Back in 2018, we had a little competition called Death Match that pitted four startups against each other to win a coveted gold chain. It was New Orleans and the Bloody Marys were flowing. Conversational A.I. solution TalkPush, and its flamboyant CEO Max Armbruster, was there to pitch their company to a panel of judges. We go back to that moment in 2018 and then catch-up with what's been going on since then, as well as glimpse into the future for what's next.
Enjoy.
TRANSCRIPTION SPONSORED BY: Disability Solutions partners with our clients to build best-in-class inclusion programs and reach qualified, talented individuals with disabilities of every skill, education, and experience level.
INTRO: Hide your kids. Lock the doors. You're listening to HR's most dangerous podcast. Chad Sowash and Joel Cheesman are here to punch the recruiting industry right where it hurts. Complete with breaking news, brash opinion, and loads of snark. Buckle up boys and girls, it's time for the Chad and Cheese Podcast.
Joel: Oh yeah, it's your favorite guilty pleasure, aka the Chad and Cheese Podcast. I'm your co-host, Joel, joined as always, the feral kid to my Mad Max.
Chad: Yes.
Joel: Chad is in the house and today, holy shit, it's a reunion show. You know him, you love him, you remember him, let's welcome Max Armbruster to the show.
Max Armbruster: Thank you, thank you.
Chad: CEO, founder, Talkpush all the way from Hong Kong, now you're in Ireland. So tell us a little bit about that real quick, what was the big move? What happened?
Max Armbruster: It's a personal move. We're a company which is not anchored to any specific location we're kind of nation-less and so Ireland is good because we cover a lot of customers in Latin America and Asia, so I have more normal hours now and Hong Kong was becoming very restrictive for certain citizens, it was hard to travel, you couldn't download TikTok and other things like that. So it was time to go.
Joel: The real question is, have you embraced Irish Whiskey? And if so, which is your favorite?
Max Armbruster: I go buy scotch from Scotland, I do drink Jameson when that's what's on the menu. So I'm equal opportunity drinker.
Joel: I'm sorry, the correct answer is Green Spot for everyday.
Chad: Just something you should have on the mantle.
Joel: Yes, yeah, something you should have on the mantle.
Chad: So back in 2018, Max actually got on stage during a little thing we call Death Match, and this was back in New Orleans, and we had four on stage, Max got up there and he gave it his all and we're gonna listen to that first pitch.
Joel: He was dressed quite captivatingly, if I remember. There was a t-shirt with a boxing hairy chest on it. Am I remembering that correctly, Max?
Max Armbruster: Yes, that was me. I went up on stage with boxing gloves and a naked torso that was on top of like a kitchen apron.
Joel: Yes, yeah, I remember it pretty well. Nola was out of control that morning.
Max Armbruster: That's okay. It wasn't my actual torso, so we don't know each other that well.
Chad: It was a good entering point though.
Joel: We're gonna hear Max's original pitch back in 2018, and then we're gonna get a refresher, an update on what's been going on with Talkpush. Here we go.
[music]
Joel: Alright, next we have the CEO, Max Armbruster of Talkpush, bring it. Push it real good.
Max Armbruster: I was so scared of this, I was so scared of that moment.
Chad: Alright, Max. You can hear it. So two down. You ready?
Max Armbruster: Ready. Ready.
Joel: Love it.
Chad: A little anxious.
Max Armbruster: A little bit. Yeah.
Chad: Would you like me to get off the stage so you can do the shit?
Max Armbruster: No, I think we can switch.
Chad: Fight.
Max Armbruster: Let's do this. Alright, thank you. Thank you, Chad. Good morning. Talkpush in two minutes. Recruiters like to have it all. On one hand, they wanna use AI and big data.
Max Armbruster: I'm out of breath.
Chad: I can tell.
Max Armbruster: And want to automate all the boring stuff, like scheduling interviews, doing background checking, re-engaging, all databases.
Chad: This is where he almost faints.
