September 22, 2022

Robin Choy, CEO at HireSweet

Robin Choy was born in France, founded HireSweet in 2016 - we help companies hire for roles where they lack applicants (like software engineering, sales, etc.). Spearheaded sales and marketing from the 1st to 1000th client, grew the team to 40 people while raising a single Seed round of $1.5M, and got selected by Y Combinator.

Julian: Hey everyone. We're here on the behind company lines podcast with Robin Choy CEO and founder of HireSweet HireSweet was founded in 2016. They help companies hire for roles where they are lacking applicants. HireSweet has two main products HireSweet marketplace and HireSweet CRM, which we'll dive into in a moment here.

But Robin, thank you so much for being on the show. Really excited to chat with you, learn more about your background, your experience, and honestly, just to jump right in. What were you doing before you started high sweet. 

Robin: So thanks a lot for having me. And it's always a good experience to be a, a podcast guest.

Yeah. actually a bit stressed out because I host a lot of episodes, but being a guest is a completely different game. So thanks for having me the before I started the company. So I just started the company fresh out of sChoyol with doing maker founders that were in sChoyol with me. And we started a company hire suit as a sChoyol project.

So. I was 

Julian: in sChoyol. I love that. What, what drove you to use this in sChoyol? Were you seeing the recruiting side from [00:01:00] inside of sChoyol and they wanted to tackle it? Where did the conception of this idea come from 

Robin: as. most of the, most of people working in recruiting will say, I kind of fell into recruiting.

Yeah. Didn't expect to go there. Wasn't trained for this. And that's something that you hear a lot for people from people working in recruiting, mostly recruiters. And that was the same for me. So we actually worked on a project. So two of my co-founders Paul Ismail, we worked on that, on that sChoyol project originally was a common track between different specializations.

I was already friends with Paul and then we met with Ismail and started working on a project mostly based on Ismail's expertise and Ismail's expertise. Doing competitive programming, it was a coach for the French team in the international Olympian informatics. I didn't know there was such a thing as the international informatics.

So that's what we discovered. And remember we were just fresh out of sChoyol, so we didn't know much. I [00:02:00] had internships, it was always passionate about financial markets. So I worked in a bank before, but didn't really know much about the corporate world. So we started working on this and we had that idea that we could leverage and use Ismail's expertise to to have an application in a corporate world and competitive programming in a corporate world, translates into software engineer, engineering assessment, and a lot of companies actually needed to assess and still need to assess the software engineers that, that they hire.

And we built a very quickly, a top. Assessment service for companies hiring software engineers. And so that was basically first, it was kind of a market technology push. So we started with the technology and the expertise that we had PTO first product, and then started to talk to companies about that.

Which was good because we had a product so we could talk to companies and they will actually buy from us. Or at least we were trying to see if they [00:03:00] would buy the product. But pretty quickly we realized that most companies didn't have any problem or they did have an, an assessment problem, but their biggest problem was on the sourcing.

And so all these companies basically told us the same. We're we'd be happy to pay for your product. We'd be happy to use it, but the problem is we have no software engineer to take those tests. We really struggle getting good software engineers. And actually when we do get them in interviews, we're not too worried about spending some time with them because we just wanna make sure that we do assess them well.

And there are not that many people that we wanna automate the whole thing. So. Lot of companies told us in, in the end, we'd be happy to pay for the product, but you have to find people for us. So that's kind of how we fell into it. And we realized that the problem was that especially tech startups didn't have enough applicants on software engineering rules.

And that's [00:04:00] how we discovered that the. Recording world is not that easy, that simple, that transparent it feels like you just have to post a job online and then you get applicants. And on the other end, as an applicant, you just go through different jobs and apply. And then if there is a good job for you, you get higher in the end.

This is much, much more complex than this. There are a lot of biases. There are a lot of problems, lack of transparency. So when you discover this again, 23 or 24 years old. Suddenly it's a, like, everything is possible, right? There are so many problems at every step that you can solve using either the expertise as again, competitive programming or also other expertise that we quickly built around artificial intelligence and, and data science on how to improve the process as a whole.

