How Human-sounding Bots will shape your Sales, Marketing & Support Calls: Ricardo Garcia-Amaya




Transcript:

Thank you. I love that intro solid. Awesome. So great to be here. I actually lived about five blocks from here for 20 years before moving to San Francisco. So it's pretty nice. I want to get started asking anyone here if they can tell me take a guess as to what this is anyone. No guesses.


I'll tell you no guess again. This is the cover of a 1960s report and speech recognition from the IBM TJ Watson Center. The reason I bring it up just to bring a little bit of perspective that we've been we decide to is the engineers of Brilliant Minds have been working 16 years 60 years on getting this right to the magic moment. When you see a human having a natural conversation with a bot or with a computer now in the 60s, they had awesome strides. They didn't quite get there. So let me just take you to 1999. I'm a freshman in college and I'm doing my summer internship. I did my summer internship at IBM TJ Watson Center in Yorktown Heights.


And I remember my first day. I walked in into the cafeteria and I just remember seeing a sea of just white hair as far as the eye could see and I mean I was 20, so I thought that probably most likely I was the youngest person under the age of what seemed to be like 75 that's Sure, that's not true. Love sure but that's how we felt. That was the first thing that I got out of that internship the second part. I got to work under the team of Dr. Nahum. He is considered the father of speech recognition. He's been working on it since the 80s until last year.


He was the global CTO of IBM for speech recognition and I was under one of his team's and text-to-speech and later. I learned in 2006 that IBM pulled the plug in that because the technology wasn't that made amazing strides right IBM's incredible leader, but they couldn't get to the point of Technology wasn't there to get him to the point where you have that natural conversation between a human and a robot. So that takes me to the next date important date 2018 probably most of you remember this this is Google's I/O last year and they feature duplex a bot is calling and setting up a hair appointment and the bod sounds perfectly human.


It breeds like a human. It talks like a human and take pauses like a human. The other person on the line has no idea that it's a human. So, is that good? And in many ways this this year, I mean 2018 last year Google was firing the first shot to the race of building voice pots or commercial voice pots voice, but for sales service and marketing calls specifically among the many other thousands of applications, right? So that's a big year and that takes us to today which I will be talking about how companies can leverage voice Bots for sales marketing calls to develop 222 build that Roi So, my name is Ricardo Garcia Maya. I am the founder of VO IQ.


We are platform that helps businesses build voice. Bots deploy spots for sales marketing and service calls campaigns. I want to talk about three things today. The first one I want to go under the hood. I want just to get a little bit of appreciation of what makes a voice bought a successful product to make very successful calls service marketing and sales calls. Then I want to be able to share a handful of use cases a house sales and marketing leaders are using the many amazing features of voice Bots to generate Roi and hopefully you guys can get inspired and go back to the sales and marketing teams in your companies and share those with them and lastly I want to share a couple of two Colombian so we say a couple and we've literally mean to but that's not true in the English language. I've to use cases that I'll share with of two specific companies that we work with in the way that they generate Roi based on leveraging this voice pots. And again, hopefully you guys walk out with a couple of awesome ideas on how to build that are on your own companies.


Usually there are a couple people in the audience that are saying okay robocalls is the Here Comes voice Bots kind of Robo calls and steroids, but there's nothing farther from the truth and will be able to go through that through the presentation. So let's go straight take a look under the hood. And with this is a chatbot conference. We know we are familiar with the most common form of a conversational chatbot. It's on the lower right-hand corner you enter an input of text and it goes through natural language understanding and you get an output of text right would hopefully the right answer. So thanks to that technology. We have the conversation. Chatbot now, what do we need to get it to a voice bot or voice assistant on the consumer end.


Well, we need a layer speech recognition as I mentioned took 60 years to get here. So thanks to that. Thanks to the fact that we have a I now these two technologies actually start working together and you can actually have what most of you know as the Google assistant or Alexa in living room in your kitchen. So, how do we make a voice Bob make a call? We need one more layer. And that is the programmable Telecom layer. Now. This wasn't always around either if you go 2010 and back only big Telco only big telecommunication companies could build applications and not until the past 10-15 years. You have companies like Amazon next Mall opening it up.


