Kshitij (also known as KJ) is a recruitment industry icon, recognized globally for his passion for solving the most difficult problems in the human capital ecosystem with technology. KJ founded Joveo with the mission of delivering the right “job to everyone” (a phrase that translated into the name of the company itself).Follow
On today’s episode of the RecruitingDaily Podcast, William Tincup speaks to Kshitij from Joveo about how the industry has failed adtech.
Some Conversation Highlights:
Listening time: 26 minutes
Enjoy the podcast?
Thanks for tuning in to this episode of The RecruitingDaily Podcast with William Tincup. Be sure to subscribe through your favorite platform.
Announcer: 00:01 This is RecruitingDaily’s Recruiting Live Podcast, where we look at the strategies behind the world’s best talent acquisition teams. We talk recruiting, sourcing, and talent acquisition. Each week we take one overcomplicated topic and break it down so that your three-year-old can understand it. Make sense? Are you ready to take your game to the next level? You’re at the right spot. You’re now entering the mind of a hustler. Here’s your host, William Tincup.
William Tincup: 00:35 Ladies and gentlemen, this is William Tincup and you’re listening to the RecruitingDaily podcast. Today, we have KJ on from Joveo, and the topic today is going to be fantastic. I’ve been waiting for this podcast for a long time. It’s doing better it’s, how the industry has failed ad tech. So we’re going to jump right into it. KJ, would you do us a favor and introduce both yourself and Joveo?
KJ: 00:57 Absolutely. Thanks, William. Thanks for having us here. Hi, everyone listening out there. I am the founder of Joveo, but prior to that, I founded the company acquired by Indeed. I’ve been in the space for about close to 20 plus years now, having been a part of the strongest ship in the town that time called Monster from 2001 to ’12, and I’m constantly intrigued, excited, disappointed, and a mix of emotions about the state of our industry and what we can do better and could not have asked for a better topic to talk with you, William. So, that’s-
William Tincup: 01:36 I’ve gotten in trouble for saying that HR tech in general is about 10 to 15 years behind sales or marketing technology. So I’ve gotten myself into some hot water because when someone shows me programmatic, or they show me something, the CRM solution, I’m like, “Yeah, Salesforce has been around for 25 years, so this isn’t new, it’s new to us.” And for a lot of practitioners, these things are new, but they’re not necessarily new in terms of, other industries are doing these and actually doing them rather well. So what do you see in ad tech? When you look at ad tech, you’re like, “Oh, wow. They’re 10 years ahead of us.”
KJ: 02:27 If you got in trouble for what you said in the past, then I think I’m in a very, very big trouble today. So I don’t think we are at a point to even start to compare ad tech with the state of recruitment advertising honestly. It’s not even Apples to Apples in the mind, I think it’s so far off. I think the right way to look at, and should’ve tried to bring an Apples to Apples comparison, think about ad tech industry as an adult and the adult was a child sometime. HR industry or the ad tech part of the HR tech industry today is in a child phase. So the best way to compare it is what was it like when ad tech was in the same similar stage and how it evolved from there? What are the evolutionary triggers that made it become what it became today? And therefore the conversation could become a lot more interesting or it can be, but if I were to answer that question succinctly, I would say, at least 10 years, if not more behind the ad tech space.
William Tincup: 03:36 Oh, yeah. Yeah. So let’s deal with it. I love the way that you phrase this, “evolutionary triggers.” So if we’re a child and again, we’re talking as adults and about a very specific part of our industry and the technologies that serve that. If we’re a child, what’s adolescence looks like?
KJ: 03:56 All right. Sure. So, let me to start with the definition of child in ad tech.
William Tincup: 04:01 Yeah, yeah.
KJ: 04:01 In the early 2000s, and then that will allow us to have a baseline of how we compare the current state of, I don’t know, what’s a better word for it, but ad tech and HR tech.
William Tincup: 04:12 Yeah. I think that’s the best way.
