We’ve been hearing a lot about Recruiting Automation, and Chatbots, AI & ML, but where are we with this tech in HR and Recruiting?

In this episode of DriveThruHR, host Michael Vandervoort interviews Jonathan Duarte about the role of chatbots in recruitment.

Jonathan shares his extensive experience in the technical recruiting space and provides an in-depth explanation of chatbots, their functionalities, and benefits for recruiters. The discussion includes historical examples like the U.S. Army’s Sergeant Star and Intel’s implementation, showcasing the efficiency of chatbots in handling large volumes of inquiries and pre-screening candidates. The conversation also touches upon the cost, integration with existing systems, and the personalized nature of chatbot interactions, emphasizing their growing relevance in modern recruitment processes.

Michael VanDervort, Co-Host of the DriveThruHR podcastDriveThruHR podcast and Jonathan Duarte, CEO of GoHire, chat about chatbots, Apply by Text, and detoxing from another HR Tech in Las Vegas.

Recruiting Chatbots Transcript:

# Michael VanDervort – Jonathan Duarte – DriveThruHR podcast

[00:00:00] **Michael VanDervort:** Good afternoon good afternoon for me and good morning for our guest. This is Michael Vandervoort and we’re doing another random episode of DriveThruHR. Today the guest is going to be a guy I’ve known for a couple of years. His name is Jonathan Duarte and we’re going to talk about chatbots.

[00:00:18] The show is everything you ever wanted to know about chatbots. But we’re afraid to ask, and Jonathan was kind enough to give me a demo for about an hour and a half earlier this week. We won’t do that, but it was for some pretty interesting stuff. Hey Jonathan, why don’t, before we get started, introduce yourself, tell people who you are and what you’ve done in life and that kind of thing, and then we’ll dive into the topic, okay?

[00:00:40] **Jonathan Duarte:** Yeah, sure. Thanks very much, Michael, and great to be on the show. I guess there is an old saying, long time listener first time caller, right? I started in the TA or recruitment marketing slash internet recruiting space way back in the 76 K modem upload speed time of [00:01:00] 1996.

[00:01:01] Wow. And back in those days, we were we were literally just moving people off of the newspaper onto the web, but there was a lot of I was working with a lot of technical recruiting companies at the time technical staffing. And it was all during the First. com boom, and most of the teams at that point didn’t really even have we were still fax based for the most part and so I I grew up in this era of and I was only like three years out of school when I started my first company, GoJobs, and it’s been a 21 year ride in learning deeper and deeper about what TA is all about, what the recruiting positions about, what recruiting technology needs to do to make the the whole mission that much easier.

[00:01:45] And from day one, my first thought process in this whole thing, being young and naive and having never started a business or a software company or anything before was, there’s this. Big staffing industry out there, [00:02:00] right? We can change the world and get rid of recruiters if we can just get job seekers connected to employers faster.

[00:02:08] The big picture always out there was, Hey, and then you realize that’s not really possible. But what you can do is get rid of a lot of what we call the friction points. And so that’s what I’ve been doing over that over the last 21 years is building everything from job boards to Job distribution systems to job posting systems.

[00:02:27] I was in the background screening business for a little while. And now I got into built a new company called GoHire and we’re building chatbots and we’ll talk more about that in a little bit.

[00:02:38] **Michael VanDervort:** Yeah. So great. And you live on the West coast. You’re in San Francisco, right?

[00:02:42] **Jonathan Duarte:** Yes.

[00:02:43] In the depths of Silicon Valley.

[00:02:46] **Michael VanDervort:** Yeah. Cool. So lots of. Lots of different news. That’s a whole nother topic and a whole nother show, but lots of different news coming out of there the last few weeks and always the epicenter of stuff it’s funny you mentioned fax machines.

[00:02:59] My very first [00:03:00] HR job, I went into HR in 1986. And I, after I was on the job about three months and somebody from our corporate office called and said, Hey we’re sending you a fax machine and I was like, Cool. What’s a fax machine? I knew what it was, but I had literally never used one.

[00:03:17] And we they delivered it the next day. And it was one of those giant rolls of toilet paper kind of thing. Faxes where you have to yeah.

[00:03:24] **Jonathan Duarte:** Yeah. So we set it up like a What was that? Thermal paper, right?

[00:03:27] **Michael VanDervort:** Yeah, thermal paper. Yeah, thermal paper.

[00:03:29] **Jonathan Duarte:** And you had to put the resumes right away into the file cabinet, which is, really useful, otherwise they would they would bleed out and you wouldn’t be able to see them in a week.