So that's how we started. First we had that expertise. We pushed that to the market and the market told us. That's good that you have that expertise, but we don't actually need this. We did. We need that other thing from you. Can you [00:05:00] help with this? And, and then that's when we basically said yes, we can.

Julian: I love that. Yeah. It eChoyes a lot of different stories. I've heard from, from founders on the show where it's, you know, they, they have one product that they're, that they're bullish on and that they're putting out into the market, but the feedback then helps them adjust to identify. More. So what a customer's problem is and you have to, it's not even a pivot.

It's really just like a micro adjustment. One, one founder said that product is it, product market fit is a point in time or something like that. Somewhere, something along those lines I'll have to review what exactly he said, but it's really interesting how you know, the feedback in, in the market and your customer will help you.

To identify what that fit is. But tell me a little bit more about the gap that that is you know, relevant in, you know, the recruiting world in applications. You said, you know, companies, it seems like they just post a job. There should be plenty of applicants. And then, you know, developers should just, you know, apply to many jobs and, and get the one that they want.

What is the disconnect? Is it the [00:06:00] communication? You mentioned something about transparency, you know, where, where is it that you see the disconnect is between the two parties. 

Robin: F frankly it's everywhere. . Yeah, but mostly yet it has improved over the past six years when we created the company. I would say that the, the main problem lies in the recruiting process in itself that is basically selling or buying a job depending on where you look at.

And it's a single, there it's a single opportunity. It's the most, it's the rarest community. It's just one job. And every time it's just one job. If there are even if there are different openings, it's just one job for one person. And so basically you could hire anyone or you could hire no one. Also there is no pressure on time.

And so the result of this is that a lot of companies are really stressed on the person they wanna hire. [00:07:00] They wanna keep their, they wanna keep a assert the upper end in that relationship and that's the same for candidates. And so in the end, everybody ends up retaining a lot of information. So the exact nature of recruiting makes it so that everybody will place all the actors in recruiting need to keep a lot of information for themselves.

And so companies will pass job. But there will there will just mask most of the information that will make it interesting for candidates and candidates will apply, but there will keep some of information for them. So there is this lack of information, transparency, and also the fact that not only the message itself lacks information and transparency, but also the message.

Lots of time doesn't even reach their recipients. So you will pass the job online, but the person that you're targeting won't see that job, or you will send them a message on LinkedIn, but they just won't check their line inbox. So not only is the message bad and companies [00:08:00] and candidates like have really improved this and we really see much more transparency.

And that's one of the topic that we discuss a lot on our own podcast as well. So much more transparency to the recruiting process. Mm-hmm . Companies know that they have to be proactive about transparency and give that transparency to candidates. There is much more transparency on candidates as well with public resumes, LinkedIn, et cetera, etcetera.

So that improved. But then the second problem is the message not re reaching the recipients because there is so much nose very difficult to get signal and, and that's another problem. So that's the reason. Recording is so hard and so interesting at the same time. And I don't believe those problems will be solved anytime soon.

So you don't have to you can't solve those problems, but rather you need to find a way around those problems. Yeah. 

Julian: What, why do you think they hold withhold information? what do you mean? Like, well, yeah, no. In terms of candidates and, and, and and companies, do they, do you so 

Robin: very [00:09:00] easy companies don't want to, they're hiring for a role.

First, they don't know exactly who they're hiring for because they just say, okay, well, we'll figure it out. So sometimes they don't just return information on purpose. It's just, they don't have that information themselves. So that's the number one symptom of a bad recruiting process. You'll have a recruiter hiring for a.

But they don't exactly know what, who they're hiring for because the hiring manager doesn't know either. So they just, that's the worst thing recruiting is we'll know it when we'll see it. Yeah. So that's the number one thing. And then the second problem with information and that's the same for candidates, right?

A lot of people don't exactly know what they want in their next step. They believe they want a a better work life balance, but maybe this. Something that they will get, or they will, they believe that they want to work in that specific industry or want to avoid that specific industry, but will end up there anyway.