KJ: 04:15 So in the early 2000s, there was no measurement of the quality of traffic. It was always about my word is the truth. The source of truth was the supplier, or for that matter, the early stage platforms, they had absolute power. And absolute power, what does it do? We all know that. And there were all kinds of wrong things that were happening, the fraud, the bots, the foreign flex duplicates. And it was a place where you could not connect the dots from the original traffic to a lead getting concluded into a purchase, that’s what the consumer technology was all about, to make sure that a consumer is able to buy that product. And so there was no connecting the dots at all, there’s no tracking, it was just a Wild, Wild West.
Now, if you look at today, and I’m talking about three years back, I remember the day in the early stages of Joveo, when we would go to our customers and we would show that, “Hey, 20% of our traffic is bots.” That was our moment, they’re like, “Wow, we are paying 20% of our entire dollars for bots? The clicks that never existed.” And not for nobody’s fault, I’m not saying it is all intentional. A lot of bots are done to scrap somebody’s systems and jobs. But the payment was happening nevertheless, now foreign clicks, duplicate clicks. And I can keep on going on about some of these kind of clicks. And now, as you could imagine, at that point of time, there was no down the funnel tracking. Nobody knew, if I’m paying X dollars for a certain media, how many of the clicks that I generated result into down the stream funnel conversions, whether it is a shortlist or into your hire. And when that measurement does not exist, it is very hard to improve your buying decisions, right?
You may be buying from someone and I heard, and I’m not talking about too long back, about three years back, a customer of ours, a large hospital chain in the US showed that they spend almost $50,000 in the ballpark on two different job boards having made one hire. Only when we showed them the down the street conversions and the other job board there is spending $10,000 or $20,000 because the cost per click was more expensive, gave them an average of $500, $600 a hire. So it’s only then that this industry started trying to say, “Let’s connect the dots downstream.” Now the next stage, and which is perhaps the defining moment of ad tech stage was the implosion state.
In the stage one, there were massive number of accusations in ad tech space. I can literally count 10 in one fingers. And this implosion is also, we can call it a realization stage, where one of the first companies in ad tech space was a DSP called Rocket Fuel, they went public. If I’m not wrong, it was like a $3 billion to $5 billion valuation. And as soon as they went public, they had to clearly talk about their margins. And the margins were in such, there were 70%, there were 60%, 70%, 80%. In that phrase, that customer said, “Take a break. You said you were a platform, but how come you’re making that much margin? That means what we are paying for is not what we are getting. There’s a lot going in your pockets.” And they’re like, “Hey, you are not a platform, you’re a middleman.” There was a line between the fair and unfair that was getting lost. And people started questioning that, there was a lot of realization.
People started asking for, “Hey, we are not going to have ad tech project. We are not going to have a completely opaque or blind networks, we want some kind of a transparency work into it. People started questioning a lot of sub-syndication that was still happening out there. And in that stage, when the hands of the suppliers and DSPs were being forced to change, the sophistication of the wrongdoing in that industry became much better. That means the bad things were still happening, but the bad actors became a lot more sophisticated, so good and bad traffic was getting mixed and adulterated. And that was a time we started seeing the retargeting and the use of the word like data science and machine learning and AI. But in reality, it was nothing but retargeting, it was a dumb retargeting to say the least. It is not even getting better at it.
William Tincup: 08:59 And now retargeting was just basically to track clicks?
KJ: 09:02 Correct.
William Tincup: 09:03 Yeah. Okay.
KJ: 09:04 And that was late 2000s. And now I would say we are more or less in the A and B stage, moving from A to B. And in some ways we are in the B phase, some ways we are still getting out of A to B. During the B phase to C phase, there was a conversation about standardizing the definitions of, what does the quality mean? The relevance of the traffic. Back in that ecosystem, the suppliers and the real buyers, which were the large giants like Nike, and GMS and those of the world, they came on the table together and said, “Let’s agree on the right definition of track.” Google and Facebook in the early 2010s became very, very big, so that also for the hand. There were platforms that came up as a clear SaaS and managed services platforms, right? It is no more as, “We are the source as well as a platform.” The exchanges, the ad networks evolved.
