Rashmi Vittal, John McCrea
December 18, 2020

Powering 2021 B2B Marketing with Conversational AI

Guest: Rashmi Vittal, John McCrea - CMO, Conversica, Amplify.ai

In the premiere episode of Renegade Thinkers Live, Drew was joined by CMOs Rashmi Vittal (Conversica) and John McCrea (Amplify.ai), to discuss all things conversational AI. We take stirring things up quite literally here at Renegade, so in this episode, we sipped in the latest AI marketing insights along with Letherbee’s original label gin (which was amazing, by the way).

It was a fun episode, filled with AI conversational marketing use cases, wisdom from Dr. Benjamin Franklin, and a Live audience Q&A. Even Google Chrome makes an appearance! Tune in to hear some incredible use cases, learn why John and Rashmi both say that “bot” is a terrible word, and see just how Intelligent Virtual Assistants (IVAs) can transform B2B marketing in a digital-first world.

What You’ll Learn in this Episode

  • How conversational AI can optimize B2B marketing
  • Common conversational AI mistakes made by B2B marketers
  • How virtual assistants are becoming more human

Renegade Thinkers Unite, Episode 219 on YouTube

Resources Mentioned

Time Stamped Highlights

  • [1:11] Digging into Renegade Marketing with Rashmi Vittal of Conversica
  • [6:11] AI Virtual Canvassing with John McCrea and Amplify.ai
  • [9:08] Defining Conversational AI
  • [10:48] B2B Use Case: Century Link Sales AI Assistant
  • [13:57] The Omnichannel Authority of Conversational AI
  • [18:51] How Conversational AI Builds on Data Sets
  • [22:08] Will Virtual Assistants Replace BDRs?
  • [23:42] Can AI Have Emotional Intelligence?
  • [27:46] How Data Privacy Rules Are Evolving
  • [32:03] Why You Should Stop Using the Word “Bot”
  • [34:37] How IVAs Can Align B2B Marketing and Sales Teams
  • [38:08] Augmenting Sales Teams with Conversational AI Tools
  • [41:11] Human AI Interactions for Social Betterment
  • [50:30] Choosing an Intelligent Automation Platform for Your Business

Transcript Highlights: AI Conversational Marketing with Rashmi Vittal and John McCrea

Cold Open: Introducing Renegade Thinkers Live

Drew Neisser: Hey, it’s Drew, and this episode is actually from a November livestream which is part of our new series called Renegade Thinkers Live, and this was the first episode. There are definitely a whole new set of logistical challenges and plenty of surprises when recording live, but it was truly delightful to sit down with two savvy CMOs—Rashmi Vittal of Conversica and John McCrea from Amplify.ai—to talk all things conversational AI.

You’ll learn a lot from this episode—I know I did. Did you know that bots—sorry, I should say “Intelligent Virtual Assistants” of IVAs—have come such a long way that they can not only filter through high-quality leads and, in Century Link’s case, deliver an 8:1 ROI, they can also align marketing and sales teams and speak to customers in a remarkably human way. Rashmi and John have a lot of great insights to share in this episode for B2B marketers. I hope you enjoy the show.

[1:11] Digging into Renegade Marketing with Rashmi Vittal of Conversica

Drew Neisser: Welcome to the premiere episode of Renegade Thinkers Live, a show that promises to be long on marketing insights and short on over-hyped buzzwords. We’ll be drinking in the latest tactics and the coolest gin. Yep, this may be the only live marketing show that also features a gin tasting. This show is being livestreamed via our new friends at Restream, and if you want to drink along at home, today’s gin is Letherbee. I’m your host, Drew Neisser, live from my home studio in New York City, and as I like to say on my podcasts, “Hello, Renegade Thinkers!”

On today’s show, we’re going to have a conversation about conversational AI. If you don’t know what that is and why it’s likely to transform B2B marketing, you’ve come to the right place. We’ve got two human experts to help you understand how savvy marketers are already using virtual assistants, AKA bots, to accelerate the sales process and improve the customer experience. It’s exciting, it’s mind-boggling, and it’s a bit scary, so I, for one, am really glad we have a gin tasting that will be starting about 25 minutes from now.

Alrighty then. Let’s meet our guests. First up is Rashmi Vittal, the CMO of Conversica, a Silicon Valley veteran with stints at IBM, Oracle, and SAP, and perhaps most famously known as the star of episode 201 of Renegade Thinkers Unite. Rashmi, welcome to Renegade Thinkers Live.

Rashmi Vittal: Hey, Drew. Happy to be here and great spending this time with you.

Drew Neisser: Well, some might think you’re a glutton for punishment for being interviewed by me twice in a matter of months, but we’ll let you off the hook here.

Rashmi Vittal: Hey, you toss a little gin in there and I’m gonna show up.

Drew Neisser: Well, there you go. All right. I knew there was a good reason we had that. Anyway, let’s start on a very human note. What makes you happiest in your life?