[00:03:37] **Michael VanDervort:** And, it didn’t take me long to figure out, oh, there’s fax machines, and then, oh, there, oh, yeah, there’s computers, and so there’s, I probably should have a desktop computer, and, anyway, I advanced technology. Pretty quickly for an HR person who had never seen a fax machine before.

[00:03:53] But I told you when we did our, when you did the demo of your product for me the other day, I know what a chat bot is, but, [00:04:00] and it seems like people are always talking about it, but it’s I don’t really know that I really don’t know that I. I think I know, right? But I don’t really know what’s under the hood.

[00:04:10] So I think the first question, about, I think the first question that you want, maybe you want to share our listeners is the same, that what is a chatbot? And, why would you start a company based around chatbots?

[00:04:22] **Jonathan Duarte:** Okay. Yeah. So here’s the easiest way to imagine a chatbot is Most of us are using SMS and even internationally SMS or text messaging.

[00:04:33] Imagine I’ve got Michael’s phone number on my phone, and I know you’re a recruiter, and I know you’ve got a job opening. I’m just going to text, Hey, Michael, do you have this job opening? For, let’s just say it’s a retail say a front desk at a hotel. I could type you a quick message, which is, use my thumbs instead of a desktop use your mobile phone that everyone’s got for the majority.

[00:04:57] And you could then [00:05:00] get that information and respond back. You would say, Oh, Hey, John, we’ve got this one within five miles of your house. This one position. Here’s a link to see the job description if you wanted to. If you think you’re qualified, you can click this link and you can start answering some questions and I’ll see if you’re qualified.

[00:05:16] And then you can have a conversation and forget what the actual conversation is. But you could respond back and forth, one person to one person. The chatbot, which is really just a computer on the back end, could mimic Michael in this situation, and say, me, John, I’m responding, I’m interested in this job, and you Michael being the chatbot just uses what we call natural language processing or artificial intelligence.

[00:05:47] I don’t want to go deep into that stuff, but It just responds back and ask questions that are of nature. They could be structured. They could ask you pre qualifying questions. What’s your name, your first name, last name, etc. And then [00:06:00] you store all that data. In the database, very much like you would with your applicant tracking system.

[00:06:05] The difference really between this kind of tool of having, doing it on your mobile phone versus just filling in a long form for an apply, even if you have a mobile optimized website. And you get someone from Craigslist to enter the landing page for your applicant tracking system to start entering their information.

[00:06:26] The problem is it’s a one way push of data, which means you have to fill out, maybe there’s 20 questions. First name, last name, address. Give me all of your, all of your personal information, all of your resume, try to do all this stuff in 20 questions and then click send. Chatbots are different because they’re conversations.

[00:06:44] You can start Asking more and more questions and then depending on the job, depending on the company, you can respond differently to the candidate. So it’s not one size application form for every single user. It doesn’t have to be that way. It can be [00:07:00] tailored to that type of position as well. Does that kind of give you a better kind of handhold SEO standpoint?

[00:07:05] Yeah.

[00:07:07] **Michael VanDervort:** Yeah, I think it does. And I think what’s interesting, I told you this again when we talked the other day, it seems to me that in, and it’s natural that in the, at least in the HR space, technology is quite often Led by attracting talent, that kind of thing, or compliance, but the talent attraction, people are

[00:07:23] willing to put money into products, because they always need talent, right? There are other things that It’s always a moving, it’s a moving target. Yeah, so I, I, I don’t work in the recruiting space, and we’re not gonna, I’m not ready to jump down that rabbit hole just yet.

[00:07:38] But as I looked at what chatbots are doing, like what you were talking about, and some of the other examples that I’ve seen, it’s I can imagine a lot of other ways. That they could be used, and that’s what got me interested. You mentioned that there were a couple really, really early on, early adopter examples, and one was the surprisingly maybe not surprisingly, because it fits the talent theory, was the U.

[00:07:59] S. [00:08:00] Army with a chatbot called Sergeant Star, is that right? Yeah. And then another one, I think the other company was Intel. Maybe I, if I remember correctly, you want to tell us about those examples that are, 10 years old. If I remember what you shared with me the other day. Yeah,

[00:08:13] **Jonathan Duarte:** exactly.

[00:08:14] Exactly. The U S military, when, obviously the army is employing a lot of people and they have, I wouldn’t call it attrition. I’m not sure, military recruiters probably knocked me over the head on this, but it’s not a lifetime job for the majority of people. So they come in and maybe they’re in the service for three to five years of, excuse my ignorance.

