Ep #33, Chris McGrath with Tangwork
Chatbots are an increasingly popular topic at internal communication events. But there's a sense of skepticism from some communicators about how successful they will be. And I know many more aren't even sure what they are.
In this episode of ICology, I spoke with Chris McGrath, the founder of Tangowork, about chatbots for internal communications. He provides helpful definitions and use cases for how he sees companies and communicators using the technology. And with billions of people using Whatsapp, Slack and Facebook Messenger, it would make sense that employees are using these apps internally.
If you're curious about chatbots, you'll want to listen to this episode.
LinkedIn Group: Chatbots for internal communicators
Blog post: 10 reasons chatbots will save internal communications
October 12: Why chatbots are the next big thing for internal communicators
October 19: How to design a chatbot for your organization
Chuck Gose: In one of the very early episodes of ICology, I interviewed JoEllen Saeli-Lane, who at the time ran internal coms for CARE, a non-profit down in Atlanta. Now she's part of the CDC. Well one of the big takeaways from her episode was the line, "Go where your employees are." As cliché as it might sound, we know that they're on their phones because we, as communicators, are on our phones and mobile apps make a lot of sense for organizations. I'm a big fan of mobile apps for internal coms, but that does require employees to download a new app, app development, that investment. It's not an obstacle, but something to consider when you're looking at mobile for internal coms.
The chances are that each of you listening use some type of messaging app on your phone. It could be something as native as SMS of iMessage, or perhaps you use WhatsApp or Facebook Messenger, or maybe you use Slack at work, because there's a pretty good chance your employees are also using these apps to communicate with their co-workers. How can a communicator take advantage of this?
This is what we're going to talk about today with chatbots and there's this great quote that I found online that sort of reinforces JoEllen's advice, is rather than force consumers to stop what they're doing and open a new app, chatbots allow companies to inject themselves into the places where people are already communicating. You take out the word consumers, drop in employees, and that's why we're going to talk about chatbots today.
I want to welcome to the show Chris McGrath, founder at Tangowork. Chris, welcome to ICology.
Chris McGrath: Thanks, Chuck. Thanks for having me on the show.
Chuck: Happy to have you on. You're one of the guests I haven't met face-to-face, so we've done this all through digital and social to connect. Chatbots is one of those topics where when I speak at events, it's one of those technologies I encourage communicators to take the time to learn about. It's not necessarily that they should be using them now or they must be using the tomorrow, but it's a chance for them to understand how they could be used, so I hope that helps guide our conversation today.
Chuck: Before we get into the topic, why don't you tell our listeners a bit about yourself, Chris.
Chris: Sure. Well I got started with internal communications through intranets. The first intranet I was involved with was at Mercantile Bank in St. Louis. It was back in 1997. They had 12,000 employees back then before they merged with US Bank. About 10 years ago I co-founded ThoughtFarmer, which is a social intranet software company. Of course intranets, the primary people who are buying them and implementing them are internal communications people. The big innovation with ThoughtFarmer was that it enabled every employee to contribute content which was this blow your mind idea back in 2006 and today it's pretty much accepted. I sold out of that last year and this year I started getting interested in chatbots and then founded Tangowork in the springtime.
Chuck: Tell everybody what Tangowork is.
Chris: Well Tangowork is chatbots for internal communications. It helps you send your important messages to your staff where you know they already are in their messaging apps.
Chuck: Let's dig in now, deeper, into this chatbot topic because I could see how this could be a strange word, a strange phenomenon probably, maybe to some people. I'm going to ask you for 2 definitions of what a chatbot is. First I want to give me the 8 year old definition, if you were to describe what a chatbot is to my 8 year old son. Then give me the real technical definition of what a chatbot is.