Rashmi Vittal: Oh my god that’s deep. Wow, I think I need a shot before I can answer that question. Gosh, you know, wow, man, that’s a toughie. Okay, you’ve got me on the hook. Real quick. What makes me happiest in my life? I mean, I think it’s just feeling like I’m being the best version of myself and helping others be that as well. So, holding myself to kind of keeping in check, and don’t take yourself too seriously or don’t get too intense. Ratchet it down and have some fun, as well as making sure you’re doing good things. I’m happiest when I feel like I’m doing the things that I know I can be doing and I should be doing that allows me to be my best self. And then also, you know, I’m happiest when I work out. Let’s just be honest, the endorphins kick in and it makes you feel really darn good.

Drew Neisser: Well, that’s great, and I appreciate those comments. When you think about your career as a whole—this is the Renegade question, the toughie—what do you think is the most Renegade thing you’ve done?

Rashmi Vittal: Renegade as a whole, the most Renegade thing I’ve done. In the context of as a B2B marketer?

Drew Neisser: Yeah, sure.

Rashmi Vittal: Yeah. Wow. You know, I really loved what we did last year at the OpsStars event in 2019, where we transformed the room that they gave us into a beer hall, like a taproom. The whole idea was to come in and untap your full sales and marketing potential. That was the whole idea with an intelligent virtual assistant augmented workforce, and we had the most people in our room, people were coming in, had demos, all that stuff for sure, but it was the environment of a living room quasi gastro pub and people just had fun. They had a tremendous amount of fun with it. A lot of people were hanging around drinking some good beers. We made sure we had good beers and we had pretzels. We had stuff where people could come, sit down, have community, and talk shop. And I think that to me was pretty fun and Renegade-ish perhaps.

Drew Neisser: Oh, well, totally. Really, hearing that story, it makes me miss physical events so much. That’s one of my favorite things in all of it. We can connect the dots from what makes you happy, a little bit of fun (and it sounds like drinking), and an event that you did, and it sounds like drinking.

Rashmi Vittal: If my mom sees this, she’ll kill me. No, no, I’m really not… [Laughter]

[6:11] Virtual Canvassing with John McCrea and Amplify.ai

“We've been very involved in the 2020 election cycle, applying conversational AI to do voter engagement at a superhuman scale.” @johnmccrea @AmplifyDotAi #RTU #podcast Click To Tweet

Drew Neisser: Now, let’s welcome our second guest, John McCrea, who is the CMO of Amplify. Hey, John.

John McCrea: Howdy.

Drew Neisser: He’s been a serial entrepreneur since 1994. John co-founded two companies—MediaSpike in 2012 and Ethereal VR in 2015 and joined Amplify in 2018. Hey, John, welcome to Renegade Thinkers Live.

John McCrea: Thank you, and apologies—a leaf blower has shown up out of nowhere.

Drew Neisser: You know what, I think you’re okay. It’s funny—we had a helicopter flying over our apartment building here in NYC, so, you know, anything can happen on a live show, but you were expecting, I know, a question, but not this one: my question for you is: Why do you think the world would be better off if women ran everything?

John McCrea: Well, that’s a great question. Certainly, when you look at the mess that the country is trying its best right now to transition out of, there’s an awful lot of thanks for the leadership role of women in exercising their vote wisely.

Drew Neisser: You know, I’m absolutely in love with the prime minister of New Zealand. She’s an amazing lady and it’s just, it occurs to me and I’m sure others that there are some just outstanding women leaders in the world and we need more of them. All right. Well, that is a good subject for the two of us, but let’s talk conversational AI. Rashmi, why don’t you come back in? Thank you, Melissa, for that. We’re really testing Restream here. John, can you give an example of how one of your clients is using conversational AI?

John McCrea: Sure. As you know, we’ve been very involved in the 2020 election cycle, applying conversational AI to do voter engagement at a superhuman scale. There are so many examples there, but one in particular that I think is quite pioneering is, you may have heard of this organization VoteVets.

They’ve been doing some great ads this election cycle and really great work. Together with them, we pioneered using conversational AI to do virtual canvassing in battleground states, and that was a way to find either—in fact, we were looking for undecided voters who could be potentially persuaded and that was all done through conversational AI, and then with a seamless handover to a human volunteer.

[9:08] Defining Conversational AI

“#ConversationalAI is a series of technology components that allow a program to actually engage in a natural language conversation with a person.” @rashmivittal10 @MyConversica #RTU Click To Tweet

Drew Neisser: You know, I realized one thing right away, which was that I should have asked both of you, and I meant to ask both of you, can you give a definition of what conversational AI is in 10 words or less? Rashmi, why don’t you give it a shot?

Rashmi Vittal: Conversational AI is a series of technology components that allow a program to actually engage in a natural language conversation with a person. It’s part of an overall solution, but it is a key component to do that natural language engagement.

Drew Neisser: Okay. And John, how about a description of what that is? 10 words or less.

John McCrea: Sure. I think it’s the Turing Test in the narrow, so purpose-oriented conversation that might be hard to tell whether it’s a human or software.

Drew Neisser: Oh, interesting. Okay. And speaking of artificial intelligence, I’m just going to ask, “Hey, Google, what is conversational AI?”

Google: On the website bold360.com, they say: “Conversational AI is the technology that makes that possible. It allows artificial intelligence technologies like chatbots to interact with people in a human-like way. By bridging the gap between human and computer language, it makes communication between the two easy and natural.”

