Picture of Larry Kim for the Rethink Marketing Podcast where he talked about using chatbots for b2b marketing

How Chatbots Will Transform B2B Marketing in 2018

A billion people use Facebook Messenger every day. Consider reaching them with your B2B marketing chatbot. Larry Kim explains how marketers can sidestep the recent Facebook news feed changes with chatbots and other growth hacks.
Article Outline

A billion people use Facebook Messenger every day. Is your B2B marketing chatbot reaching them?

Larry Kim, founder of Wordstream and Mobile Monkey, believes using Facebook Messenger chatbots is a great opportunity for B2B marketers in 2018; especially, following the Facebook’s news feed change.

“Not a lot of marketers are actually using Messenger to collect customer contacts, and be able to send notifications, and interact with customers one on one through Messenger Chat on Facebook,” Kim said. “But I think that’s a really interesting growth hack for companies looking to kind of bypass the news feed and just talk to them directly.”

We chatted with Larry for the Rethink Marketing podcast. We talked about the recent Facebook changes and what it means for businesses. We also chatted about chatbots and how marketers can use them.

This transcript has been edited for length. To get the full measure, listen to the podcast.

Chatbots for B2B Marketers

NATHAN: One of your recommended hacks for beating the recent Facebook news feed change is with messaging using the Facebook Messenger app. How does a B2B marketer do that?

LARRY: Basically what it is, Nathan, like say you visited my Facebook page and you messaged me, you messaged Larry through my page. Well then we’ve now started kind of a conversation channel that we can just go back and forth forever. Just like how I collect your email, and if I’m a company now I have your email, I can subscribe you to all my podcasts, and all my product updates, and marketing updates. And so that’s kind of the same idea with Messenger. You just got to get people into this messaging channel so you can then develop kind of a one on one conversational relationship with these prospects.

NATHAN: We’re getting into an area where Mobile Monkey is, too. Can you tell us more about what Mobile Monkey is doing with this chat function?

LARRY: Basically, we are doing mobile messaging on Facebook Messenger. I think this is a really big opportunity. I think if you’re a business, and if your customers are on chat like Facebook Messenger every day, then you better have a way to send updates to them via chat. And basically what Mobile Monkey is, it’s like a really easy way to develop those types of interactions.

So for example, autoresponders, like reminders, appointment reminders. Oh, you’ve got a dentist appointment in 15 minutes. Well that should be automated. It should just go straight to their Messenger. Or appointment setting, like if someone wants to sign up for a demo, you could just take the information using chat and have that populate into your appointment book, rather than having to go back and forth over a telephone call. They’re called chatbots. The way it’s different from regular webchat is that we have these AI-powered robots that ask and answer questions based on what we’ve seen in the past.

NATHAN: So let me go back there, so I could have someone come into my website, they’re far enough in the funnel that they want to see a demo, and then they sign up for the demo, and then from there they would just get this chat through the Facebook Messenger telling them when their demo appointment was happening?

LARRY: It’s just like another communications channel. Think of it like email, except that it’s instantaneous, and two-way rather than one-way. So if a business sends you an update through email, I don’t think there’s an expectation that you can just reply to it and get an instantaneous response. But in chat, if I’m the business and I’m sending you an update, like say I’m a pizza restaurant, and I notice that it’s like Friday evening around 6:00, I could just send out a message, you want to order the same thing you had last week. And then you could respond, yes. And then the pizza could just show up. It’s more interacting than email.

NATHAN: And this is all powered by AI. So you could just let it loose and have it do this on its own? At scale too, right? It’s not just Nathan, it’s 1,000 Nathans.

LARRY: Or a million. Well, maybe not for a pizza restaurant. But for a bigger company, absolutely. So all of the things that you typically do in marketing, you can reproduce in a chat environment. Like what are the things that marketers want to do, like surveying a user. Like how did you enjoy the stay at our hotel, like a quick survey. Rather than sending an email that has like a 10 percent open rate, you could send them a chat, just send that message to them via chat, and have them respond very quickly. And maybe instead of a 10 percent open rate, you have an 80 percent response rate. Or like reminders, or all sorts of form collection. So if you want people to sign up for things, book appointments, you can do them via chat rather than a website.

And what we’re finding is that the conversion rates are substantially higher because it’s just a much more natural fit for people on the go using mobile, than trying to navigate some clunky mobile website.

NATHAN: And this is all being run through the Facebook Messenger app, right? A customer doesn’t have to download some other app to see your instant messages or your chat messages or anything.