Chris: Okay. Okay. Well this should be easy. I've got a 7 year old, almost 8, Rubin. I'll just imagine I'm explaining to Rubin what a chatbot is. Rubin, you know how you see Mom chatting on her phone, typing messages? "Yeah." Who is she talking to? "Her friends?" Well imagine if she was talking to another computer and the computer was typing back to her. That's what Daddy does. That's called a chatbot so when mommy's typing, if she was typing to a computer and the computer's answering her, that's a chatbot.
Chuck: I love it. Now the technical definition.
Chris: Sure. Well to understand chatbots you have to understand messaging apps. A messaging app can be like you said in the intro, just plain old text messaging, or also called SMS. The other common ones are WhatsApp, has more than a billion users. Also Facebook Messenger, right? More than a billion users. They crossed the billion threshold this summer. Also there's messaging within things like Instagram. Skype is a messaging app. Those are all messaging apps. In most of those we're always talking to people, right? The new thing that is now becoming really common is all these platforms have opened themselves up to bots, so you're talking to a computer program rather than a person.
I think the word chatbot is bit misleading because it sounds like it's all chatty and all about the friendly conversation but really chatbot is referring more to the channel. It's using the messaging app channel to deliver an application. The app is within another app. It's within your messaging app. Does that make sense?
Chuck: No, that's really helpful. I would imagine that one of the reasons chatbots are so popular right now is because of the popularity of these various messaging platforms. We talked about WhatsApp, Facebook Messenger. Are there other reasons that chatbots are hot right now and also, I think it'd be helpful for people, if you had a sample of a simple chatbot that someone could use right now while they're listening to the podcast that they could use, or is there an example of a chatbot that they're already using and aren't aware of it?
Chris: Well I think right now, people could say, open up Facebook Messenger. I think most of your users probably have it on their phone, more than a billion people do, and do a search for CNN. One of the bots that comes up will be the CNN bot and add it to your phone. Then you start the conversation by saying, "Hi." Then the CNN bot will answer and say, "Hi there. I deliver news to you. Do you want to read the top stories or can I search for something in particular?" I click top stories and every day CNN sends me the top stories and I can flip through them and if I want to read one of them, I click and it'll either send me a summary, right? Within Facebook Messenger, or I can link out to my browser and look at the actual CNN page. You can do that right now while you're listening. As you said, I think the reason they're becoming so common is because of the ubiquity of messaging apps.
Smartphones have just reached a saturation point where every single person, we can now expect, with few exceptions, has a messaging app installed. In late 2015, the use of messaging apps surpassed that of social networks. Social networks have really had all the press for the last 5 to 10 years. Everybody astounded at how heavily they're being used, but messaging is being used more. If you reflect on your own habits, you see that. I think about this morning already, all the different people I've messaged in several different platforms. Well, I've also used Facebook too. I think for most of us, the use of messaging apps surpasses that of social networks.
Chuck: Now a question I have around those messaging apps is, is it going to be something kind of in weird way, like social media, where there's this pressure to sort of have a presence on every messaging app or on every social media platform? Do you think there's going to be a similar pressure for people to sort of make sure they're on WhatsApp and they're on Facebook Messenger and some of these others so that they are able to message with friends, or do you think there's going to be a natural gravitation, you're just going to stick to the main 1 or 2 that you just use naturally?
Chris: Yeah, that's a good question. Of course there's going to be the dominant ones. One of the ways around that that companies are making are bot frameworks that enable you to develop once and deploy everywhere. That's what were doing at Tangowork. You create your bot once, but it can be delivered to all the different channels. To SMS or to Facebook or to Skype, to Slack, and so on. Each company that's considering deploying one, they do have to consider where their audience is. In the case of internal communications, we're talking about the audience of your employees.
Often times messaging use is geographical. In one area, Facebook Messenger might be dominant. In another area Skype might be dominant so you want to choose where you invest the effort, but you are saved, somewhat, by these write once, run everywhere platforms.