That's something that we see a lot as well. So that has to do with [00:10:00] discovery early discovery. You really need to do discovery in the company and make sure that you surface the information. Number one. And then number two is how do you communicate that information and communicating that information is always losing a bit.

Bargain power mm-hmm and that's the reason why people don't wanna be revealing information. That's the reason why they don't wanna be revealing salary ranges. Sure. Because, or, or, or sometimes another problem that we see as well is companies have that person in mind, but they will put a job online and add a lot of extra requirements.

That the person doesn't actually need to have, but they think that they need to, I don't know, like target that ideal person that would be right, exactly in the scope. They're just looking for someone who is good enough to do the job, but they will try and find that specific person. And that's retaining information because they don't wanna say we're targeting, we're talking someone with five years of experience, but we would actually settle with someone two years [00:11:00] of experie.

Because otherwise they will get too many people applying it. They don't want to go through all these applications, et cetera, cetera. So yeah, most of the time they don't have the information themselves and when they do, they try to retain it because more by getting power also, they think. Yeah, I'm speaking for both sides.

Julian: Yeah. Yeah. It, it, it's fascinating to see. And I'm always curious if it is just, you know, better to be transparent because, you know, if, if you get what you want as a candidate and you get what you need as a company, it seems to, you know, be the perfect scenario. But we do see that a lot where they put all these extra requirements on.

It doesn't necessarily, you know, candidate doesn't necessarily need to come with them, but they need to be able to have a certain personality trait to then maybe learn them. Which, which is this whole, you know, conversation and and kind of like, you know, thing you have to mend and mold until you can kind of put something into the right position.

I'm always curious because a lot of companies go, you know, think about the automated [00:12:00] route. You know, there's a lot of automated systems marketplaces that. You know, match companies and candidates and it, in an ideal world, it seems like, you know, it would work out, but I feel, and I'm, I'm curious on your take on this that it, it will never be automated because there's too many nuances.

There's too many different pieces of information you need to You know, take into account when matching you know, candidates and, and companies and those opportunities, but curious about the automation piece. Do you ever think that, you know, recruiting is going into an automated kind of more AI route or is there always gonna be that component of needing that human touch to, you know, make those micro adjustments to fit candidates within jobs?

Robin: So recruiting is irrational by nature, and we've had experiments where we there, there is a [00:13:00] very interesting experiment that we led. We give a company, a shop list of 10 people that will reach out to only 60% of them and reach out to six out of 10. We give the company the same six. They will still reach out to 60% of them and reach out to four, really?

And you give them 20 people. They will reach out to 14 out of these out of this or 12 out of these 20. So it's irrational. It's very irrational. So in the end you cannot avoid the human making the decision. Because as I said, you can build the best matching system. But if it's based on a poor discovery, if the person says, okay, I'm looking for someone with five years of experience, but they would actually settle with someone with three.

You can build a, the, the most perfect matching system around this. It won't work because it's lacking the discovery. Yeah. So my view, and though we, we do develop a lot of matching system ourselves. We, we do invest a lot [00:14:00] in, in AI. We have a six people data sense team, and that has been the case for six years.

So we built a lot of things internally. And my take on this is you. Try and replace the human. You should just try and enhance the human. So try and find that right balance between saying this will be a perfect match or this will be a good enough match. And the goal shouldn't be to say, this is the person that you will hire, because I can guarantee you that if you Choyse just one person, the company will not hire that person.

Even though that. Perfect person that just wanna make sure that they see other people as well. So the goal should me to say the AI said that this is the person that you should hire. But the goal should say out of these a hundred applicants here are the. 20% that are on the most relevant for you to look at.

And here's why, and just maybe guide your eyes, just watch out that person. You might have [00:15:00] missed that person because you might have missed that experience or just highlight that experience. And I believe that's where AI and technology is going. And that's actually a big trend in the recruiting industry.

Everybody's moving away from fully automated systems like video interviewing. At some point we talked and we went too far. We talked about. Automated video interviewing assessments and keep the human completely out of the loop with the help of saving of savings. Sure. Time simply. And I don't believe that.