Drew Neisser: Okay. Thanks, Google. A few more than 10 words or less, but it’s sorta fun to use a conversation with a bot if you will, to talk about that.

[10:48] B2B Use Case: Century Link Sales AI Assistant

#CaseStudy w @rashmivittal10: @MyConversica #SalesAIAssistant helps @CenturyLink contact 90,000 prospects/qr, identify 30-100+ qualified, hot opportunities each week, & deliver an 8:1 ROI on a monthly basis. #RTU Click To Tweet

Drew Neisser: Rashmi, give us an example of how one of your clients is using conversational AI.

Rashmi Vittal: Yeah, so our clients are a little different. We really sell into enterprise, mid-market, even SMB markets, but more in that B2B space, so around technology, telecom, manufacturing, financial services, et cetera. Let’s just pick a telecom company. Century Link uses our Sales AI Assistant to contact 90,000 prospects every quarter.

Through this, it’s contacting prospects very similar to what you would expect a sales development or business development rep do. It can actually begin the engagement with the lead or prospect. It will send out a different personalized email with every cadence. It will persistently do that until it receives a response back. When it receives a response back from the prospect, it can easily understand and decipher what’s being said, what is the intent or the request within that response. And then it will craft a message in kind appropriate to its understanding and send it back to the prospect.

Now that could be the prospect might say, “Hey, I’m sales-ready. I’m ready to have a conversation with sales.” It might say, “Hey, come check back later. I’m not really ready.” It might say, “Hey, help me learn more.”

In any one of these cases, the IVA, the intelligent virtual assistant—which is not a bot, and we’ll get into that a little later—the intelligent virtual assistant will actually take the right action and send the right email back to the prospect. It will notify the rep that owns that contact and account, and it knows that because it has direct integration into your CRM. It will do a lot of things. It’s a very busy bee, but what this has been able to do with Century Link is identify over 30 to 100 qualified, hot opportunities each week for them that has resulted in delivering an 8:1 ROI on a monthly basis.

Just imagine that. You take something like 90,000 of these prospects, you’re able to work, work, work them, you’re getting 100 hot qualified opportunities. These are people who are raising their hands saying, “I’m ready to have a sales conversation” and that is resulting in over 800% of an ROI based on what they spend on us.

It is just a tremendous force multiplier for a lot of organizations that we’re seeing, especially when you’re trying to solve that common problem that marketing and sales have, which is, sales always want what? Better quality leads. And marketing feels like they’re sending it to them. Well, the only way sales will accept those leads, or at least have a better likelihood of accepting those leads, is if they’re actually raising their hand saying, “I’m ready to have a conversation.”

Drew Neisser: Right. There’s a little bit of begging going on between marketing and sales. I’m begging you to make this an SQL.

[13:57] The Omnichannel Authority of Conversational AI

“We really believe that the #IVA itself speaks on chat, email, & SMS so that you can deliver that continuity of conversation across the channels for that contact, that prospective buyer. @rashmivittal10 @MyConversica #RTU” Click To Tweet

Drew Neisser: So my question on this is, first of all, it’s fascinating, but I don’t quite know if I got all the steps in there. At any point in time in this virtual engagement, does the prospect know that they’re dealing with a virtual thing?

Rashmi Vittal: When we talk to our customers, they say 9 out of 10 times, their customers have no idea that the person that they were engaging with was not human. In fact, it’s so human-like and so polite and professional in the way that it’s interacting, that oftentimes they’ll even say, “Gosh, I loved Ashley. Can she join our call? Is she going to join us here today?”

That just tells you that they don’t know that they’re interacting. Now, what we advise clients, you know, it’s up to you. If you want to stay in there and let’s say–you can name your IVA anything you want. If you’ve named it Roger, if you’ve named it Candace, it doesn’t matter, but at the end of the day, it’s one of those things where you could say, “This is a sales assistant, this is a business development assistant.” You could put something in there that lets them know that it’s not a real person, but oftentimes they don’t look at that detail. They just go, “Wow, this is a very natural, polite interaction. I’m enjoying this experience.

Drew Neisser: And is this interaction email primarily? Or is it email and text, SMS, as well as all of the above?

Rashmi Vittal: Yeah. So, we support three channels. We support website chat, so if you have website visitors coming onto your website, and oftentimes they’re just trying to, especially in a B2B marketing world, you’re just trying to connect them a lot to information that they’re seeking, right? So, whether that’s “I want to find the product page,” or “I want to find these resources,” or maybe they were ready to contact you or schedule a demo and you can easily get that scheduled, or maybe they need to talk to a rep or ask a question and get a quick answer back.

We support website chat, but the key thing about that and what’s really cool is that it’s not just a chatbot. We really believe that the intelligent virtual assistant itself speaks on chat, email, and SMS so that you can deliver that continuity of conversation across the channels for that contact, that prospective buyer, or even customer.

It can also then be somebody—let’s say you came in through an inbound marketing lead. Content syndication, webinar programs, whatever it is that’s flowing through the system, those leads can wind up in a campaign, let’s say in a CRM campaign. And then that will trigger the IVA to begin the outreach, much like an SDR would begin that inbound outreach via email. Then you can engage in that email back and forth.