And now any engineer from their bedroom can write through lines of code, amazing enterprise-level telecommunication application so that last layer allows us to have a voice but in this case, let's just color a voice assistant on the consumer in that can actually make calls and now we're talking about Google I/O this year. What is it a week ago two weeks ago where you can have your voice bot make calls on your behalf to set up restaurant reservations? Right? So now we know what a consumer voice about can do with all these Technologies and for the consumer it works for you, right? So what if you are a business and Enterprise and you want to leverage voice Bots for sales marketing and service calls? Well to give you an idea what I mean, if you are a business and you want some solution for chatbot automation, you can find companies like drift in a hundred others probably now a thousand of those are companies that focus on just chat pod automation. If you're looking for email automation same thing right now, I mean you can go for a company like Outreach or you can or the other hundreds of other very successful amazing email automation tools when it comes to voice body call automation not until the beginning of this year.


You had any alternative where the first company that focused on Marketing sales and service calls through voice pots and there will be tens if not hundreds of this companies in this space coming up over the next year five years etcetera. So what does that so the sales and marketing voice broad platform leverages these core infrastructure at this court Technologies and a platform anybody who has done carried out or LED any teams? Sales people on the phone. Has anybody had that experience here raise your hand? Okay, how was it? Awesome? It's painful. Yes usually is extremely painful and it's extremely difficult in order to get a sales or marketing call campaign.


You need to optimize a series of criteria to get it right? It's extremely difficult to with this with unpredictable humans and we set out to build a platform like a spaceship that basically tackles all of these with the predictability of voice, but we could do it. However, we had to start from some of the things that sound extremely simple but they're extremely complex problems. So I'm going to go over a couple of things that have voice but platform just takes care of the you don't have to think about a voice bot does not know the difference between a ringing tone or hitting a voicemail when you're calling somebody or even worse an IV an IV are so now I've ER is the hi you've reached the offices of dr. Jones, press 1 for this press 2 for that. So a platform actually allows you to navigate through all of these so you don't have to start from scratch with your voice spot. This isn't common one. If you're thinking of if you're new to voice Buds and you want to launch Voice with that make calls you can launch this say I don't know five thousand calls and then you quickly realize that of 25% of them. In the business in the Pacific and PST time zone and you just woke up people at 5 a.m.

And they're not going to be really happy with your brand. Right. So these are some of the things that a platform takes care of automatically this is another case updating lead information.


So when you hit somebody and they don't have it's wrong number a voice but I just want to continue having a conversation and they don't know what happened. So a platform will say, okay this first one is not who we thought we need to not call her and make sure we flag it as a bad data on it handling unexpected questions as well voice, but just want to have a conversation.


They they're not really ready to handle unexpected questions. This is not so unexpected. This is typical like someone might say hi, is this Linda and they say yes, and maybe they say it was this right? So the answer of Barry's not by the company, but by the many type of campaigns anybody was done sales and marketing campaigns, you know that you can look just like anybody who's launch email campaigns, right? You don't launch one email campaign per company. You have multiple products multiple geographies multiple teams you have Thousands of different campaigns for voice blessed the same thing and you need to train them per campaign. So that's something that a platform addresses for you and this is an interesting one. This is sorry. I can't talk right now again, they sound super simple. If you have humans. You don't have to think about it though it voice has she to it. These are actually pretty big problems.


Sorry, I can't talk right now. It doesn't mean I'm not interested or I am interested in means not now call me later. So a platform will handle understand that and understands and we'll call that person later. So no human interaction. It will just do it for you in this is one of my favorite ones that are probably one of the scariest things to here for somebody who leads teams of making calls on your behalf, or if you have teams in the Philippines or whatever. They are. Right? One of the scariest things is to outsource this job or have somebody walk into the office and say great remember that list of 10,000 or whatever five thousand calls. They're called right and it's a manager you are you're more interested in.


Going okay, but what happened in those calls, right? Who do we call? What did they say who confirm who answered? What attempt at what time what objections were questions, right? That's the whole point of that campaign is like sending one email campaign and not learning anything from it, right? You learn over the course of sending multiple campaigns. So that's what a platform helps you solve.