Yeah. And we're moving away from this and it's much more about rather than saying like fully automated video, interviewings more like. A human that can use assessments on the side to get rid of biases because there's, there are problems with humans as well. Sure. But it's a much more balanced approach between AI technology and the human and that's for the best.

Julian: Yeah, no, that that's, I feel like a lot of companies are going that route with their technology. It's, it's enabling and empowering people to use and [00:16:00] utilize it within their day to day, you know, task list or activity or processes because you know, it. You can't replace any piece of you know, any piece of what a human does.

You really have to enable them to do things more quickly and more efficiently and effect. But you mentioned something around testing and screening, and I'm always fascinated because I feel as though I, you know, I was in recruiting and, and did the whole gambit, you know, sales from sales recruiting to some HR people to now engineers with the company I run.

So, so we're in a similar, similar field. But the assessment piece is interesting because engineering, it seems that inherently needs a, an assessment component because you have to validate their ability to code, right? They, they have to write some level of code to have some idea of their abilities, whether, and, and within that piece, if there's a really good hiring manager, VP of engineering, who's reviewing a can.

they can see the smallest details in, in between the lines of that code. They can dive in there [00:17:00] understand their process, how they approached it, and also how they would approach other problems by some project or assessment that they gave this candidate. But I'm curious because there's so many different types of assessments.

That it's really hard to pin down. What is, you know, one that is, is generally more productive, whether it's like, you know, a take home test or something on coder bite or a video interview, or, you know, work session with someone on their team, like a developer on their team that is reviewed. What have you seen in terms of assessments have been successful for screening candidates?

Like, you know, engineers in particular and, and engineers are a. Amazing bunch because they, they do have a specialization, which I think is inherently easier to control for, in terms of understanding their ability. But I'm curious to hear from your end, what are some assessments that you've seen that are actually successful in identifying the quality of, of talent?

Robin: So I believe there is a consensus on this mm-hmm and took some time but the best way to [00:18:00] assess someone. Is to see them work on a task that is related to the task they would perform at work. Yeah. So the best way to assess an engineer is to have a pair cutting session with another engineer in the team and see how they work on a, on a.

On a problem together. And same for sales marketing just have them work with people because that's how you work once you are in the company, work with other people and try and solve problems. Yeah. The other good thing about doing a pair cutting session is that you can you can build a much stronger relationship that you've done using a standard assessment test.

And and then you can use this to also. Build a competitive advantage of how that person can. You grow as an engineer. So that could be that's an assessment session, but that's also almost a coaching session. Because mostly for software engineers, the problem is that a lot of companies are competing for the same people.

So it's [00:19:00] not actually very hard to know who's good with not. But then once you figure this out, you have to actually hire them, close them. Yeah. And so using that per coding session can also be a good way to to do that, but you're doing a lot of software engineers hiring as well. So how, what do you think about 

Julian: this?

I agree. I agree, honestly. I mean, you hit the nail right on the head. It's, you know, I am frustrated when, when clients are so bullish on a certain assessment because it's a point in time and you don't really understand the relationship. You know, it's hard because, you know, coming from a non-technical background it's hard to communicate that understanding, but after having so many times through this recruiting process, the most successful clients either have a very specific test that is related to a task.

So it, and it's more, and it also depends on the role. So it's like, you know, whether it's a full stack engineer versus a QA engineer versus DevOps engineer, there's different assessments. I think for QA testing automation and manual. You can definitely have a test. I think that's a little bit more you know, that's a little bit [00:20:00] more accepted in my book because you know, that's what they'll be doing.

They'll be assessing code that they get from, that's been pushed from their engineering team and then they'll deploy it or they'll, they'll say it's ready for deployment. And then for DevOps and, and full stack engineering front end backend, that's a little bit difficult because you don't, you don't understand how you'll be able to work with them unless you have a working session, how they understand problems and there's so.

Times in a Workday where you do get something wrong, but you, you correct it with your colleagues and then you, you don't get it wrong the next time. And so that process tells you so much, how smooth can they and how quickly can they adapt to the code? And, and the, and, and the problem that they've seen, and then not, not have that problem arise again, you know, kind of proactively avoid that.