The cool thing too is that, depending on where you are in that engagement, if you wanted to actually, for your buyer’s journey, if you said it made sense for that email engagement to then switch into an SMS text, then that’s another option that can happen. We use email to drive the opt-in, so everything is privacy compliant on mobile, and then we can continue the conversation on SMS text. The IVA can actually speak across chat, email, and SMS, and can do so driving that continuity of message based on where the buyer is in the journey.

Drew Neisser: John, what do you make of that relative to what you all do?

John McCrea: Actually, it was so fascinating to hear because it’s very similar to how we think of it although we’re on the B2C side. We use the phrase “branded virtual assistant” because often, our client is a brand and we’re helping them engage with prospects—consumer prospects—and customers at a much higher scale. But that notion that the conversation is stateful and that it’s omnichannel, it’s just that in B2C, in addition to webchat and SMS, it’s everything from Facebook and Instagram and Messenger and WhatsApp, but also all the way over to conversational ad units. And now with Google Business Messages to conversations that can arise out of Search or Maps that have continuity.

[18:51] How Conversational AI Builds on Data Sets

“In conversational AI, ideally, you would want to have a large conversational data set.” @JohnMcCrea @AmplifyDotAI #RTU Click To Tweet

Drew Neisser: You wrote an interesting article—I think you just published it yesterday and I love this line. It said, “The first 10 billion engagements are the hardest.” I had to laugh because that’s a lot of engagements, John.

John McCrea: It really is. And I used that to frame the idea that conversational AI is hard, it’s just now crossing into the mainstream and we’ve had to overcome a significant chicken and egg problem, which is: Machine learning is at the core, but machine learning has become so successful now in the world of AI because of the existence of large data sets, so for image recognition, if you have a large number of photographs that you know what they are, you can train the AI on that.

In conversational AI, ideally, you would want to have a large conversational data set, but most NLP engines have been trained on book prose, which is to say perfect spelling, punctuation, sentences. And what we see in B2C messaging-based communication is a lot more vernacular, slang, emojis, misspellings. We had to build our own NLP engine that could be really ideally suited to that.

Drew Neisser: Got it. And my question for you­­—oh, we just put the link to John’s article in the notes there—Rashmi, to get your system to work, how did get enough information into the system so that this virtual assistant was really smart enough?

Rashmi Vittal: Not unlike John’s situation, I mean, we’ve been around for 12, 13 years, right? We’ve had the benefit of having hundreds of millions of interactions flowing through our system, which has just really created robust AI data models in our platform that allows then our IVAs to perform with really a high degree of accuracy.

We have a 98% accuracy rate in the way that we comprehend and then take action on any messages that we’re receiving. That’s really high. And in fact, I would submit that that’s higher than your average sales development rep in terms of the level of accuracy and decision-making that it’s able to do. That’s really important. There’s no doubt that you need to have a data-rich backbone of the various interactions feeding your systems. And I think, with the focus on the use cases that we support, we’ve been able to really get that groundswell of data input into our system so that it can be that accurate.

[22:08] Will Virtual Assistants Replace BDRs?

“Instead of having to hire those 10 SDRs that you think you need to hire, maybe you only need to hire four or five, and one IVA will cover the rest.” @rashmivittal10 @MyConversica #RTU Click To Tweet

Drew Neisser: So, if I’m a BDR—that’s brand development rep—should I be looking for a new career?

Rashmi Vittal: So BDR…A sales development or business development rep?

Drew Neisser: Can you tell me…which is going to be replaced by your Ashley, the virtual assistant?

Rashmi Vittal: Yeah, so it’s going to be that inbound SDR. They better get on that AE track pretty fast. But in all seriousness, I think where you see this going is—I’ve always said, what can be automated will be automated, it’s just a matter of time. What we’re finding is that the whole pursuit around lead qualification, really finding whether those leads are sales-ready or not especially on those inbound leads, is something that we’re really doing well.

Now to the point of what we’re seeing is, are we right now entirely replacing SDR or BDR teams? No, we’re not, and we’re not suggesting that either. What we’re seeing is that, instead of having to hire those 10 SDRs that you think you need to hire, maybe you only need to hire four or five, and one IVA will cover the rest. That really creates a different dynamic within how you’re thinking about your sales workforce in the future because if this is able to serve up more, then you just need more people to take it on the receiving end and push it down from that SQL all the way to Close 1.

[23:42] Can AI Have Emotional Intelligence?

“We've had in-depth conversations with 100s of 1000s of Trump supporters...We have found ways to deescalate, to get them to engage, and to feel like finally, someone's listening to me.” @JohnMcCrea @AmplifyDotAi #RTU #AI Click To Tweet

Drew Neisser: Gabi Zijderveld, a CMO that has been on Renegade Thinkers Unite and is in the world of emotional intelligence asks… And I have to read her first question, hang on, let me pull it up. The first question was related to the fact that AI is transactional and lacks emotional intelligence. First of all, do you agree with that? And second of all, how do you fix that?

John McCrea: You know, I think it comes down to what is kind of a new job title, which is, in this world of conversational AI, the UX design challenge is fundamentally different than what a generation of web designers have become so good at. This emotional intelligence piece of it really is fundamentally about design that’s understanding where are you likely to get friction points, how can you phrase answers to that that give the equivalent of empathy while in the absence of actual emotion.