Max Armbruster: But on the other hand, nobody wants to be hired by a robot, people wanna talk to people, they wanna have a personal, individual conversation, I'm a little bit out of breath from running on stage. And so it's very hard today with an ATS and a CRM to get both of those experiences, the automation and the personal touch, until Talkpush. So what we do is we work at the front of the marketing of the recruitment funnel, and we automate the initial engagement with a candidate with conversational agents that optimize conversion rates. So that translates into a 70% reduction in marketing cost per hire, because those conversational agents aka chatbots who are talking on WhatsApp, Facebook Messenger, on career sites, they're really good at getting the candidate to that next step. Once you're at the next step, then you wanna talk to a recruiter and with our experience, the handover is seamless, so you go from talking to a bot, if you're a candidate, to a recruiter seamlessly and the recruiter can then take over...
Joel: Seamlessly...
Max Armbruster: From their mobile app and talking to the candidate. So we use the word augmented recruiter to talk about what the recruiter can do.
Chad: And when you say that this is seamlessly, red flag.
Max Armbruster: So much more with our system, this is why millions of candidates are already being processed on Talkpush with big brands like Adecco, Accenture, AIA, Starwood, using our system. They come to us because they realize the way you talk to candidates is a very core part of your employer brand, and they use us to level up the recruitment experience.
Chad: Excellent.
Joel: You said augmented back in 2018. That was awesome. Something that we would say very much in the next four years.
Chad: This is the where are they now episode, right? So once you got off that stage, what was the start of the journey to get to here? What's been different? Other than a pandemic. Right? Talk a little bit about that. What's changed?
Max Armbruster: First after that stage experience, I had to go get a Xanax, just to come back to baseline.
Chad: Some makers markers, some shit like that.
Max Armbruster: I think it's quite honest, the presentation that I made, because I really focused the pitch on the front of the funnel and that's what really chatbots we're good at. It's a UI-type technology that helps to increase conversion rates at a certain point in the journey. But because at that same time we were talking about augmented realities and machines and we saw a lot of science fiction come out, people would imagine that it could do so much more, that every conversation was contextualized and personalized.
Chad: Yeah.
Max Armbruster: And so I think there was a bit of a moment of deflation in the industry of what conversational AI could do since then. But we kept on just working at the front of the funnel mostly. And only recently have we started contextualizing conversations for the later stages, being able to answer questions from candidates when they've been hired or just before they've been hired and asking them additional information. So that's... That's from the... There's just broad technology standpoint, from a company standpoint, growing. Obviously the pandemic was a shocker to the system for us in a period of two, three months, we went from... We tripled our volume and, we did not triple our revenue at that time. We just tripled our volume, which meant we had to spend three times more money on Amazon. So that was a bit of a learning curve.
Chad: Yes.
Max Armbruster: And we were kind of a semi-remote already, but we basically went full remote and are now nation-less as I said in my intro.
Joel: You've taken very little funding according to Crunchbase about $1.3 million while companies were raising tons of money in that time period. What was your mentality around money? Are you just profitable enough to like not worry about raising money? How does fundraising play into your business model particularly over the four years since we last spoke?
Max Armbruster: And if you're gonna play videos of me from four years ago, back to me four years from now, I have to be cautious with what I say, because fundraising is not a philosophy or, it's just a tool that you use when the time is right. And I suppose the time wasn't right for us because we saw that every customer deserved a lot of hand-holding, a lot of attention, and therefore, just doubling or tripling the size of the team wasn't gonna crack the case on conversational AI, that you needed to kind of work through the pains of the technology where you could gradually automate some of the jobs. And that takes time. So I think our time wasn't there, even though there was obviously a lot of money in the market. Yeah. No regrets. We're around. We're still growing and still building while others have come and gone.
Joel: Any plans to raise in the future?
Max Armbruster: Sure, sure, I don't think money is dirty, just 'cause I'm French.
Joel: Amen to that.
Chad: Talk about growth. I mean, you guys have... You've been a big player in APAC, but that's been kind of like your core and not to mention what South America, Mexico. Talk a little bit about growth where you saw gaps and you wanted to fill those gaps. What areas.