[00:08:34] But it’s a constant feeding. They have to keep getting more recruits into the system. And the numbers are gigantic. They’re looking at, 40, 000 recruits every year coming in. So even your largest retailers in the U S could find, there’s a lot of similarities between the recruiting structures that need to happen.

[00:08:54] So when you’re done, when you’re working on that large of volume, there’s a lot of hand holding, too because military is a [00:09:00] little bit different than, say, working a job at JCPenney or something. You’re, and it’s a little bit older. The benefits are a little bit different. You’re going to be out of town.

[00:09:08] Those questions would normally be handled by a recruiter. We would see that in a first round interview. And what the U. S. Army did about 10 years ago now they built what they call Sergeant Star. And Sergeant Star, if you go to GoArmy. com, you’ll see him up in the right hand corner. And if you’ve been in the Army over the last 10 years, you’ve probably gotten to know him very well.

[00:09:29] And what he does is he answers questions, free form text questions, and answers them with a short snippet response probably about, an hour long. It’s about 150 characters or something about a text size, and it sits on the website, and then as you ask questions, it can send you to the right area of the website to get more details.

[00:09:50] And so it’s really a search engine that’s personable, and it keeps track of all your questions, and it gets you to the right place, and then you can actually do other [00:10:00] things where you could probably apply through it or it could send you to the right pages. So it’s really more of a an index or a searchable way to find data, especially if you’re in find information, especially if you’re in a big retail organization or you’re in a company that you really want to sell your like employer brand or why work for XYZ Corporation.

[00:10:20] It’s just a, and it’s just personable. It just feels like you ask a question, they answer it. The other thing about chatbots, and we’ll talk about the Intel example in a second, is that chatbots can be seem, they seem almost anonymous in a, in essence. We’ve all gone to I think that the other kind of chatbot I’ll describe, the Sargent Star was fully automated.

[00:10:42] And some of the stats that came out of it, just to complete the conversation, were in 10 years, it’s answered, Sgt. Starr’s answered roughly 16 million questions, and annually does the work of approximately 55 recruiters. And those are stats out of the U. S. Army. As a big organization, you could just start counting [00:11:00] the dollar savings.

[00:11:01] automate frequently asked questions that candidates might ask. So that’s one great example. The other example is employee self service. And those questions again the biggest impact is larger HRIS or larger companies. Is if you’ve got candidates even in a dispersed workforce if you could answer questions 24 7 have your normal HR Self service and or call center but if those questions could be answered via the intranet or via a mobile phone or an SMS or some sort of chat system, that would allow candidates to ask questions or employees to ask questions 24 hours, 7 days a week and 365.

[00:11:44] So you can expand your offering to your candidates or to your employees and give them the information they want when they want it without them having to have to learn. The entire intranet to find out where the information is. The ca the case with [00:12:00] Intel is they built a chatbot I believe 10 years ago, and they created, I believe, about a thousand questions that they wanted this chatbot to answer.

[00:12:10] And again, it’s fully automated on the intranet, and it answered all Employee self service. Now, I don’t know tons of details, but I have talked to Intel employees who know the system and who have actively used it over 10 years, and it works. It’s a first line of defense so it’s like your level one tech support, if you will and when you go to I think another way to envision this is if you go to, say, Xfinity or Comcast or your cable provider Oftentimes, there’s a little chat window that you can talk to customer support.

[00:12:43] That’s the, that’s our, from a website, I think that’s probably the best use case that most people are used to. But instead of talking to a human in recruiting and what we can do, I mean you can still build, we have as well, chatbots where you can, have live support 24 7 or just on a [00:13:00] certainly hourly basis, but you can have the frequently asked questions answered by the chatbot that way you reduce the cost internally of the recruiters, but candidates and employees can get the information they’re looking for very quickly.

[00:13:13] Does that kind of help? Yeah, it does.

[00:13:15] **Michael VanDervort:** Yeah, I’m trying to think. I think you mentioned another thing that there was a, was it Maya? Is that the, where you said, I think we talked about that. There was a big blast of press. And I don’t want to like, I’m not trying to call anyone out, but you had some interesting comments on that.

[00:13:30] I thought.

[00:13:31] **Jonathan Duarte:** In October there was a lot of, October 2016 there were some press releases that came out that I think started this scared tactic. We can call it a tactic, but it resulted in a lot of people freaking out about what is artificial intelligence and are the robots going to take over in recruiting.

[00:13:52] And, everyone’s got their own opinion of that. My my thought is that what happened was it was great [00:14:00] that a lot of people got interested in the topic because of this kind of almost threatening But the reality is, chatbots are not going to take over 80 percent of the recruiting roles, which is the sum of it.