Drew Neisser: Rashmi, what do you think? AI just doesn’t have that emotional intelligence and is that a limitation?

Rashmi Vittal: Well, like anything else, I think the emotional intelligence can be programmed and built into the AI models. It takes time and it takes a lot of testing to make sure that the nuances of the language don’t get misinterpreted. But I don’t see any reason why you couldn’t build that as long as it’s appropriate to the use cases and the personas that you’re supporting, in terms of who’s on the receiving end of this. I think it is important and I think you will see that evolve over time.

And if I would add, not only do you need to see that, but you also need to think about empathy comes with memory, the experience one has had with that person. I mean, there’s empathy in real-time like when you hear something for the first time, even if it’s from a stranger, how do you react to it? But there’s also this notion that, you know, I submit that IVAs are going to really help along the entire customer revenue life cycle, not just in one part of it.

From website visitor to prospective buyer to customer to advocate, right? There are different instances where an IVA is going to be used, and the key is can that IVA remember and have memory and context about what that journey was between the customer and the vendor that they’re interacting with, and if they can, can they apply that to deliver on that, continue to deliver on that human-like personalized experience? And my answer to that is, yes, it’s just going to take time to build out, but I think it’s totally doable.

John McCrea: I would add if I may one thought here, which is, you know, when we took on the political market opportunity, we knew that we were going into the very deepest end of the pool with respect to conversational AI.

Right now in America, political discourse is so polarized, it’s broken. Humans are having trouble talking to each other in an empathetic way and we’ve found at scale millions of voters… For example, our clients have all been democratic, but we’ve run into voters across the spectrum and we’ve had in-depth conversations with hundreds of thousands of Trump supporters. Many times, they come in quite angry and we have found ways to deescalate, to get them to engage, and to feel like finally, someone’s listening to me.

Drew Neisser: Wow. I have to get clarification from Rashmi. IVA. The “I” stands for what?

Rashmi Vittal: Intelligent.

Drew Neisser: Intelligent virtual assistant. Okay. Intelligent, got it.

[27:46] How Data Privacy Rules Are Evolving

“We're going to continue to evolve the data privacy rules and regulations to make sure that not too many things are falling through the cracks that might feel too creepy.” @RashmiVittal10 @MyConversica #AI #dataprivacy Click To Tweet

Drew Neisser: Some of this is kind of mind-boggling. It also—I’m imagining this scenario. I happen to have a very public footprint and I’ve seen applications—I think it’s xiQ—which has the ability to just look at someone’s social profile and then create a persona for someone to sell against.

I’m imagining that your system could also, in either case, suck in all of the social data that is out there and therefore be so much smarter and potentially empathetic to this individual. How much of this process is gathering some of that external data? And at what point do we cross the line from helpful to creepy?

Rashmi Vittal: Well look, I think in any situation you need to follow the data privacy and compliance regulations that are out there. I’ll take this from my perspective—it comes to, what is the profile of the customer? And there are different ways to build a customer profile. There are unidentified unknown ways, right?

Like if you think about cookie IDs and then what it captures and behavioral data, et cetera. And then you think about deterministic data, which is the customer has voluntarily provided information. They have opted in. If you take that opted in data and you can use that to create a profile to drive the personalization, that’s great, and that’s internally managed and maintained data.

When you go outside of the four walls, you’re going outside and you’re probably connecting through an API, and most of the APIs you are connecting into, the data you’re getting access to has in and of itself privacy constraints or restrictions, et cetera. So you’re working in good faith with the vendor on the other side, the partner vendor that you’re dealing with, but the information that you’re acquiring is indeed information that the customer profile has allowed you to receive. You are then using that as building out further your profile.

But I think, look, as we continue to expand the boundaries on this, or test, I’d say, the boundaries on this, yeah, I think over time there’s gotta be additional—not additional—we’re going to continue to evolve the data privacy rules and regulations to make sure that not too many things are falling through the cracks that might feel too creepy.

Drew Neisser: John, you have anything to add on this issue of privacy?

John McCrea: Yeah. We definitely aren’t doing that kind of scraping of public information, but we do find that having a conversation with someone through a messaging channel, there’s an intimacy to it and a willingness people have to share.

What we think works well all along the customer journey is, the more that the person shares, the more that we can personalize the experience for them in a way that’s adding value. In e-commerce scenarios, where we’re often integrated with the store itself, like through a Shopify integration—in that way, we can also get buyer history. Essentially welcoming someone back, knowing who they are, knowing what they like, what they care about, kind of like a concierge.

[32:03] Why You Should Stop Using the Word “Bot”

“#Bot is a terrible word.” @JohnMcCrea @AmplifyDotAI Click To Tweet

Drew Neisser: John, I feel like yours is more of a bot, and Rashmi, you’ve corrected me and called this a virtual assistant. Is there a difference between a bot and a virtual assistant?

John McCrea: Well, first of all, “bot” is a terrible word.

Rashmi Vittal: Thank you!

Drew Neisser: Well, that’s a bad word, but okay, fine.