Max Armbruster: Growth can, as I said, you have to have happy customers with a stable technology to grow. So that's the basis. And we lost a bit of time chasing smaller customers for a while. And I think every entrepreneur wants to please everybody and wants to please other entrepreneurs, and so they want their technology to be adopted by all. But the reality is our technology, conversational AI is most applicable in high volume situations where there's a lot of repeat. So you have to embrace working with corporates, that means like investing in infosec and security and integrations and workday integrations and things like that.
Chad: Things that suck. Yeah.
Max Armbruster: Yeah. So it's a growing journey for sure. And that's where the growth is, and where the growth was low was in the S&B market, even though we spent a lot of time on it. And I like the emerging markets because they are a little bit more slow growth, but you can kind of leapfrog from Excel sheet straight to an automated recruitment platform like Talkpush without going through the ATS step. Whereas, in more developed market, perhaps in the western market, everybody's got an ATS in place, so then you have to integrate with the ATS and then sometimes it's... It just makes things too complicated. So there... We have two happy customers, the ones that never had an ATS and the ones that are working on a really big one like Workday because, Workday is not going away, but they still need another solution to handle all the transactions that we do.
Joel: Interesting. I wanna talk about sort of overall branding of your business and when you first started, chatbot was perfectly fine to say, everyone was okay with, we're a chatbot, and over the last four years, it's conversational AI, which you've used in this interview, and chatbot has become a bit of a dirty word. Is that fair? What's your take on the use of the word chatbot?
Max Armbruster: We all started interacting with them for customer service, and the thing you want when you talk to customer service is to talk to a human, and you're like, "Yeah, what do I have to say?" What's the combination of letters and numbers I have to put on my phone so I can get to a human as fast as possible? 3, 2, 1 or two good. I got somebody on the phone, thank the lord. That's the mentality for most people when dealing with chatbots. And I think that's why it's a dirty word, is because, it's trying to say, let's just get rid of that thing. Let's get past it. Let's get to the next step. I think that conversational AI speaks to a grander vision where the AI will not just be the first step, but maybe the second, maybe the third, maybe all the steps. And some employers now are doing all the steps in one go and the chatbot is still the interface. There's an AI in the back that decides for the 10% of candidates that did get shortlisted, what's gonna happen to them and what questions are they gonna ask. And to be able to contextualize the answers at different stages of the journey. That's like doing double the amount of work for only 10% of the volume. That's what has to happen to go from talking to candidates to talking to applicants. And then if you go from talking to applicants to talking to hires, that's another 90% drop.
Max Armbruster: So again, you have to triple the amount of work to impact only 1% of the volume. So what I mean to say is that the last mile is very hard to do 'cause you have much less data and there's the economics are harder to build a full end-to-end platform.
Chad: Seems like a maturation through data. And really... And correct me if I'm wrong, but the thing that, what really distinguishes one conversational AI company versus another is the experiences and the data that you have to crunch to contextualize the actual experience itself. Not everybody has the same data, not everybody has the same clients, the same region of the world, et cetera, et cetera, works under the same languages. So, I mean, from my standpoint, trying to like separate the two types of company, one conversational AI to another, it's really the secret sauce of the data and the behavior that you've been able to pull together over the years. Talk a little bit about that.
Max Armbruster: Yeah. If our sales team talks to a company that wants to automate, but they have, let's say, 2,000 job openings and 200 different job description, over 10 different sites, I'd be like, this is not, this is too hard. And we're not gonna do it. But if it's a company with 5,000 employees or 10,000 employees and they have five sites and three job descriptions, then, yeah, that's great, that's perfect. So that's where it's hard because a conversation can go in so many different directions if you... The Chad and Cheese Podcast is like talent acquisition tech. Straight talking, and that's it. You know what you're going for.
Joel: And sex appeal.
Max Armbruster: And the sex appeal, and the alcohol. And the alcohol to forget. Yes.