[00:14:12] What chatbots can do in a recruiting still keeping this high level but what chatbots can do in a recruiting setting is automate a lot of those individual tasks, such as the recruiting coordinators spend a significant amount of time doing things that could be automated. Such as scheduling interviews with candidates.

[00:14:35] If a candidate had access via text messaging to the calendar and could respond back quickly, and a lot of these things, what’s interesting about messaging as well, is that it’s almost instantaneous. You get your little notification on the phone. Hey, I just got a new text message that came in or a Facebook Messenger message.

[00:14:55] So I click on it and I respond back really quickly. We don’t do that with email. And [00:15:00] so as we leverage messaging, we can shorten the time period between connecting the dots and getting people scheduled, asking simple questions getting them to fill out their I 9 or E Verify information if they haven’t done it.

[00:15:15] Those little tasks that we have to do in recruiting with every single candidate that we’re going to screen, A lot of those things can be automated by the chatbot. Is that going to remove recruiting coordinators jobs entirely? No, there’s still a lot of things they do. But any of the the repetitive manual tasks a lot of them can be and you won’t see it happening right away for everyone, but over time, and there are companies already implementing this stuff, and it’s working and they are starting to save cost savings by not, Engaging the candidate using messaging because it’s instantaneous versus email where we already know when anyone on the call who’s recruiting millennials, anyone below the age 34 already knows this, is trying to [00:16:00] get a millennial to respond or anyone of that age group to respond to email It is already fighting it.

[00:16:06] We hear about it from on LinkedIn. No one responds to emails anymore, and that’s more than the professional side, but in I think the response rate in emails is less than 20 percent now. In email, it’s probably less than 15 or less than that, depending on if you’re doing bulk, but the idea with messaging is that you get a 30 to 60 percent response rate, and it happens quickly versus, Email, you get a very low response rate, and it takes a couple days, possibly, if candidates even use email at all.

[00:16:34] I think we’re seeing a convergence, and let me, pitch that here in a second, is that messaging, we’re seeing a convergence that’s causing this AI chatbot conversation to happen. The first convergence is, we know that email usage is dropping off, and the response rates, the spam rates, all this stuff is causing it.

[00:16:55] Make it harder for people to recruit using email. In certain segments, not all, of course. [00:17:00] And then secondarily, we know that messaging has become more popular. In 2015, messaging apps became the number one application used on mobile phones versus even social network. Messaging is being used more than any other app on a phone.

[00:17:16] And with this down use of Email and the uptick of messaging and then the automation of messaging. We have this clear path that if we can connect messaging, the value of propositions of messaging and the value propositions of automating workflows. There’s a huge future of things that are possible, and I think that’s where we are in this whole conversation, is we’re just tipping our toes in the water there are companies, early adopters, engaging in it and that’s where we are as a company, is working out with early adopters and engaging Thank you.

[00:17:53] Companies that could have the immediate impacts of it. So it’s not fully scalable yet.

[00:17:57] **Michael VanDervort:** You you were kind enough to show me under the [00:18:00] hood a little bit the other day, how, how a desktop looks and stuff. And I thought it was, I thought it was fascinating. The so to go back I worked at, a large retailer and some of the things you cited, like we, we would, and these numbers are, somewhat aged and probably a little bit hypothetical, but we would receive over a million resumes a year.

[00:18:21] We had roughly 60. Percent turnover in the retail industry, which was considered good when I first heard that when I moved from manufacturing into retail, my mind almost exploded when they said we had 60% turnover because I came from a place that had under three, right? And then I learned that some retail and some restaurant companies literally actually have higher than a hundred percent turnover in a year.

[00:18:41] Yeah, it’s all relative, but it’s mind boggling. And the other thing you mentioned I think is that or did that you mentioned the other day, but I don’t know if you mentioned it today. So there’s tons of paper coming through lots of applicants, many of which, as you’ve pointed out, don’t even necessarily have email addresses at home.

[00:18:57] Let alone at work, right? Like [00:19:00] the Publix workforce was nearly a couple hundred thousand people and probably 150, 000 of them didn’t have a work email address. So they used other things. They used there to talk and communicate their cell phone and that convergence was almost driven by those kind of business circumstances in some ways, I think.

[00:19:18] So can you take, anyway, the reason I bring all that up is, I think it’s really cool how, you mentioned the convergence, but I also want you to talk a little bit about behind the scenes from the recruiter who’s working, the real person that’s working behind the chatbot and how that sort of transitions over into human hands at a certain point.

[00:19:36] Yeah.