Chuck: I think that'll be good to know for people because it is that challenge of if you're having to recreate something for several different platforms, or is there 1 instance where, regardless for example, if we're going to talk about internal coms, if you're going to go more into that here in a second, that they sort of do, like you said, write once and deploy many times. I think that'll be relief to a lot of communicators to know that that is possible versus having to create something for every platform out there.
Now you wrote a blog post I found. It was called The 10 Reasons Chatbots Will Save Internal Communications. Now we could talk offline about whether or not internal coms needs saved or not, but give the listeners what you think are maybe the biggest 3 reasons that chatbots will impact internal coms.
Chris: Okay, okay. I'll admit that the blog title was maybe a bit of an overstatement. However, yeah, here's a few things. First of all, chatbots reach employees who don't have email or computers. They're often referred to as the disconnected workforce. If you think about industries like retail or healthcare or hospitality, industrial, these are all large segments of the workforce that have traditionally have not been in front of computers and often times have not even had a corporate email address. However, all these people now have smartphones, they have data packages, they're using messaging apps. Chatbots are a way to reach them.
Another reason why chatbots are the next big thing is chatbot messages are easier to read. The medium forces you to be concise. You can't display an essay over SMS, right? You've got to get right to the point. This forces internal communicators who are sharing a message to really distill their message down to the most important 2 or 3 points and send those. The employee receiving those messages is going to appreciate that effort and they're going to digest it because now they can keep up with the most important stuff without wading through a lot of fluff.
What could be viewed as constricting, the short length of the messages, is actually what makes the messaging platform so powerful. Another one, one perhaps we don't often think about is that chatbots are fully accessible by users with disabilities. Now you don't think about this too often until you have a disability. It's the rare intranet that is really usable by someone with, say, a visual impairment. It takes a lot of work, a lot of development work to make an intranet fully accessible, or a social network say, fully accessible. With chatbots, it's the underlying messaging app that looks after the accessibility and these are huge companies that take this seriously. Facebook Messenger is very usable by even someone who's blind. Because chatbots deal primarily in text, a screen reader can simply read those messages.
There's a lot of companies out there that aren't really compliant about what's required of them federally, in terms of support for users with disabilities and chatbots can look after that for an internal communications team.
Chuck: You make a great point. I want to comment on 2 of them. First, the last one, that I will admit, that's an area that I hadn't thought of, of it having a potentially profound impact for connecting employees who might have been inadvertently sort of left out of the conversation because there it was difficult for them to participate, can now stay informed in a way that's easy and intuitive to them given whatever their disability might be.
Then also the second point you made around brevity. We're now sort of all guilty of being headline readers, and I think now that's the challenge communicators face is making the content interesting in a very short period of time so that people know, okay, is this relevant to me or not? It's sort of, in a way, what Twitter has done to a lot of people where some people get frustrated by the challenge of 140 characters. People like me like the challenge of getting a message in in 140 characters because you do have to be sort of creative and very concise and to the point with your words. I think those are sort of great examples, one being making your communication better, and then the other one making your communication more accessible to all audiences that are out there.
Chris: Yeah. If I could add one more, it's that chatbots are both push and pull. If you take email, for example. That's a push medium. As an internal communications team, you choose when to push that message out. An intranet, in contrast, is pull. Employee needs to make the decision to go visit the intranet and read the article. Chatbots are both push and pull so you can send a broadcast message out, like email, but then you could also have an employee pull a message by typing a chat like, "Show me the schedule for tomorrow," or, "What's the news?" The chatbot can respond to that instantly. Push and pull makes it a unique channel for internal communicators.
Chuck: That's a great point because I think that leads into this next question I have for you is, using this sort of push and pull examples, what are some uses for internal communicators with chatbots and maybe give an example of each. You mentioned news being one, but what are some others that maybe people haven't though of where it would be a way for, you said either a company to deliver content to employees, or an employee could request content to be delivered to them.
