
Offer actual guidance!
While this should be a given, far too many chatbots offer advice that goes only half the way. Most
Give your visitors enough options, and a clear picture of your services. Don’t just ask ‘how can we help you today? - give your visitors a selection! Direct them accordingly! Web design success depends on how much you assist the visitor.
Chatbots are one of our industry’s crowning achievements, yet so many aren’t using them to their full potential. You can take the above tips and ensure your chatbot is primed, polite and ready to engage your visitors.
Time to get chatting!