John McCrea: And it’s a word that’s gotten worse since the 2016 election. It’s now associated with, um, Russian bots and trolls. I think also, you know, chatbot is a term that evokes technology from five years ago or something like that. Very simple things that are more TAP oriented and which fundamentally don’t do natural language processing. I think we’re a generation beyond that and we’re, now with conversational AI, climbing out of the trough of disillusionment from the chatbot era.

Drew Neisser: Got it. I think. Okay. So, I am not allowed to use the word “bot” for the rest of the show.

Rashmi Vittal: Please don’t.

Drew Neisser: Perfect.

[34:37] How IVAs Can Align B2B Marketing and Sales Teams

#WhyAI: “What if you could actually surface more sales-ready opportunities earlier and faster in your sales cycle so that your salespeople can focus on those?” @rashmivittal @MyConversica Click To Tweet

Drew Neisser: Let’s face it—probably a bunch of marketers are listening now or will tune in later on. What do you think are the most effective ways that marketers aren’t using the tools right now that they should be? Let’s talk. I know that, John, you’re a B2B marketer who sells to B2C organizations, but we’ll start, Rashmi, with: What are they missing? Why isn’t everybody doing this right now?

Rashmi Vittal: Well, I think it’s a great question. I think there’s a lot of confusion in the marketplace. Look, if you look at the LUMAscape and the MarTech stack, there’s like, what, 5,000 different vendors? I mean, it’s daunting for marketing leaders to figure out how do I stitch together the right tech stack to support my business. But that being said, time in and time out again, again and again, it always comes back to this issue of, “I don’t believe”—like, marketing leaders are saying this—”I don’t see the return on the marketing investments I know are good that I’m making.” And sales leaders are going, “I’m just not getting enough pipeline to help me close my deals.”

How many of you have your weekly pipeline calls and forecast calls where this is the conversation, and then you go scurry off and have your side conversations with the leaders to figure out how am I going to fix this?

Well, what if you could actually surface more sales-ready opportunities earlier and faster in your sales cycle so that your salespeople can focus on those? It’s tough because here it’s taking technology, but it’s also working alongside people to determine if this is a qualified opportunity or not. I think it’s just something where, from a marketing leader’s perspective, they’ve got to get their head around because then it could start solving this chronic problem that they keep seeing.

And then from sales leadership’s perspective, hey, look, what the pandemic has taught us is that salespeople need to be digital-first. They have to be digital-first now. Gone are the days where you can go have a beer and close the deal, right? Like you could have a beer; it’s just now through Zoom.

Being a digital-savvy salesperson is going to require more than just getting on a Zoom call. What it’s going to require is actually being able to augment your sales teams with these type of intelligent virtual assistants or these AI technologies to help bring them more pipeline into the fold. And if it’s a higher quality pipeline, then they’ll have a better chance of closing it. That’s really kind of the thing, I think, both from the marketing side and the sales side.

Here’s one anecdote: We survey, we run events and we survey the people who’ve attended—even I just came off of another event people attended. You know what one of always the top five topics that comes up is? It’s, how can I drive better marketing and sales alignment? Well, here’s a tool that could actually help you do that if you implement it along with your people in your process, and it’s totally feasible to do

Drew Neisser: Wow. We were going to solve all the problems. It’s funny, I suspect if I were starting a B2B marketing organization from scratch, which was well-funded and I was in Silicon Valley, I would start with an IVA and I wouldn’t even hesitate to it.

[38:08] Augmenting Sales Teams with Conversational AI Tools

#WhyAI “If your human team can't jump on it right away, there's a possibility for emotional contagion in the negative.” @JohnMcCrea @AmplifyDotAI #RTU Click To Tweet

Drew Neisser: The challenge I imagine you’re facing is the one that Brent Adamson talks about—the simple pain of change versus the gain of change, and they can’t see the gain of change relative to the pain of change. John, what slows your prospects from taking the leap and adopting your technology?

John McCrea: I’d say that, to really push conversational marketing, AI-driven conversational marketing, is a little bit of an evangelical sale at the moment. What we found though is there’s a crawl walk, run, which is any consumer brand that has any level of engagement has a really interesting pain point where a small human team is trying to keep up with an ever-growing volume of engagement, say, on Facebook and Instagram.

In the time of the pandemic, there’s been this hockey stick growth in e-commerce and in social engagement, so the volume has gone up, and some of that volume is positive stuff, some of it’s negative. If your human team can’t jump on it right away, there’s a possibility for emotional contagion in the negative, and so being able to, even without doing conversational engagement, using natural language processing to identify places where the humans ought to do hiding of a comment, liking of a comment, public reply to a comment, doing those as suggestions for the human team and then once we prove that the AI is capable, doing that in an automated way 24/7. It’s a super easy sell.

Drew Neisser: Interesting. What I’m hearing, and we had this conversation recently with another CMO, this notion of augmented intelligence, it really makes a lot of sense to me when you start that way. Where you ended, John, was replacing the human but a large part of this is augmented where we’re really helping you. And that’s where a lot of robotics is going as well. It’s not the robot doing the work, it’s the robot helping the human do the work.