Chris: One example is scheduling and we're reading right now about a lot of personal assistant chatbots. There's a new one called x.ai. It's a personal assistant and you and I could be saying, "Hey, Chuck. Let's book a meeting for next week," and then we can ask the personal assistant, "Hey, can you find a time that works." It can actually mediate our calendars and find a time. These personal assistants, that's interesting.
Another one is computation. You can ask Siri, "Hey, Siri. What's 23 times 47?" and Siri will give you the answer. You can also imagine, say, a chatbot that's interfaced with your CRM or your salesforce and you can type, "What were sales last month?" Another use case would be engagement. Now an example might be the Whole Foods recipe bot. They deployed that earlier this week so you can add Whole Foods to your chatbot and say, "Recipe for chocolate and orange," and it'll send you some sort of cake that's got chocolate and orange in it and a list of ingredients. I think what that does, I can't actually go and buy the ingredients, not through the chatbot, not yet anyway. It's engagement building, right? It makes me think well of Whole Foods. You can imagine maybe something similar delivering a service to employees that builds employee engagement.
A fourth use case would be purchasing. Right now you can install, or sorry, I said install but you actually just add. You add the bot. You add the Dominoes bot to your Facebook Messenger and you can order a pizza through chat. The first order might take a little longer but what becomes really cool is you say, "Hey, Dominoes," and it says, "Hey. Do you want to do the same order as last time?" And you just click Yes or type yes in your chat and you got a pizza on your way. You can imagine doing this kind of purchasing within a company for common office supplies, say, if you need to reorder pens or Post-It notes.
The last use case is really the CNN use case which is also the Tangowork use case, information consolidation. Taking your news items or your alerts or your schedules and broadcasting those out to employees or letting employees request them. I think that's the last and most powerful use case for internal communicators, using it as a method for getting your news out.
Chuck: Yeah, I'd think it would be interesting too to see if there would be a way to deploy something like this, and I'm being facetious here, everybody loves calling IT to get support for something. Well I'm thinking if there was like an IT chatbot that you could, whether it's logging the ticket, whether it's getting resources for more information, something to where you're not on hold the whole time but you're able to use a chatbot to maybe initiate a process to then get help later on.
Chris: Absolutely. That's a good use case as well. I could see that working.
Chuck: Now as we've talked, chatbots do reside inside a messaging system. In a way that's sort of a limitation to it, but every technology has some sort of limitation to it. Are there any other current limitations that chatbots have?
Chris: Well the chat user interface is certainly limited, right? It would be a difficult thing to design a home, say, via chat, right? I think for the foreseeable future, or probably forever, the graphical user interface will still have an important place in our computer use. It is interesting how some of the messaging apps have been pushing the technology such as Facebook Messenger and Kik have been introducing simple user interface elements to the chat. Instead of typing yes or typing no, you can now display a button that says yes or no. They're experimenting here with bringing a little bit of the GUI into the chat. We'll see where that goes.
Another limitation of course is the brevity of the messages which we've talked about, and I think that could be a strength, but it could also be a weakness. I don't think you're going to redline a legal contract over chat. It's just too long.
Chuck: Now do you see the possibility of where, let's say whether it's in Slack or WhatsApp or iMessage where then that chatbot is, say, the medium between the employee and another piece of software. Maybe they want to post something on the Internet. Do you see where a chatbot could be the facilitator of something like that versus the employee going in to the internet software and uploading or formatting an article or page or posting an image, do you see where a chatbot could sort of be that facilitator for things like that? Is that possible?
Chris: Yeah. Yeah, it's absolutely possible and Slack has really paved the way there with those kinds of bots and introduced the concept. Again, you and I could be in a conversation and we can say, "Hey, what you just said there, we should post that to this forum and we can just message the internet bot," say, and it shows up on the forum. In fact, that kind of interact where we're bringing the application into the messaging app is flipping on its head a model that we've seen over the past 5 years where the final step of the evolution of an application seems to be adding chat.