I agree. I think the life coding sessions are the. I would say the strongest way. I, I wouldn't say the only way, but the strongest way for you to identify whether or not this candidate is good for your, your company and your team. I think that's just, you get such a a real life [00:21:00] example of what that would be like.

So, yeah, I, I completely agree with, with with your sentiment there shifting a little bit more towards HireSweet and, and how your company's going. Tell us a little bit more about the traction the companies you're working with and where where things are currently. 

Robin: so we're working mostly with tech companies, mostly.

We have two products again, as you said. So there is higher seed marketplace. It's much more service driven. Marketplace is basically we have a team of talent acquisition managers doing sourcing all day long, engaging with the community of candidates preventing the candidates and then showing them up in.

In a platform that the companies can use. So this is mostly useful to early stage founders, early stage C. It's a contingency fee base. It's 18%. So it's rather expensive. It comes with good service as well. It's always cheaper than a hiring agency. But mostly people that use it are in a rush or need extra support or that extra service.

And they always pair with so that's for the first product and the second product. Is a [00:22:00] recruiting CRM that we launched about a year and a half ago. Again, based on all the technology that we developed since 2016 and it, this one is a subscription based tool to help recruiting teams be more efficient.

Again, we target mostly the same companies, recruiting teams in tech startups of one to 20 people in the rookie team. So that often translates to companies of one to 2000 employees. And we work. We do work with a lot of companies. So we worked with about thousand and 500 employee companies over the past six years.

That again, we're old a lot of them on the marketplace products and we're growing fast. So we we grew two X since last year and know the biggest what's very exciting now is how do you scale two products especially this recruiting CRM, which is growing fast, but scale very. We wanna turn the recruiting CRM into a complete talent acquisition suite.

Yeah. And there's a lot to cover and that's actually solving the same problem that we saw in 2016. That's what do you do when you [00:23:00] don't get enough applicants? How do you do it? Number one, you need to reach out to passive applicants. But how do you do this in an efficient way, tracking your data, measuring things, running experiments.

That's the number one thing. And then you also wanna nurture existing candidates, nurture people that you've already engaged with. And mostly these people are most likely to reply. So it's very important to keep that, that very white glove relationship with them one on one relationship. But then again, how do you do this on a database of a hundred thousand candidates?

How do you do this at scale? How, how can you scale while remaining a very personalized, very. Again, white glove approach. And how do you get data on everything? Because data is key to measure again, what's working, what's not. And are you going in the right direction? Are you even. Are you even likely to hire those 10 people that you need to hire for next month?

Are you on the right path? So that's very exciting. Again, lot of growth, lot of things to do here. We're very [00:24:00] at the beginning. So it's funny to think that even that six years later, we'll still pretty early on. We believe in the trajectory of the of the company and will still believe that there is so much to do.

Yeah, because again, you look at the recruiting industry and there are so many problem. That you just don't know where to start. Yeah. Problems everywhere. And and yeah, so that's, that's where now? 20, 22. 

Julian: Awesome. Awesome. What's what's the biggest risk you're facing today?

Robin: There is, there is a market risk. There is always a market risk mm-hmm one risk that we could see there has been a big trend since 2016. That is more and more a candidate's markets. Mm-hmm originally it was software engineers. Now we've seen that on track drivers. Nurses is getting harder to hire everywhere.

Yeah. So that's the reason why we built HireSweets. And again, our value proposition is to help hire on roles where you don't get enough applicants. Now, if the market shifts and it starts being [00:25:00] a company's market again, like it was maybe in early 2010s mm-hmm that would be a risk to the company because we wouldn't be solving.

A problem anymore. Same as when we started. So I'm pretty sure we would be a job adapt to this. And also, I don't think this is going away. And I think this is a very deep trend, but that's a, that's a big risk. And then the second big risk is poor execution and that's the risk for any company. We wanna do a lot of things.

We, we, we believe that we can. Build an entire talent acquisition suite, but that has proven to be very difficult. Mm-hmm from a software perspective from a go to market sales and marketing perspective. So I hope we'll be able to execute well, we have the team for it. We have good investors. We have good clients.

