Chatbots are becoming less of a fad, and more of a permanent fixture in customer service delivery.
They’re great support for your customer service team, and they’re on track to becoming the next necessity for businesses wanting to keep up with technology innovations.
But that doesn’t mean that chatbots are without their faults.
In fact, chatbots are just as susceptible to the seven deadly sins — Gluttony, Sloth, Greed, Wrath, Lust, Envy and Pride — as the rest of us. A bad chatbot deployment could quickly leave you repenting.
Here, we outline the seven deadly sins of customer service chatbots, and how to avoid them.
Greed: hiding information
Don’t have your chatbot greedily hoarding all the information your customers are seeking.
Think of your chatbot as a tour guide, rather than a gate keeper.
It’s there to show your customers where the treasure is, not to distribute the requested treasure.
For example, a chatbot is great as part of an interactive FAQ experience. It can help customers access answers quickly and easily, and might solve simple queries on the spot.
But you still need to provide that in-depth FAQ page for your chatbot to help customers navigate.
This gives your customers the option of getting their answers directly from the chatbot, through self-service, or by using the chatbot as a stepping stone that signposts them to the right place.
In empowering the customer, you avoid the chatbot sin of Greed.
Envy: pretending it’s human
Chatbots can convince you of remarkable things — like the necessity of clicking a link or giving away your credit card number. Mostly, though, they just want to make you believe that they're real breathing humans.
- Esther Inglis-Arkell
Your chatbot might want to come across as human, but it’s not. One of the most common chatbot pitfalls is treating it like one.
So, allowing your chatbot to be envious of humans is the second of the customer service chatbot sins.
When a chatbot is trying to appear human but can’t mimic basic human conversations, customers are left cheated and frustrated. They aren’t stupid. They know when they’re talking to a human, and they know when they’re talking to a machine.
Now, this doesn’t mean to say you must climb onto your nearest rooftop and announce to the heavens that you’re using a chatbot. Nor does it mean you shouldn’t try to help your chatbot appear human in intelligence.
It simply means knowing the abilities and limitations of your bot.
Sloth: making your chatbot do all the work
This chatbot sin continues from Envy. (That’s right, the balance between bot and human is a biggie.) Automated bots are great, but let’s not forget about human interaction.
Chatbots are good at some things. As technology advances, they’ll become better at more things. But they can’t do everything
(Even with all the lessons in being human you’re going to give it.)
Humans are simply more flexible than bots, and they can handle difficult or obscure queries far better than any bot can.
So, avoid the chatbot sin of Sloth by integrating your chatbot with your human customer service team. Don’t leave your bot to do all the work.
Instead, get it tag-teaming with your human agents to provide a fast, efficient customer experience that blends automated help with human empathy.
Wrath: difficult transitions between bot and human
Smooth transitions between the chatbot and your human representative are essential. The last thing you customer wants is to spend time re-typing their issue to an agent after a bot was unable to help.
The result will be the next deadly sin: Wrath.
Fortunately, this chatbot sin is easy to avoid. You need to ensure that either the thread between the bot and customer is visible to the agent when they join the chat, or that the bot provides the agent with all relevant context before transferring.
When the bot decides to transfer the customer, it should inform them of this decision, and give them the opportunity to leave at any time.
This way, the chat remains seamless, the customer knows what’s happening with their journey, and they don’t feel like their time has been wasted.
Lust: getting a chatbot because they’re ‘cool’
There’s no denying it: chatbots are cool. But rushing to get one without the foundations to support it will cause your customer service to fall short of the hype.
A poor chatbot deployment will only lead to frustrated customers wondering why they’re using a channel that can’t help them. When you show off your cool new toy, it’ll be less impressive to the customers left in the cold.
‘Resist the temptation to chase every chatbot that comes along’
– Chris Connolly
You need to plan before you go live with a chatbot launch. Avoid the chatbot sin of Lust by ensuring that you know how, when, and why you want your chatbot to function, and what queries you want it to handle.
Don’t cut corners with your chatbot’s introduction. Instead, prove that your chatbot is there for your customers, not just a mere gimmick.
Gluttony: forcing your bot on your customers
Forcing your bot on your customers is a sure-fire way to dampen their experience.
When a customer lands on your website, they don’t want to be bombarded with an over-friendly chatbot asking them for their email address or other information. Nor do they want the same generic message popping up the moment they land on each new page.
To avoid this chatbot sin, all you need to do is exercise some moderation.
Wait for the visitor to show interest in your site before you reach out. Maybe they’ve started browsing your help area, or been stuck on your product page for a few minutes. That is the time for chatbot to offer help, or to remind the customer that support is available.
It’s also important that your chatbot doesn’t demand the customer’s contact information before it will help – there will be plenty of opportunity to ask for it when it is needed.
Don’t make your chatbot gluttonous for customer data like email addresses. It’s there to support them, not eat their contact information.
Pride: ignoring the risks
Your chatbot will not be perfect. They’re young, they’re still learning, and they’re not infallible.
In fact, up to 70% of current chatbot interactions fail.
So, the final customer service chatbot sin is Pride; a false sense of complacency that ignores the risks of the other six sins.
It’s easy to be pleased with your cool new chatbot, to think that yours is better and assume that it won’t run into any chatbot pitfalls. But it’s worth remembering that even the most high-tech chatbot still wants monitoring.
Without it, incidents like Microsoft’s Tay, or a malfunctioning chatbot providing incorrect service or broadcasting sensitive data, can happen.
Just as you need to teach your chatbots, you need to monitor your chatbot’s progress and be ready to accept when a chatbot isn’t working properly.
Maybe it needs more data input, or maybe a chatbot isn’t suitable for that specific task. Be willing to tweak the duties of a chatbot, so it can handle the tasks it’s given, and give the extra bits to your human team members.
Eternal damnation, or eternal bliss?
These chatbot sins won’t send your soul to eternal damnation, but they also won’t send your customers to service bliss. In today’s competitive market, customer experience is something that can make or break your customer retention rates.
Chatbots are a great addition to any support team. They work extremely well as a tool to deal with FAQs, simple queries and friendly service.
Just remember to check on them – they are serving at the frontline of your customer service, after all.