You think about WebX has chat in it. Google Docs has chat in it. Dropbox has chat in it. Our intranets have chat in them. Our Voice over IP software has chat feature as well and it seems like the last thing that every kind of software adds is chat. Now we're seeing the model flipped on its head where the chat comes first within your messaging app and you add the application into the chat. Now we can initiate a WebX conference from our chat. Now we can edit a Dropbox document from our chat. It seems that the model's being flipped on its head and that opens up some interesting possibilities.
Now if an internal communicator, say they're listening to this and they say, "Okay, I want to build a chatbot for my company," or, "I want to start laying the groundwork to do this," what's sort of a good first few steps for them to take in doing so?
I've developed a model or a project plan around this based on the experience I've had in deploying more than 100 intranets and it's very similar in that you want to start with your business objectives. You want to look at what is the overlap between what the business needs, what your employee wants, and what chatbots are capable of. That overlap is your fit, your sweet spot and you want to find from there your business objectives.
That's often done with maybe an affinity diagramming session with your stakeholders where you brainstorm and then you vote on the results. I also like taking those results and putting them into a user story in a who, what, why format. Something like, as a CEO, I would like to send every employee the daily key performance indicators or KPIs so that everyone gets a quick daily reminder to keep rowing in the same direction, right? Your who, what, why. When you've defined a bunch of user stories like that that you want your chatbot to execute on, then that's going to guide all your development from that point forward. I think that that's a first step to finding those business objectives, getting them into user stories and once you've laid that foundation, you're ready to start building.
Now this is a world that you've made a big investment in by starting Tangowork. I would say you've probably given the chatbot concept and idea a lot more thought than the average internal communicator has. Given your expertise, what would you say, where do you think chatbots will be in the next 2 years, 5 years, 10 years? What's a reasonable time frame and then where do you think it'll be?
Chris: There's a Bill Gates quote that I love and he says, "We always expect things will change dramatically within 2 years," or, "We overestimate the amount of change in a 2 year time period and we underestimate the amount of change that will take place in 10." I don't know exactly where chatbots are going to be in 2, but if we go out 10 years, I think we'll see that the operating systems will have completely evolved to put messaging apps first. In fact, rather than all the icons on the home screen, we'll see some thing that's more time based and people based where chatbots and real people are our interface into everything that we do, especially on our phone and on our tablets.
Chuck: Now recently I was at somebody's office and it was interesting. We made this sort of analogy and this is probably going to age us a little bit but we were comparing it to when iPhones and cameras started having the reversible camera and people were able to take selfies and how weird it looked to see people doing that, whereas now it's 100% common. It's not weird at all. When we were in his office he had an Alexa or the Amazon Echo and he was in his office and he said, "Alexa, dim the lights," and the lights dimmed, which is exactly what it's supposed to do. Then we thought now that seems kind of strange but in 2 years, will we just be talking to things and not having to physically go to switch or physically make a change? Will we just be talking to them to make that happen? That's where I wonder with chatbots, will it just become so a part of what we do that we wouldn't really think of it as a chatbot, it'll just be a part of what we do?
Chris: I think so, especially when we look out 10 years it will probably have transformed us in so many areas. I think we'll see a distinction drawn between voice-based chatbots and text-based chatbots, ones where we're touching and typing and ones where we're talking, because the text based chatbots, when we text we tend to be curt, especially when we know we're talking to a bot and we're just going to give commands like news or show news. When we talk it's hard to be like that. A couple days ago I said, "Hey, Siri, how long does it take to get from EWR to downtown Manhattan?" It couldn't figure out what EWR was but that's the airport code for the Newark Airport and I was trying to remember, is it Newark International, Liberty Newark, something airport. Anyway, in the end I had to bring up Google Maps on my computer to get an answer to that question.
In the future I expect we'll be able to throw anything at Siri or bots that are like it, or Alexa, and we'll get an excellent response. I think we'll have seen the text-based one merged to this balance between typing commands and touching things because a graphical user interface is so efficient. We're going to see some interesting changes over the next few years and chatbots, I think, are going to be at the center of them.