LARRY: So the business, that pizza restaurant that’s conversing with you, would show up in Facebook Messenger app as just another contact. So you probably have Larry as a contact and all these other contacts. And then you’d have this restaurant. But you could just say, hey, I wanna order something, rather than picking up the phone and calling them. It’s kind of a really cool tech. I’m actually very excited about it. I think every business should have one of these things. And that’s kind of why I decided to make the big move here in terms of wanting to create another business. Because if it wasn’t exciting, I probably would’ve just stayed put.

NATHAN: Well, it presents different challenges for marketers, too, or whoever’s running the app. There’s long-form content you have on a blog post. And then you try and make that shorter in an email communication. But this is even going to be shorter. You don’t have to worry about, ‘hey, Nathan,’ or anything like this. You can just ask the question. Maybe you could even suggest some answers using emoticons or whatever, like how did you enjoy our stay at our hotel, and happy face, sad face.

Chatbot Strategies

LARRY: I think it involves rethinking content. I mean, of course, you could send out links to new blog posts to everyone via chat. You could blast all 10,000 subscribers, like oh, there’s a new podcast, here’s the link.

NATHAN:      And that’s actually good.

LARRY: The problem with that is – that actually works – but I think that’s not the right way to be using chat. I think the way you use chat is that you break down your message into an interactive story. You just say, ‘hey customer, we have this new feature released, just FYI.’ But then the person on the other end can ask questions about it, like how much does it cost, and then you get a response. Can I book a demo? And then you get a response. So it kind of broke up this longer form piece of content into kind of like a choose your adventure powered by AI, where the software learns about your content and can answer questions about it, as opposed to sending it all at once in one big email or one big blog post.

NATHAN: One of your last comments there was learning about you and your business. How fast is it to get up and running with one of these things?

LARRY: It’s kind of like having a kid. Initially, they don’t know how to do anything. But then it starts to learn things. I have a 3 year old. This kid is just soaking up words like crazy. He can read and all this stuff. The more you teach it, the better it works. So one of the things that it does, is it shows you what people have asked in the past. And then you can then train it with specific types of triggers and answers, so that the next time somebody asks for what are your hours or whatever, it’ll give the right answers.

Chatbot Learning

NATHAN: I recall that Microsoft got in a little bit of trouble introducing a bot, and then people started gaming the bot, and it ended up becoming a racist. What are the protections against that from happening to your bot, to a company’s bot, the pizza bot?

LARRY: That’s called unsupervised machine learning. Here, you actually have to train the bot. It’s like say you had a receptionist to answer the phone for your business. I got to believe 70 or 80 percent of the questions that they get are similar. And so you can just kind of train this chatbot to handle all these questions, and then kind of redirect to the appropriate person at some point when they can’t answer any more questions.

NATHAN: I’m wondering, does this fit in at all with voice search as we move forward and we have more and more devices in our homes that are voice activated and that type of stuff?

LARRY: The voice component happens at a higher level. So, voice search and chatbots, if you want to talk to your chatbot or you want to enter a voice query into Google, all that’s happening is that there’s kind of a layer of taking your speech and converting it to text. It shouldn’t matter whether or not people are typing in questions or asking questions using voice. It doesn’t make a difference. But I think the broader theme that you’re talking about here is just mobile. You’re more likely to be on the go using these computers and trying to figure out the answers to the questions that you have. I think that increases the importance for companies to have ways to provide and engage with customers the information that they’re looking for in kind of more lightweight ways.

So if you can imagine like scanning through a big website or kind of download an update, some kind of app for some company, it’s kind of clunky if all you’re looking for is to book an appointment.

NATHAN: So, there’s that primary purpose, that personal one-on-one engagement, short engagement with the customer and the pizza joint. But does the pizza joint get secondary information as they just start engaging with their audience, the million Nathans out there, and better understanding that Nathan likes pepperoni pizza on Friday nights at 7 p.m., sort of the big data mining that you get from all this sort of engagement, does that happen?

LARRY:  Yes. It’s just programming; it’s just like a framework. You could store as much information as you collect. If you’re storing pizza orders, you can absolutely write something that notices that these are the top three pizza, and so those are the ones that we should feature on the welcome page. When someone starts a chat session, ‘hey, these are our top three specials for today based on orders this week.’ Would you like to place an order for one of these three? To answer your question, yes, you can use data and kind of customize the chat experiences in any which way you please.

Unicorns and Donkeys

NATHAN: Larry, I don’t think anyone’s allowed to have a conversation with you without mentioning unicorns. So this is our opportunity to just ask, what do we need to know about unicorns or how does unicorns fit with the chat discussion?

LARRY: I just think that the email open rates are donkey open rates. Typically, these days if you send a blast to your list, what’s a good open rate, like 10 or 15 percent, would you say? But chat, because it’s new, and because you get notifications, that has kind of the unicorn of open rates. It’s typically 70, 80 or even 85 percent these days. So I think people should try it out.

What's New?