So hopefully that gives you a little bit of appreciation of what goes into a successful voice spot launching Marketing sales and service campaigns whole campaigns rather. So the second thing I want to share some use cases and how some sales and marketing leaders are leveraging awesome features in voice box. So the first feature is human sounding so the reason I say that they right now there are tens potentially going to get to hundreds of companies that just focus on one thing and that is creating voices for voice pots and when you buy these Voices they are they're human. They sound like a human but they're not perfect. They don't sound perfectly human you can get them there, but you need to train it, right? So taking it back to the Google I/O last year in order to get that voice bot to sound perfectly human. They spent four months with several Engineers.


Just making sure that that voice but could do one thing and one thing only and that is to make to call and set up a hair appointment. Right? So now you can do that. It won't take you, you know, several Engineers from Google and in four months, but you can do it in hours or days and over the next year. The very likely will come sounding perfectly human to start breathing like a human taking pauses etcetera, but you will always have to train it according to the campaign that you lunching second one really interesting thing.


We found no correlation between someone hanging up the phone because it's a voice pot. The reason why people hang up the phone is highly correlated with Fact that you don't have the right ICP. They're the ideal customer profile. So that is the correlation not the fact that it's a voice but and I tell you I'll tell you why that is you the voice thread is able to have perfect.

Well, not perfect. Yeah. It's a conversational over the phone but not only that but it's much richer because it's tied in from your CRM right very different from I just had this conversation with somebody outside very different from an inbound bot when some when you need to guess. What is it that people need and outbound sales and marketing call? You are just pre-qualifying somebody and you're the one asking the questions so it's my way more narrow. So there's less room for error when using a voice but in an album fashion, so another cool. This is a good example. Let's say Karen fill.


This is a real type of customer example car insurance has a page of forum marketing Forum where they get somewhere around half a million to a million people every month asking for car insurance. They need to qualify them. They don't want to qualify them with a hundred questions. They want to ask them three and then they have a voice but that calls them right after they fill out the form to ask him to additional questions in this case. Hi Karen, you know, we just want to make sure that your Toyota Camry how many miles is a have? Right? And so that's further pre-qualifying somebody without using a human now another awesome feature in the same example is once you says yeah, I know it doesn't have over a hundred fifty thousand. It has 90,000 miles the voice that says great. Let me introduce you to do you have time and I can introduce somebody just to take care of this car insurance, right? So you can transfer to a human or a sales agent. There's another voice cloning one of the coolest things that are happening right now.


This is an example of a use case real estate agents. They spend 80% of the time outside of the office. They hate wasting their time, but people not showing up to the open houses Etc so they can leverage a voice but with their own voice to call their prospects and remind them to show up or Confirm that they show up and they get a text knowing whether they will show up or not. So this is one of the things that they don't like doing but it would love to be able to solve that's one of the areas.


This is another way that you can use it with a brand you can have the gecko from GEICO will be able to call you and said, you know, hi Josh congratulations mate, you've been approved, right? So this is yet another huge example of what you'll probably be seeing soon enough. You can have the most interesting man in the world selling those say keys to you know, restaurants and bars around the world. So those are the type of things some of the easier things is well. He's or everyday thing schedule and rescheduling anybody whose does sales demos, you know, that industry standard is 50% of people show up.


You can get it to 75% If you call them if you use an additional Channel other than email that's whether you use humans or voice pots the cool Vegas would voice about you don't actually do anything. It just happens by itself The Voice let's call people that did that that are about to show to a demo or the people that did not show up to them. He calls them to help reschedule that demo without sales people having to do that job trigger workflow. So voice but can trigger automatically themselves. And so let's say you want to send a campaign of 10,000 emails. You want to have a voice but called only the people that open the email and click on a link you can have a voice but do that right to further maybe introduce that person to somebody else at the moment that they have the highest level of attention. You can train it to handle objections, you know, I can afford it right now or like, oh, we already your use your competitor. You can trade The Voice spot to do that analytics reporting huge every voice but is a hundred percent predictable.