Chuck: Yeah, it's interesting. We'll go back. We talked about our 8 years olds, or your 7, soon to be 8 year old and I'm personally not a big user of Siri but my son is. He gets really frustrated when Siri does not come back with the answer that he is looking for or I think the other night, once again, he's looking at it completely open ended. He asked Siri, "What do you think about Nerf guns?" Siri clearly didn't know what to say, "Well this isn't about me, it's about you." Or something like that. He really wanted to know. I think he was curious to see what the answer was going to be and so as that generation keeps coming up through the work force, they're going to be used to interacting with machines, with software in a way that you and I have sort of had to learn. It's just going to become a part of their DNA in how they operate and what they do.
Chris: That makes sense.
Chuck: Tell everyone more about where they can learn about you, Tangowork, and if there's any other resources about how to learn more about chatbots.
Chris: Sure. Well I have two webinars coming up. One is "Why chatbots are the next big thing for internal communicators," and that's going to go over a lot of the material we talked about today, but also there will be working demonstrations of the software that we've talked about, and then a second webinar, "How to design a chatbot for your organization." That will go more in depth into that question that we touched on briefly. To sign up for either of those just go to Tangowork.com and go to the blog and you'll see those webinars there.
Also just started a LinkedIn group, Chatbots for Internal Communications, and it's my hope that in that LinkedIn group we can forward the conversation that we're having right now. LinkedIn groups are chicken and egg. There's no discussion until people join, but love to have your listeners join us to talk more about this chatbots for internal communicators and just search for that group in LinkedIn.
Chuck: Yep, and I'm now a member of that group and I like the idea because one of the challenges with LinkedIn groups is that sometimes they're so broad that it's tough to get a conversation focused, but I think it'll be interesting to see if you can get some of these communicators who are interested in chatbots and start having some real conversations, some troubleshooting, some examples, because in the world you and I are in on the technology side, especially in the internal communication side, they want to know who at other companies is doing what they're doing, and I think through that LinkedIn group that's a great opportunity for them to discover other internal communicators who do have that interest and can begin building that community. If it as big as ... If chatbots are as big as now that messaging apps are, then it's going to be something that they are going to need to learn about and what better way to learn than from their peers.
Chris: Yeah, that's my hope too.
Chuck: Well, Chris, I want to thank you for coming on ICology. This is a great topic. Some might think it's super forward thinking and pie in the sky, but I more reside on your side of things where I think it's going back to what JoEllen Saeli-Lane said where, "Go where your employees are." You said a billion people are using some of these apps. Your employees are using them. That might be something to begin checking out as finding out how are people using WhatsApp, how are people using Facebook Messenger, and begin scoping out and using those plans you laid out for everybody to figure out how would they like to use a chatbot for their company. I would recommend for them, probably you would too to start with something very basic but also something very core to the business so people begin to see the value, and then grow it from there outward.
Chris: Yeah, that makes sense. The best chatbots are specific. If it's too open ended, it becomes too difficult to program, but if you focus on a few core objectives and design it to fit those user stories that you prepared, then it will be a success.
Chuck: Well I agree, Chris. Thanks again for coming on the show. ICology now has a home you can visit at learnicology.com to catch up on old episodes, read transcripts, blog posts, check out events that I'll be on or ICology is sponsoring. That's learnicology.com. Also keep an eye out for Something Else. Again, this is the video series where I ask the internal communications community a question and you provide the answers. The first one will come out in early October. Please follow ICology on Twitter, @LearnICology to hear about Something Else as well as to pick up show announcements and other IC news. If you're not already a subscriber, listen to ICology on iTunes or wherever you get your podcasts. Simply search for ICology. If you like what you hear, reviews are always appreciated, especially those 5 star reviews. If internal communications is your passion, ICology is your podcast. Thanks for listening in.