[00:19:37] **Jonathan Duarte:** All right. And I’ll talk real quick about the the convergence of the moving over to messaging. Here’s what if you’re listening and you’re trying to figure out, okay, does chatbots have anything to do with my recruiting? Or why would I know if I have a problem that it, chatbots might be able to solve?

[00:19:54] And one way to think about it is this, is if you’ve got candidates coming in, From, say you put an ad [00:20:00] on Indeed, Craigslist, Monster, it doesn’t matter where you acquire a candidate. They go to your applicant tracking system. A lot of companies don’t have, but I know a lot are getting there, is better stats to say, what’s your conversion rate from someone landing on a job detail page to actually completing the application process?

[00:20:19] Candidate comes in from Indeed. How many of those candidates actually complete the application? And what I know, and because I’ve run job boards and what a lot of companies who have the data will show you, is that your average conversion rate of a candidate is somewhere between 8 to 12 percent, maybe higher, maybe lower, depending on your industry, etc.

[00:20:41] Let’s just use an even number. Say it’s 10 percent. If you have to buy a hundred candidates, or a thousand candidates to get to your website, and you convert ten of them to get a hundred applicants Or, they just show up. There’s lots of different ways, but you’re getting a 10 percent conversion rate and If you [00:21:00] have a workforce that’s under 34 right now, there’s statistics that say, and Mike, you alluded to this, was 25 percent of the U.

[00:21:07] S. workforce right now either does not have a personal email address or does not check it on a regular basis, and that may sound really strange to a recruiter, but the reality is your barista at Starbucks, Starbucks does not have a Starbucks email address. The server at Cheesecake Factory does not have a Cheesecake Factory email address.

[00:21:31] The hourly workforce in the United States does not have email addresses from a corporate side. And then secondarily an example that I have, and you listeners might have this as well, I have a 12 year old daughter, my wife emailed her something a couple weeks ago, and my daughter’s like Mom, I only have an email just to sign up.

[00:21:51] I don’t use it. So Email for that generation, and I’m 48, that generation is just using it because to get access to the rest [00:22:00] of Google’s features, or Apple’s features, you have to have an email address. But other than that, do they open Gmail? Never. They only use it as a tool to get through the gate.

[00:22:10] Here is the question, the answer to your question was, how do you know if you have this problem? If you’ve got candidates that you’re looking at in the hourly workforce, and you’re not getting enough candidates to come through the pipeline, so if you’re high volume, or you find yourselves playing phone tag with candidates because they’re not responding to emails, You probably, and I would ask you to test this, and you could probably do it with your own cell phone just to test this out, is if someone’s giving you their, their their SMS number or their mobile phone number, find out how many of those people would allow you to, if you sent them a message about a job ad, would they respond to you?

[00:22:49] And if they would, which they would. My guess is they probably will because it’s easier. If they do, then shouldn’t you scale that up across your entire company? And that’s [00:23:00] where we are as a, as an industry right now is trying to figure out. How do we give recruiters the ability to use SMS, not on their mobile cell phones, but to leverage the technology, that, and Messenger, and a bunch of other chat systems, so that for the hourly workforce, the 40 million people that are got, say, a 50 percent churn every year, how do those people communicate?

[00:23:21] They don’t use email, so how are you going to communicate with them? And SMS in the U. S. is one of the biggest tools to do that. That’s what we’re looking at, as in, how do you automate those conversations so you can scale it up? How can a recruiter have 10, 000 contacts that they use SMS to contact?

[00:23:36] That’s really difficult. If you Use a software tool on the back to answer pre screening questions, answer questions about the company, or the job opening, or what’s available right now, or are you 21, you need to be 21 to work in this job. Those are things that we can scale up. We call them just, your average first round screening.

[00:23:56] There’s another thing that the chatbot can do is, [00:24:00] imagine you place an ad from Craigslist and it goes, the response is not to have the candidate or call to action, isn’t to have the candidate go to your applicant tracking system with the 20 questions because you know you have low conversion ratios.

[00:24:14] If you’re in that high volume area, if you’re looking for candidates that need, that If you are probably using SMS, give them an SMS phone number to apply to, and when they apply to that number, get all their contact information, and then we have a dashboard from our site that allows the recruiters to actually communicate with the candidates that way.

[00:24:33] But if you tested that out with a couple of ads on Craigslist and say, hey, text me, I know they do this in the trucking industry quite a bit see if it works. If it works for you, then you might have a solution to a problem, which is an acquisition problem. And you might need a new tool, which is text messaging, in order to get over that acquisition problem.

[00:24:52] **Michael VanDervort:** The it’s funny, I haven’t done recruiting, active recruiting, in about 10 years, and I never was a big time, I was a generalist doing [00:25:00] recruiting, but Craigslist was one of the more weird, but more effective recruiting spots for certain industry, or certain niche jobs, like you mentioned, like trucking, or it doesn’t fit everything.