We spend a lot of time with our clients. Yeah. But that's always a risk. Yeah. 

Julian: Yeah. What's you know, if everything goes well, what's the long term vision for higher.[00:26:00] 

Robin: again, as I said, the CRM turns into a full talent acquisition suite. The hubs companies really provide that white glove applicant. Not only to experience, not only to. 10 applicants at a time, but to hundreds of thousands, how can they scale that much more transparency brings much more transparency to the market as well, and a direct communication between candidates and companies.

So help, not exactly solve that communication and transparency problem, but at least improve it. That's for the CRM marketplace. I would say mostly the same, a great marketplace where companies can hire candidates. And everybody has much, much more information on both sides. And this will go through adding extra salaries, both for candidates and for applicants.

I don't believe that we, we moved wave from solely software engineers for, well, no, we still haven't addressed. Blue collar jobs. Mm-hmm this is something so long-term vision. You're saying everything goes well for [00:27:00] HireSweets. I believe we'll need to address the jobs at, at some point in the next 10 years, but that's not something we'll do for now.

Julian: Yeah, amazing. Amazing. I, I, I love the way you're approaching it. You know, tackling the problems that I, I think will always be there. Whe whether it's skewed towards there's more jobs in the market, or there's more candidates in the market companies will have trouble hiring. Because one it's just risky to hire, you know, and, and have a bad hire.

But two. The benefits of a good hire and making that, you know, that count is so exponential in what it could add to your business that it's really such a value add when there's a company or a third party that can help facilitate that process. And, you know, I battle with this all the time because there's, you know, this whole battle between external and internal recruiters.

But I see them, you know, working together a lot of ways. You know, it's awesome to see a tool like HireSweet, be able to kind of bridge that gap where it can be possible. Right? The, the problem is lack of applicants. And then the second problem is lack of good applicants and making sure [00:28:00] that those hires count.

So it's amazing to see that that's the direction the company's going. And I'm excited to see where you all, where you all land. And bonus question. I like to ask all my guests, since we're, we're running into time here what books or people have influenced you the most? You know, as an entrepreneur, but also as you run HireSweet or just in your life.

Robin: So I'm a big book reader. So I could go on in here, but let me just quote a few. I've had a, a real epiphany reading thinking fast and slow mm-hmm from Daniel Kenman. This is a great one about decision making biases. Great. One, if you are more into fiction and closer to. Docu fiction, maybe all books from Michael leis Lewis, especially Moneyball mm-hmm

And it's a great one, same about biases and how to basically create a team, an, a team and a team of a players. And recently I was I was mind blown by way of the Wolf from Jordan Belfort. I didn't expect [00:29:00] expect much from the book. I was a bit. I didn't really know what you think about it. I dunno, really what you think about the person, but they'd say actually great book.

So, but more of a 

Julian: sales book. I love that. Well, thank you so much for sharing. Definitely always gives me some great homework to do last little bit. Where can we support HireSweet? I always like to give my guests a chance to tell us your socials, your plugs of when we work, we find you Twitters LinkedIns, but please go ahead, tell us where we can support HireSweet and and learn.

Robin: So the best way to do this is to follow me on LinkedIn. I'm Robin Choy CEO at HireSweet. And I post on LinkedIn on recruiting daily. I would say that probably 50% of what I post is really good and 50% is really crap. So I'll make you, I'll let you make your own decision opinion and you can also follow.

Or if you, if you check me on LinkedIn, you'll see that we have a podcast called a players on recruiting. You check it and you'll see, there is a lot of very interesting guests, lot of very interesting episodes [00:30:00] and including very famous guests. So I'll let you make your own decision, check me on LinkedIn and you'll find all the, all the links.

Julian: Awesome. Well, thank you so much, Robin. I really appreciate the conversation, especially learning from someone else who's, who's working in a similar capacity and I'm really excited to see where your, your company goes. And I'm definitely following you on LinkedIn. I love the tips and love the podcast that you run as well.

So thank you so much for being on the. 

Robin: Thanks for you Julian!.

Other interesting podcasts