New Landscape for Customer Interaction: Chatbots, AI and Messaging
There’s been a lot of hype about chat and messaging tech.
I thought it’d be useful to delineate some of the different terminologies and practices that are hot topics for the digital world (and maybe your website).
Live chat – Continued Dominance and Ongoing Optimisation
- The oldest and most trusted technology on the board
- A real-time, text-based, human to human messaging solution
- Typically delivered in a one-to-one setting
- Great for connecting your website visitors with actual members of your team
Good ol’ live chat. It’s the backbone of digital engagement strategies because it works. We’re talking 5 times higher conversion rates and 50% more leads. The key thing for live chat is that someone from your team needs to be available on the other end of your website; that human-to-human engagement and experience is critical. These conversations happen in real time and both the awesome dev team at live chat and managed chat providers are constantly working to improve its efficiency. It’s a lean, mean, customer-serving, lead-generating machine.
Messaging – Growth
- An asynchronous, text-based, human to human messaging solution
- Great for follow-up or enquiries that aren’t time-sensitive
Some exciting innovations are happening in the online digital comms world, and messaging’s picking up a lot of steam. But what’s the difference between messaging and live chat?
Messaging’s big differentiation is that it’s asynchronous; your team doesn’t have to be there for an interaction to take place.
So, when you as a customer send a message, you might get an automated response saying ‘we’ll respond in 30 minutes.’
Now, from a business perspective, you might not need that instant availability at all times, so a delayed comms portal (like a decent messaging app) can step in. And that makes this easier to resource than chat. But, from your customer’s’ perspective, there are challenges. And those challenges beg the question: Do customers expect an instant response if they use something that looks, feels, and acts like live chat?
Messaging solutions will do things like capture email addresses so that your team can then initiate an email-based conversation once someone does come online to check these. It’s a bit more of a convoluted journey for both your team and your customers. Messaging’s trying to mirror that Whatsapp style of comms that we’re hearing a lot about.
But there are some challenges on a website if we’re not offering that instant response because people will drop off your radar. All the research points to the fact that there’s a correlation between the speed of a response and customer satisfaction. So, bear in mind that even though messaging might be super easy to manage, you’ll need to work to ensure response times aren’t negatively impacting your goals with this platform.
Parting thought: If you’re using this in a sales setting, speed is of the essence and I always steer people to manage this in a Live Chat-type way, where instant response is possible.
Bots – Lots of Buzz
- A real-time, text-based, machine to human messaging solution
Chatbots are essentially a pre-set, animated, predictive order of question-and-answer with various logical options to flow through. One delineation I want to make here is that a bot is not, in and of itself, intelligent. A bot is just following a pre-set, pre-ordained set of questions and responses. If you ask it something out of its remit, or if you misuse a word in your question, or if you break its logical flow, often these things break.
Bots have a lot of excitement around them because of the automation aspect, which is a similar thread to why messaging is so appealing – it’s the idea that we don’t have to be around all the time or respond immediately (and therefore, they’re a lot easier to manage and maintain on your site). Some people are really excited about this because staffing your website chatbots means you don’t need as many humans (a bit sad). If you’re talking with your team about cost-effective support, chatbots might be a lower-cost alternative to a resourced human engagement strategy on your digital platform.
On the topic of chatbots not being human, bots don’t sleep so you can run them on your site 24/7, which is pretty neat.
Now, there’s a lot of traction for bots that have handoff processes whereby if the bot does get stuck because the human it’s talking to isn’t perfectly logical, it can reach out to your team and pass off to a human agent who can take your customers on the rest of their journey.
The bot script and the bot management process is going to take a lot of work to get right. There’s a number of articles and threads (just have a Google) out there suggesting that bots are going to create a whole new category of employment avenues (bot knowledge managers, bot flow managers, bot flow optimisation managers, bot…we don’t know yet). So in essence, the script and the process and the approach that the bot takes needs to be managed and optimised in much the same way that chat flows and scripting needs to be managed, optimised, and the knowledge and technology need to be revised and optimised. So, this is not a shortcut to create a great experience: it’ll require much of the same effort as live chat.
Parting thought: Still lots of work to be done before they become massively useful, but right now chatbots could be great for simple FAQs, order updates, or ordering a pizza. In more complex or value-driven transactions, you may want to be leaning more toward the human to human interaction.
AI – More Buzz than Bots
The difference between bots and AI is that AI is intelligent. It gets better over time. It betters itself automatically through a sequence of machine learning, Big Data, and all those other buzzwords you wanna throw at it.
Now, full disclosure: These solutions are still quite nascent. There’s not much on the market we see being put into use effectively where it’s taking a machine learning rather than bot approach. A lot of the talk still has been around bots (and there’s the misconception that bots are actually machine learning or that they’re intelligent in some way, iterating themselves over time).
So, it’ll be interesting to see where this technology goes in the next few years. Maybe it’ll convince a customer chatting with a bot that they’re actually chatting with a human? Skynet is here.
It’ll be really interesting to see how these quadrants combine over time, because there’s so much overlap and opportunities for interaction methods and modes in the text-based communication space where we mix H2H and M2H together.
The take away: M2H is nascent with large potential in low-value interactions, but still imperfect; as AI develops, I expect we’ll see a lot of movement on the machine front.
H2H is the known way of working with customers, but it’s rarely optimised. Still companies struggle to activate messaging and chat to their full potential with channel-specific training. And maybe that training will soon be extended by some cool AI integrations. Some centaur experience is on the horizon.
If you’re looking to grow your business in 2018 and beyond, you need to make sure your digital engagement strategy is aligned with your marketing and CX goals.