[00:25:11] But I, when I said I was going to try it at one of my employers, they looked at me like I had lost my mind, but it actually produced some good leads. I was like, you got to fish anywhere. You got to go where the fish are. So anyway so tell us a bit. So the, so just to go to reframe the other question I asked a second ago was like, so on the front end, the chatbot does a lot of the prescreening.

[00:25:31] Just using technology to people. It’s no different than the old kiosk. You answer questions at a kiosk. Have you ever used that? And there’s knockout questions, right? It’s just a different way and it looks more It’s a it feels perhaps a little more personal because even though they’re technically talking to a machine They’re getting immediate responses on their phone, etc.

[00:25:50] And people are used to communicating that way.

[00:25:52] **Jonathan Duarte:** And they nearly sleep with it, too.

[00:25:54] **Michael VanDervort:** Exactly. I do. But at some point, if somebody makes it through all those [00:26:00] nets and knockout questions, at some point, the tool has a back end on it, and there’s a person there, right? So tell us a little bit about that.

[00:26:07] Yeah.

[00:26:07] **Jonathan Duarte:** Yeah, so there’s two things. That’s a technology allow you to do. The first is if you get a candidate, and these are the kind of words that they’re using. Recruiting is the qualified available and interested if you pre screened out to get to that point. The next thing you probably want to do is push that candidate into your applicant tracking system.

[00:26:27] So that’s one of the things we’re working on with a lot of the ATS is right now is how do you take That pre screened data that we’ve gotten on a candidate and push it into the ATS, therefore the client the recruiter in this case doesn’t have to use an additional tool. They might, we just push it right into their workflow.

[00:26:43] So that’s a process that’s ongoing. The other process is, again, if you’re in high volume, which, again, it’s a different segment than, say, your professional side but it relates very similar, is if candidates Fill out all this pre screening [00:27:00] information. On our dashboard, a client can be sitting there and respond back to candidates just as if they were using their cell phone.

[00:27:08] But you can imagine your cell phone, number one, you’ve got to use your thumbs. But if you had a desktop and you’ve got, 40 to 60 openings in high volume in the RPO world, maybe up to 100 and you could sit there and respond back to candidates. You can look at the candidate’s profile, you can see their pre screening information, and then you essentially just send them a quick message and say, Hey, I’d like you to respond and actually apply for this job.

[00:27:33] So again, it depends on the setup of the client. Again, it’s all customizable for the most part. But if a client If the recruiter wants to do pre screening before the applicant tracking system because of the volume play, then the recruiter can sit there and ask candidates, review the profile information, and the ones that, answered correctly and got through the QAI, Can, they can just send them a response, here’s a quick link, go to the ATS, [00:28:00] but they already know who the candidate is.

[00:28:02] And here’s the difference in the funnel, it’s hard to describe, I like to do this stuff visually, but what normally happens, say if you’ve got 300 candidates that look at a job on Craigslist, 100 of them go to your landing page for your job, and only two make it out of the bottom of the funnel who actually applied.

[00:28:19] You’ve got a 98 percent drop off rate, which is really difficult. Especially if you’re in a high volume. So the other side and one of the solutions we’ve done, we’ll publish soon enough, is get the candidates to apply directly off of the Craigslist ad. Say, text blah, blah, blah to this phone number, blah, blah, blah.

[00:28:37] And as soon as that candidate starts that process, you get their phone number. Then you get, and then the chatbot asks for their first name. And they, then they ask for their last name. So within three questions, ten seconds, you’ve got 95% of the people who clicked that link have given you all that information, which now you’ve got 95 candidates.

[00:28:56] Now you don’t know all the stuff about them, but [00:29:00] they’re starting to fill out the process and you can interactively engage with them from a desktop application that allows you to send SMS messages. If, remember we were talking about Sergeant Star as well, if you had a Sergeant Star kind of application or a career chat on your website.

[00:29:18] You could use the same dashboard to communicate with candidates that are coming in through your career site. You could also use the dashboard to communicate with candidates coming in through Facebook Messenger. And what we see is, if you’ve got this the candidates are not email based, and therefore having problems getting into your applicant tracking system, Or it’s getting difficult to get responses from that user base to get them into scheduling.

[00:29:44] And so you’re sitting there on the phone, playing a lot of phone tag. If you feel like you are seeing this in your company or you’re, getting beat over the head with it every day, I suggest try to go to this kind of text based solution. And then [00:30:00] the handoff, as you asked about was, The other thing I noticed was you could use the dashboard to then communicate with the candidates.