Joel: By the way you can't see this, but Max is drinking wine while we're doing this interview, so he may forget this, so you mentioned your AWS bill going up exponentially without an increase in revenue, obviously, COVID was a curveball in your business, what was another thing that caused sleepless nights, was unexpected, 'cause I know a lot of entrepreneurs out there, startups that are listening are dealing with things all the time, and people don't really appreciate, every day it's putting out new fires, what was a particular hot fire that you had to deal with since we last spoke?
Max Armbruster: Traumatic experiences from the past, we lost our head of engineering in early 2019, and when you have a platform that handles 10 million conversations, 10 million candidate interactions, 10 million candidates a year...
Joel: It hurts...
Max Armbruster: Yeah. That's a scary moment for sure. He was our, kind of our CTO or founder. And yeah, that was a pretty scary moment. I think early 2020, when COVID happened, all of our pipeline basically disappeared from one day to the next, because everybody froze their spending, and as a cash flow positive company that relies on our customer to pay the bill, that was actually more scary than if we had been a VC-backed company with three million sitting in the bank, so we had to make a few twists and turns, but, yeah, and I think all good, you sign up to be a company builder so that you have a few scars and lessons.
Joel: You sign up for the roller coaster, right?
Max Armbruster: Yeah, yeah. So those were probably some of the low lights.
Chad: Well, talk about the rise after COVID, because we started to see, obviously, there was a huge dip, and then there was an explosion, and then money was... It seemed like money was fucking everywhere, so what happened at that point. Because you see so many companies when they're looking at expanding and contracting, they have to do that around talent as well.
Max Armbruster: The explosion, as I said, was one in traffic where it went from, it tripled in a period of three months for us, and while it was difficult to process from a cost perspective, it meant that the time had come. It basically meant that for a certain category of employers, the first screen would not happen with a phone screener, there wouldn't be no more phone calls, and once you've turned that corner, it's not something you can go back to. Automation is like a one-way street where you don't go back three years later and say, I remember the old days when we used to go through the fax machine and read the resumes and decide who to call? Once you're past that you're past, and it was the same thing with some of the pre-screening that we were put in charge of. The recruitment teams were cut by 20%, 30% of head counts, the volumes were triple, the amount of positions that were available to fill were lower, so there was no other solution than to use us for our customers and transformation that would have taken a couple of years were done in a couple of months.
Chad: On the crack. That's what I'm hearing. They were on the crack.
Joel: Well, is that the Irish crack or the...
Chad: The Irish crack.
Joel: I don't know if Chad knows what the Irish crack is.
Max Armbruster: It means it's a good bit of fun, that's all.
Chad: That's all it is. That's all it is.
Joel: Chad loves crack.
Chad: It helps me stay awake and get work done. So let's talk. I was lucky enough to actually judge a competition for Recruitment Hackers, which is something that you guys do, and one of the things that I love about a lot of the new tech that I'm seeing, but I really love about what Talkpush is doing, what you've been doing for a while, is getting candidates through the process incredibly fast, and one of those things was via voice, which I thought was incredible, and I think these positions that I was actually applying for, I was mock-applying for to these real companies, they were looking for customer service individuals, and one of the questions were around being able to speak fluent English, so that whole process of using me, my voice. What was that all about? So there literally was, I think, five questions that were asked and then you asked for an audio sample, so can you take us through that and take us through kind of like the completion ratios that you're seeing from start to finish, because as we've seen with Appcast data, 92% of individuals who start an application do not complete it, so only 8% complete. What are you guys seeing from completion rates? And then talk a little bit more about the voice tech that you guys are using.
Max Armbruster: Okay, on the topic of completion rates. Be very wary of what the assessment vendors are saying, because you have to compare apples with apples, the completion rate for somebody who has already been through an interview with a live of human is gonna be drastically higher than somebody who comes from Indeed or Facebook or somewhere else. In our case, our customers, a lot of them were sourcing through Instagram and Facebook and word of mouth, and so they were very sort of low engagement candidates, and when our customers tried to move them to long-form assessments, they would get a 15% completion rate, so it's still better than what the 7% you are mentioning, but low enough that it would make it impossible to hit their quotas, their hiring quotas, and so we look looked for a solution, we said, "Okay, well, is there a way that we can collect this... " Because the assessments that we're talking about were long-form, 20, 25, 30-minute language assessments and people would just drop off, and I feel for the candidates, 25 minutes to apply for a job when you're looking for a job? That's worth a lot of money, because you could be applying to five other jobs in that time, so it's obviously annoying, and so we went around and looked for technology partners that could analyze audio data.