John McCrea: Right. And often it’s like a small human team that’s paid 9-5 in one time zone, so they start their day out trying to dig out of a deep hole that could have been handled in an automated way overnight, or they could start with a report that says, “All that stuff? You only need to focus on this 10% of it instead of waiting through all 100% of it.

[41:11] Human AI Interactions for Social Betterment

“Our thesis is that we're living in a messaging-first world.” @JohnMcCrea @AmplifyDotAI #RTU #AI Click To Tweet

Drew Neisser: Gabi, who’s been true to her name, I guess you could say, has another question for us. She’s going to just kill me when I see her next. Do you have any experience applying your conversational AI in healthcare for assisting—I love this question—assisting in detecting depression or suicide prevention. What other applications do you see being enabled beyond marketing that’s for social betterment?

John McCrea: The specific question—I haven’t seen that, but it’s, I think, a fruitful area for folks to pursue. One of the deployments I’m most proud of is, during the pandemic, the Indian government asked us to deploy an omnichannel system to help them battle COVID-19 misinformation and to help better inform the public.

One example of that is through Google Business Messages, people searching for the nearby food or shelter of which there are like almost 12,000 of them around the country, you could search, get the result, and instead of call, it had message. Folks could tap in, find out: What are the hours of operation? Does this place have food or not? Do they still have food? Does the shelter have availability? Being able to really help with that kind of civic engagement at a millions-of-people, superhuman scale is really rewarding.

Drew Neisser: There’s a little irony here to me in that we’re using these robotic, smart computer things that are not bots to synthesize and go through and try to help people when we know that they were responsible for misinformation. It’s sort of ironic that we’ll use the same tool to help provide real information that once provided misinformation. It makes total sense to me.

This is the moment in the show where I say, let’s ask Ben Franklin. It’s a little bit of a random thing, but Ben happens to be my hero and so I get to do it. Now, I could just randomly go to a page and pull up wisdom. One of the things that he said was, “A good example is the best sermon,” and certainly you both provided great examples. But the other quote that I wanted to share that sort of made me laugh, let’s see, which is great, and it speaks to where we started, Rashmi. It’s “Let thy discontents be secrets.” You know, just keep them to yourself for the moment. Just sort of let it be; be a happy person. So anyway, thank you, Ben, for joining the show.

By the way, one of the reasons why I love bringing him up is, this was a guy who really believed in technology, and his one regret about dying—he had an incredible life—was that he wasn’t going to get to see where all the inventions were going to end up. When he saw the first hot air balloon launched in Paris, some observer said, “Well, what good is that?” and Franklin rebutted, “What good is a baby?” He had the ability to see that these technologies could grow into something, and he didn’t know where they were going, but they could be something amazing.

What’s interesting to me with today’s conversation is, I suspect we’re at the tip of the iceberg. We’re just getting started with conversation AI, and it’s really important that the people who are using it and applying it are trustworthy and that they bring data, that they bring integrity to this. At what point in time do you feel like this industry needs to be regulated? I know this is a hard question to ask, but we’re in the middle of an interesting time.

Can we trust the practitioners, whether they’re marketers or companies like yours, to not use these tools in nefarious ways? I don’t know. It’s sort of one of the criticisms that advertising has always felt—you’re selling things to people that they don’t want. That’s a tough question to wrap your mind around, but Rashmi, help me out here.

Rashmi Vittal: Well, I mean, I think part of that that we’re really trying to solve for—now look, stronger in Europe and Canada and some other regions than here in the US. I mean, California obviously has taken a big step forward in driving better consumer data privacy laws. But one of the things I’d say is that, as a marketer, if you’re running your organization with integrity, you should not be doing outreach to customers who are not interested in hearing from you in that sense.

There are different tools, again, if you use technology to aid you in this, there are different tools to keep you in check. So have they opted in? Do you have that consent? Are you managing that consent? You know, if they’ve asked to unsubscribe are you doing it in a proper way? You have the right to be forgotten, and you are being forgotten, are you truly being erased and removed?

There are steps that need to be taken that are already established today that we as operators within our company have to make sure are being done. It isn’t just on the marketing leader’s shoulders; it’s on the responsibility from the CEO across the board in the organization.

I would say, in terms of how can AI continue to grow, especially conversational AI, to grow to where it might be more manipulative is maybe where we’re heading with that? Yeah. I mean, you always have to do an ethics check in terms of, are you pushing the boundary for good? Are you pushing the boundary for good innovation? Or are you pushing the boundary for what can be interpreted as bad innovation?

So, is there a silver bullet, one answer way of determining that? No, you have to sit down and evaluate what it is you’re doing. Like some people may even think, “Well, hey, if I’m building empathy and emotional intelligence into my AI, is that now considered manipulative?” And I would just say, “Well, if you’re expecting that experience from a human-to-human interaction, why shouldn’t you expect that experience from a machine-to-human interaction?

Drew Neisser: Interesting. John, what do you think?

John McCrea: Yeah, it’s an interesting question. Obviously, you know, bad actors could wreak havoc with any new technology, this one included. I think for the most part on the business side of things, existing laws probably are good enough. I think sometimes we try to regulate a new technology too aggressively. It doesn’t allow for some of the best experimentation and blossoming. The idea that—I think California law has it—they really want to be clear that every interaction is with software, not with a human. It’s not clear to me that that ends up benefiting people.