[00:30:06] It’s almost using your cell phone, but if you’re going to talk to a lot of candidates, You don’t want to be using your thumbs. You’d probably rather be behind a desktop. And then, additionally you have a centralized pool of all the data of all of your recruiters on all of the candidates. It’s that integration piece, but it’s in sales and marketing, we call marketing qualified leads versus sales qualified leads. In the TA side and recruiting side, we’d call them like, like your CRM or talent network versus your applicant pool. So you’re essentially building up this huge And huge, because in this one case study we have with this one client that we’ll be publishing, normally her applicant pool would be considered two on an ad.

[00:30:51] They would only get two people through the process. On the talent management side, that same ad would drive 95 people that she could [00:31:00] contact. If you’ve got an ad and they’re not getting through to the applicant tracking system, you can go out and reach out and grab them and push them into the ATS.

[00:31:09] And our client that we’ll talk about did just this. Candidates were coming in from an ad on Craigslist and going through pre screening. She jumps onto the dashboard. sends a response back to say, Hi, Nancy, this thanks very much for filling out the information in a short form. Could you go over here to our ATS, fill out a form, and within an hour, the candidate was getting scheduled for interviews.

[00:31:32] It’s just that fast. If you were to try to do that, Via email or phone. You, it might take days to get a response and get somebody back into the system inter

[00:31:42] **Michael VanDervort:** interesting. I scheduled 45. Yeah. I thought we were gonna do 30 minutes. I scheduled 45 minutes. We’re we’re like eight minutes and 20 seconds left on this.

[00:31:50] Yeah. So it’s just, it’s flown by. I, and I’m not trying to shut you up, I’m just I just, so I wanted to do a reset for listeners. Yeah. Our guest is today’s Jonathan Duarte. Jonathan [00:32:00] runs a company called GoHire. They build chatbots aimed at the recruit, at the talent acquisition space. And we’re going through what chatbots are and how, how they work.

[00:32:08] We’ve got about just a little under eight minutes left, Jonathan. I guess one, one question is I it’s a two prong question. One is this technology overlays, you have an applicant tracking system. You have others, all this other stuff that HR people use, this overlays most of those pretty seamlessly, or are there some challenges to implementation?

[00:32:26] That’s one question. And then the second one is, in a very, I know you have had a chart and table, but in a very broad sense, can you give us an idea of the implementation costs for something like this?

[00:32:35] **Jonathan Duarte:** Yeah, the, okay, so the first question is, how does it fit into the workflow?

[00:32:38] For smaller recruiters, it’s really, our smaller agencies and things like that, You it’s really simple. There is, we do have a dashboard right now that you can do on the front end and other companies as well would have the same thing. But the goal is to get them into your applicant tracking system and, or make it as seamless as possible.

[00:32:55] We understand, obviously, that’s the biggest thing I don’t want to provide [00:33:00] and require somebody else to have an additional tool that they need to use. I would foresee that in five years from now, you’ll be able to do all this stuff within your ATS anyways. Some companies are getting there quicker than others but it will take some time because it’s still new.

[00:33:15] And then cost structure from chatbot side. A basic SMS or text chatbot that has, say, like five pre screening questions for three users. It costs 12, 000 or 15, 000, depending on what you’re putting into it. The Expense, why you think, oh my gosh, it’s a little bit more expensive than I was thinking about, is because messaging actually costs the SMS side, every time you send and receive a message, does cost money.

[00:33:43] It’s unlike the web where you just pay your internet bill. This is there is expenses involved with that side. And then it goes up. If you want to do multilingual, if you want to have lots of users, lots of jobs, and you can integrate straight into your ATS you could do it multinational. So the [00:34:00] enterprise things go up, significantly and get really complex depending on what kind of use cases you want to do.

[00:34:05] **Michael VanDervort:** Yeah, and you can also integrate, like Facebook messaging other, not just SMS texting or email. You can, it pretty much will accept any kind of messaging service. Obviously, there’s probably additional programming and stuff to that, right? Yeah, I

[00:34:17] **Jonathan Duarte:** think one, and we didn’t really touch on it that much, but is if you’re familiar with using Facebook and you’ve got a careers page.

[00:34:23] On the careers page, you sometimes you might hit somebody’s career page and see this little window pop up like a little message. If you’re on messenger, we’re starting to see that more and more often. And those little message windows is Facebook messenger. Sometimes they’re automated and they just give you a quick response.