Max Armbruster: And with one minute of data, we have 95% accuracy on something that if you did 25 minutes, you wouldn't get anything better in terms of generating a score for your English proficiency with multiple parameters, like vocabulary, fluency, pronunciation. And we can capture this data talking to people on WhatsApp or on Messenger or on a website. And so, because it's just one or two minutes of audio, it doesn't feel like a test. It feels like, "Hey, tell me about yourself. Why do you wanna work here?" Which is a different experience. It's more like, alright. I'm just saying what I... Why I wanna work here, I'm not being tested. And the experience of not being tested is better if you're talking to leads, if you're talking to applicants, maybe our shortlist applicants, then an assessment is more suitable. So that's our approach on that. And it's working well, always trying to reduce the workload at the front of the funnel.
Chad: So what's the new percentage that you're generally seeing? Just kind of like an idea for completion rates?
Max Armbruster: Yeah. It's about 60% for social media traffic. So we used to have like 70% completion rate for people to just apply, but now it's down to like 50, 60%, with this extra audio question. So it's not a big drop. And if you get qualified candidates on the other end, that's a huge cost saver.
Joel: Like most things in France, it tends to be cool before it's cool anywhere else. And you were work-from-home before, it was cool. Talk about... I assume you're in favor of it, but companies out there that are thinking, do we go hybrid? Do we go full work-from-home, and no office whatsoever versus your asses back in the office? What's your stance on all that? And what tips might you give companies who are looking to go work from home 100%?
Max Armbruster: You have to hire people who are very good at writing, you have to embrace asynchronous communication. You have to reduce the amount of meetings, and you have to have a very tough layer of management because the far from the heart, far from the eyes, far from the task sometimes. So it's not for everybody. I wouldn't advise it for everybody. I think it's great for us and it keeps our people happy and we managed to keep them. And that's great and obviously it's got all the benefits of expanding your talent pool. This year we didn't hire a lot, but we hired people in... And someone in Venezuela and Nigeria and Brazil, where we don't even have customers, but we still hire there because the talent is everywhere.
Joel: Talkpush engages with leads in real-time when they click on ads in Facebook, Instagram, TikTok, Indeed, et cetera. What's your take on how employers are leveraging social media? Which platforms get the most activity, just overall social media in recruiting?
Max Armbruster: Instagram and Facebook. Facebook gets a lot of bad press, but it has the most advanced advertising platform of all. You can do the most amount of targeting. They made things very difficult in the last two years because there was a layer of banning targeting ads. Because they wanted to be compliant with equal employment opportunity law in the US. And so they said no targeting at all, which slowed down the adoption of this wonderful tool, which is where you could say, "I wanna do an ad just for a woman, and just for a woman of a certain age." And why wouldn't that be a good idea? It's a great idea. But out of caution, Facebook kind of scaled that back a bit, but without revealing too much, you can kind of work around it and you can run ads that still, to target certain populations. And Facebook gives you the best cost per lead and the best targeting options.
Chad: So let's talk about the evolution of the relic that we call an applicant tracking system. We're seeing more of tech stacks and what I like to call core talent platforms. I see Talkpush and Paradox. You guys have really created an engagement layer that is necessary to feed the beast being that that core talent platform. Right? Can you tell us what you think the future of this stack looks like and how it's powered?
Max Armbruster: Adding a chatbot to an ATS is kind of like... I don't know what a good allegory is, but it doesn't work because basically an ATS is, at the foundation is just a place where you send email updates to your candidates and where you manage a career website. So if you don't need a career website and you don't need email, then kind of like, why do you need an ATS? Then you have to rebuild everything. So I think that it's gonna be hard for the traditional ATSs to really leverage messaging and messaging is what we do, is texting, WhatsApp, Messenger is how we communicate with candidates from the start to the end. Now we can send a link to somebody on WhatsApp and say, click here to see your employment contract and sign your employment contract. Nobody else is doing that. So where do I see... What was your question? How is that gonna evolve? Well, I hope it's gonna take off.