So we you're able to capture everything in real time, you know, exactly what's happening. You have live QA so they QA will tell the lead the sales or marketing lead exactly where it is. What type of objections are coming up etcetera to how to make it stronger the campaign in the future and hundreds of other tens of other amazing features. Here's one of the coolest one in the future live online demos. So if you sell a 10,000 dollar product, you're going to have a human doing that live demo, but if you sell a product for a hundred two hundred dollars now, you can actually afford having a voice bot doing a live online demo for you and that's coming up over the next three four years lastly case study two case studies. I'm going to run through this. How am I doing in time? Okay. Okay good.


So these are the two use cases. I'm sorry case studies the way the two companies leverage voice Bots to generate Roi the first one I'll just say is one of the one of the largest car hailing company. They also do food delivery and the way that they went to Market in the United States. They needed to have two teams one creating a list of half a million restaurants the other one calling them and they needed to prequalify them at three questions one is first make sure that the restaurant exists second whether they offer delivery and third whether they have multiple locations, these were the three questions they need to do they actually did it in the US. Yes. It was extremely expensive but more importantly took time for a company Venture back to that. They're more interested in time rather than money now they needed to expand in Latin America. In the decided to go with the same Playbook now the smaller number of restaurants, but the same type of calls right now one of the issues and they have humans making this cause that's how they started and one of the issues that they're calling, you know auto body shops instead of restaurants.


They're calling restaurants have been closed for the past 10 years, right? So they have they were dealing with 30% bad data and it was really bad. They were wasting their time. So we start working on a pilot campaigning three countries. We chose a Rodrigo a voice of a voice but that is not doesn't have a Mexican accent doesn't have a Spanish accent. It has a neutral an American accent and you're going to see a ton of voices that you know to fit your specific. I don't know if we have a voice. I don't think we have sound on this one. But yeah, believe me sounds great.


All right, so then we build a campaign with those questions that they needed to ask to prequalify them the reports in real time. The analytics show them within two weeks exactly what's happening. I mean its real time. So it's not like you need to wait for it. And at the end they save about 25,000 per week for by not having humans doing these pre-qualification questions and also they got to sign up 15% more the reason why is you can't call a restaurant during lunch time? So don't ever try that. I mean to ask like three questions, right? If you call a restaurant make sure you order that's the only reason they care right second when you call and you have six questions to get to before selling somebody something they're going to hang up on you. But if you have a voice pot pre-qualifying before you have the human actually making the real cell, then you can actually sign up more people and of course everything is a hundred percent documented in their CRM, which is something that salespeople are not known. For love Which is documentation because it's hard and you know, you're not making the sale.


So this is another is my second case study. This is a big Health Care call. That's what I'm going to call it. One of the largest websites that provides advice of what type of personal insurance you should select until very recently they bought an insurance company. So they started selling Insurance themselves. They ran into an issue, which is that people were not renewing their insurance when they needed to so they developed email reminders Etc. that there wasn't really quite working and they decided you know, what let's test is let's just the internal team jumped on the phones and start calling people as they were they're starting to expire and suddenly they saw the needle move and they said, huh? Okay, does that mean they were going to have to build a team or what are we going to do? So luckily they were using hops pod as their CRM and they started looking and they found us on the integration page after 2 minutes. They were integrated with us. They already had a workflow because they had these emails ready.


Set up remind people as soon as they get expired and all they did is create a voice pot inside Hub Spot that followed and they mapped it in a way that this people will be called to confirm whether going to renew or remind them and at the end well, they didn't have to build a team of humans to do this. So it costs them a ton less than actually having humans doing it and also the increase the customer retention rate because when people are making calls that will not make a hundred percent of the calls that they need to do and they won't do them at the right time with the right Cadence a voice but we'll do that every single time. So that's how you can actually get better results. There's a bunch of different kinds case studies examples of how they're being used in other Industries. This is idc's as 30% of Enterprises will use voice bust up our Marketing sales and service by 2022. So I hope that today that is this inspire you to take it back to sales marketing service and start using voice Buds and hopefully be one of these 30% of Enterprises will be using it over the next year's. Thank you. This is my email.

Anything should be an email and I'll be here to talk. Thank you very much.

© 2017 by Chatbots Conferene TM.