[00:34:40] So if somebody asks a question. But what I like to do, especially the prospects while we’re talking to them, is go to their Facebook careers page, click on the button that says send us a message, and then I send a message and see how long it takes somebody to respond to it. Because I’ve never found anyone that responded quickly unless they have a canned automated message, which is [00:35:00] fine, but it just repeated everything else that was on the page.

[00:35:03] That’s how we actually sell it to clients is that Okay, what if somebody clicked on that little button? Then they could have a virtual recruiter answering all their questions about your company. It could do job search, can find out about your employer branding, can find about your corporate corporate vision and mission.

[00:35:18] And you, from a company side, You’re not hiring anyone else. You’re actually eliminating workload for somebody who has to sit there and respond back, but it’s now 24 7. And so it’s just another way of engaging the audience in a scalable way.

[00:35:33] **Michael VanDervort:** Yeah. Just about four minutes. I have one last question.

[00:35:36] And actually this was, as we were talking earlier this week, I noticed a tweet that was directed at you from Meg Bear. And she mentioned that she had been talking, not to you, but to someone else about chatbot personalities. And I was like, they have them? And it was more of a rhetorical, smart ass question.

[00:35:54] Yeah. There’s probably some legitimacy to that. That you don’t, you probably do want it to have a certain tone, right? [00:36:00] Just some of that kind of design stuff. Yeah,

[00:36:02] **Jonathan Duarte:** yeah, absolutely. You don’t, first I think it begs the question is every, in the opening conversation we have with every chatbot, it doesn’t matter if it’s web based a career site chat, a Facebook messenger chat, or text based chat Our philosophy is every single time we want to tell the user that we’re communicating with that this is a computer, this is a virtual recruiter this is a chatbot we use those terms synonymously, whether people understand what they mean or not.

[00:36:32] We don’t really know, but we are trying to say that this is not a human. And we think that’s imperative for everyone. Wrap it back. And then, where the personality comes in is if the chatbot can’t answer a question or gets an error. And it happens all the time. When, you and I ask each other a question oh let me ask it a little bit differently.

[00:36:52] Or, it, we, In a normal communication, there’s always back and forth and banter and things. So in that banter that goes back and [00:37:00] forth, the best way to do this is make it personable. And even though it’s corporate in many cases, you still want to have a personality we can, our chatbots can tell funny jokes, everyone’s programmed, all, they’re all programmed with a default set of, if I don’t know the question, I do funny things, like one of them that I think is funny is That my son came up with, who’s 10, which was I’m a I am a robot.

[00:37:25] I feel like I need a little WD 40. I don’t really understand what you said. You do those kind of things, and, you laugh about it, and you go Hey, it makes it takes the edge off and says, okay, I get it. Now, what’s cool about the chatbot on the other side is that you don’t have to spend time if you really just needed a quick answer as a recruiter from somebody.

[00:37:45] You don’t have to spend 15 minutes on the phone listening about somebody’s day when all you needed was yes or no to an answer. But the chatbot, really because there’s no psychology to it, They just answer questions [00:38:00] and, but Meg’s question was around, interesting, like the names. Our chatbot, the one that we made public in November is called Gobi.

[00:38:08] And it’s I don’t know what the word is, but he has a tie, if it’s a he but we don’t really, Give him a sex, other than on his picture, he’s got a tie if we call him a he. But the idea is, from our side, it didn’t really matter to us. And I know there’s a lot of female ones and we seem to be one of the very few that does have a male name or something along those lines.

[00:38:31] **Michael VanDervort:** I don’t mean to cut you off, but we’re literally down the last 30 seconds. Tell people where they can find you Twitter or whatever. You

[00:38:37] **Jonathan Duarte:** can Yep, you can reach me at JD at GoHire. com or GoHire. com and love to, answer questions and teach you more about what chatbots do.

[00:38:48] **Michael VanDervort:** I really appreciate the time you gave us today and earlier this week.

[00:38:51] It was really helpful. There, there’s a probably, we could probably do another half hour, but anyway, we’re going to get out of here. I’ll send you the link, Jonathan, for the show and appreciate you being here. [00:39:00] Thanks.

[00:39:00] **Jonathan Duarte:** All right. Great. Thanks everyone. Have Bye-bye, Mike.

[00:39:02] **Michael VanDervort:** Bye.

[00:39:03] What was that from the director of the Invisible Man. What’s wrong with daddy? He got infected Lum house, producers of the black phone. Something’s happening to

[00:39:11] **Michael VanDervort:** me. Can you hear me? You.

[00:39:12] I can’t understand you.

[00:39:13] **Michael VanDervort:** Mommy, he’s

[00:39:14] changing. Wait! Daddy, is that you?