[laughter]
Chad: From what we've seen thus far, I think you're a hundred percent correct. I think it's like putting a V8 engine in a horse and buggy. I think that's what... That's pretty much what you're looking at, is that's not... That's what an applicant tracking system would be trying to do to build it from that old horse and buggy to try to give it some superpower. It's just not gonna happen. They're going to have to restructure and rebuild from the ground up.
Max Armbruster: And another way to answer that question is to say there's an ATS for every type of candidate, and obviously there's a huge segment of the market, which is perfectly fine with the way things are. Email, long forms, whatever. There just happens to be a segment of the market where that's suboptimal.
Chad: Yeah.
Joel: Oh shit.
Max Armbruster: And just like we have ATSs for truck drivers and for healthcare workers, it's normal that we have an ATS now for high volume, for retail workers, and for people at the early stages of their career where a resume doesn't mean anything.
Joel: How has pricing evolved over the last four years?
Max Armbruster: I'm not equipped to tell you, but we are working with one of your sponsors and they're working fine.
Joel: I'm gonna let you out on this one, Max, better French based band, Phoenix or Daft Punk.
Max Armbruster: I've never heard of Phoenix.
Joel: So Daft Punk is your answer. [laughter]
Max Armbruster: That's it.
Chad: I think The Stripes are French-ish.
Joel: The Stripes?
Max Armbruster: The Stripes. It's a New York based band, but I heard that the lead singer went to like a bilingual, to a French lycée, basically in New York.
Joel: Not The White Stripes just the...
Chad: The White Stripes.
Joel: Jack White is French.
Chad: Hearsay.
[overlapping conversation]
Max Armbruster: Or is it The Strokes? Anyway, it's one of them.
Chad: Anyway...
Joel: And Julian Casablancas?
Max Armbruster: The only rock I get now is from the intro of the Chat and Cheese Podcast.
Joel: Man, you are as old as us, Max. You are as old as us.
Chad: Here's the last one, with the advent of ChatGPT, how do you think that's going to impact chatbots, conversational AI? I mean, is it gonna make it easier because you have a different OpenAI set? Are you already using it? Tell us a little bit about what you think about ChatGPT.
Max Armbruster: I had a chat with Quincy Valencia recently, and I'm gonna steal her idea and repeat it on your show.
Chad: Okay.
Max Armbruster: She said the first application for ChatGPT will be to help recruiters write job descriptions better. And that makes a lot of sense to me, 'cause I played with it and it obviously it writes well. And so that's probably the one of the hardest and most critical parts of recruitment. I don't see a short-term future for us to apply this technology in our environments 'cause we're happy with our natural language processing engine as it is. But of course I'll keep my eyes open.
Joel: There you go. Max Armbruster from Talkpush. Max, for our listeners who wanna know more about Talkpush, where would you send them?
Max Armbruster: Talkpush.com and The Recruitment Hackers podcast if you like the sound of my voice.
Chad: Oh bullshit.
Joel: Chad, another reunion show in the can we got all the feels on this episode. We out.
Chad: We out.
OUTRO: And thank you for listening to... What's it called, Podcast, with Chad, with Cheese. Brilliant. They talk about recruiting, they talk about technology, but most of all, they talk about nothing. Just a lot of shoutouts of people you don't even know. And yet you're listening. It's incredible. And not one word about cheese. Not one cheddar, blue, nacho, pepper jack, Swiss, there's so many cheeses and not one word. So weird. Any who. Be sure to subscribe today on iTunes, Spotify, Google Play, or wherever you listen to your podcasts. That way you won't miss an episode. And while you're at it, visit www.chadcheese.com. Just don't expect to find any recipes for grilled cheese. It's so weird. We out.