Drew Neisser: Yeah. It’s funny—as you’re thinking about this, I’m thinking about my most recent experience with, I’m going to call it a bot. I don’t know if it was Resistbot or Vote Blue or whichever one, but I actually found it a really good experience. It was one of these where you got a text and then it said, “Hey, opt-in” and then…uh, no, I had already opted in, then it said, “Do you know who your local so-and-so is? Do you know where your voter thing is?”

It asked a series of questions and then gave me lots of good information and links that were extremely helpful. And I think we’re at the point where there are enough people out there who just want to Google something and find the answer. They just to get the answer, and virtual assistants that can get the answer the same way that Google can get you an answer is really actually a very helpful act. I think that, as long as we’re thinking about how are we helping our customers or prospects in their use of their time, then it feels like we’re in very safe territory.

John McCrea: Yeah. Our thesis is that we’re living in a messaging-first world. Our customers, our prospects are, and our engagement tech stack hasn’t kept up with that, so we’re still offering people to phone into a business, and it’s often a terrible experience. People would like immediate self-serve that’s respectful of their time and that allows them to communicate through the channels that they want to use.

[50:30] Choosing an Intelligent Automation Platform for Your Business

“You have to ask, are you meeting your customer at the moment of when they need you along the buyer's journey?” @rashmivittal10 @MyConversica #RTU #AI Click To Tweet

Drew Neisser: I want to go human-to-human for a moment. I want you both to sort of say —Rashmi, I think a lot of folks, when they think of conversational AI, they think of Drift because I think they may have even started that language. I don’t know. I’m curious, why is your conversational AI “awesomer?” You have one minute.

Rashmi Vittal: Well, first, oh boy. First off, we’ve actually been at this since 2007 is when we were founded, so I like to actually think we’ve pioneered in this category around intelligent virtual assistants, where, very honestly, conversational AI is a piece of it.

The very reason why we don’t call it a conversational AI platform and an intelligent automation platform, because you need conversational AI, you need deep learning, you need process automation, you need a lot of pieces working together for it to truly be a human-like interaction and experience. And then for it to be able to actually take the appropriate action based on that interaction. I think John was mentioning this before—there are simple chatbots that are TAP-oriented that have very little conversational AI or natural language understanding to it.

I would say that some of these website chatbots are growing in their ability to handle it, but it’s really rudimentary. What they really are, especially on website chat, which again, you know, that is, website chat, our IVAs talk along website, chat, email, and SMS. That is something where it’s really just like an intelligent router. It’s really trying to route you to the information you seek, so that’s where it began. I think this space though is going to get really crowded in a hurry. I’m getting tired of being compared to chatbots over on the contact center customer service side, because that’s where this really began, on that side of the house. Now let’s move this to the front office and really talk about how marketing sales and customer success teams can do this better.

But no, this is the other thing I say. First off, not all chatbots are created equally, one. So if you’re evaluating a chatbot that is single-channel communication channel functional, do your evaluation. Two, not all of them are AI-driven. They might slap the tag of AI on it; it may not really be AI. And three, you have to ask, are you meeting your customer at the moment of when they need you along the buyer’s journey? It should be a multi-channel experience because your buyers are interacting with your multi-channel. So that’s why an IVA is really important to be able to do that. Then in the context and the language, understanding those intents and requests are not easy.

Here’s another thing, too. People don’t know this, right? A lot of people think, “Chatbot, oh my god, I’ve got to go into some six-month, yearlong deployment cycle. I’ve got to build this thing from scratch.” One of the key differences that we provide is the fact that we’ve pre-built this all for you. So when you deploy an IVA, which you can do in just a matter of weeks, you’re getting the knowledge of the over 3,500 IVAs that came before it and so you’re getting it trained on day one, ready to go. No coding involved. You don’t have to write a single conversation. You get a very deep conversation library that comes with this IVA, and so you’re saving yourself a ton of time and a ton of resources.

We always say, “Kick your playbooks to the curb.” You don’t need to have a playbook, a manual, to tell you how you have to configure your chatbot and write your scripts. That is something that we’ve been very focused on, trying to cut down that time-to-value in our IVAs.

Drew Neisser: Alright, we’re going to kick our playbooks to the curb. Okay. John last licks, one minute, what opportunities—who would you love to have call you tomorrow as a result of watching the show? Let’s just go there. What customer, what’s the ideal customer for you that will call you?

John McCrea: Really, any high engagement consumer brand can get benefit right out the gate. Red Bull, if you’re listening, please call.

Drew Neisser: Let’s go get Red Bull. We’ll tag them in the notes and say, “You guys need this, and you need it right away.” All right. John and Rashmi, you guys are great sports. I am so grateful for you to be on the show today and be on the premier of Renegade Thinkers Live.

Show Credits

Renegade Thinkers Live is produced by Melissa Caffrey and Maya Todd. For show notes and past episodes, please visit renegade.com, home of quite possibly the savviest B2B marketing agency in New York City. I’m your host, Drew Neisser, and until next time, keep those Renegade Thinking Caps on and strong.

Subscribe

to the Latest in

Renegade Thinking