And during the time, you start seeing emergence of the industry into the segments of the industry, the sub segments of the industry, like the supply side platforms, the ad exchanges, DSPs, data management platforms, as were trade desks, which are like managed services divisions. So there was very clear distinction of what I do versus what you do. And user cohorts were being made. Audience creation, behavioral patterns were starting to be seen. And then we come to the late 2010s, which is pretty much now, what’s going on is just industry is going to a whole different level. I think it’s gone so beyond, it’s about head bidding. It’s about Calculus tracking. It’s about, I won’t even know the PII, the GDPS come in, the regulations are tightening the ship around how things were done.
So the industry has gone through a cycle. And now in a phase of reinventing itself to see how we can do personalized targeting without knowing the person and how we can use fingerprinting technology to kind of forward that case. So if I were to call it this way, our industry is still stuck because most of the players are not talking in transparency out there. People are not able to see source, there’s no control on sub-syndication.
I can see the same job coming to us from at least 20 different sources, but the employer is still the same because somebody is sending a job to someone X, and the X is sending a job to Y, Y is sending the job to Z, and they’re all buying from each other. So the user experience sucks as a result of it, and I’m sorry for my language, but-
William Tincup: 11:49 No, you’re good.
KJ: 11:49 But it is what it is, right? So I think we are at a point where we are not truly technology first still, as much as we may say, I’m going to make a bold statement here, which I hope I don’t stand to regret later is, there’s nothing programmatic, but programmatic advertising today. It is all fixed bids, there’s nothing real time. If you think about it, suppliers of the ad tech system openly share their data, their traffic data, their data of every single page with ad exchanges and data management platforms to allow for real time bidding. In our industry, no supplier ever wants to do that because they’re all afraid and fearful that, “I’m exposing my data.”
William Tincup: 12:35 That’s right.
KJ: 12:37 And one thing that is very interesting that is happening is, this juxta position of the most sophisticated buyers who are actually dealing with our tech space, the likes of the largest staffing companies in the world, the likes of the largest gig companies out there in the world, the Ubers of the world, those are the players who are saying, “Hey, we know what entire ad tech is. We are the gurus of that.” And they are actually trying to for the hand of this industry to change. So I have a feeling that this industry that we have right now is not going to stay forever behind. They will be a generational leap that will happen assured by the big buyers in the industry who are used to playing in the ad tech space like it is the 2020s, not 2010s. Sorry, if it’s too much.
William Tincup: 13:27 No, no, no. First of all, I’ve got like 40 questions. So first thing is, if we don’t call programmatic programmatic right now, which is exactly correct, what is it that we should call it? And then-
KJ: 13:46 I think-
William Tincup: 13:46 Yeah, go ahead.
KJ: 13:48 Yeah. I think we should call it simple pay per performance today. At least what we have done is, and I was talking to a friend of mine who has built a first, truly pay per performance job board in Europe. And he is the CEO of this company. And he is like, “Hey, it’s truly pay per performance only when you track down the funnel,” because at least you know what’s converting. Programmatic is when machines are making real time decisions, and that isn’t happening because still the suppliers play ball, it would happen.
William Tincup: 14:20 And pay per performance isn’t a bad thing. I mean, that’s not a bad word, that’s not a bad phrase. These aren’t bad things, it’s misleading if we call it programmatic because it’s not truly programmatic, yet.
KJ: 14:34 Correct.
William Tincup: 14:34 It will be, it’s coming. But I wanted to back up for a second, because I think when people listen to this, they’re going to be like, “Why are we talking about ad tech?” Okay. On some level, one of the things we’re trying to solve for is source of hire, what’s working? Where do candidates actually come from without making some assumptions? Do I have the algebra right?
KJ: 15:02 Yes. Yes. I know what you’re saying. It’s a attribution problem, the classic attribution problem, the first click, the last click, the middle clicks, right? So that actually is a very, very good question. And the reason I say that is because the origin of Joveo was trying to solve that problem only to realize that the job boards will never share the rate of user clicks. So I would never know who went to… The first click happened on, let’s say Indeed. And the last click may have happened on LinkedIn, and the middle clicks could have happened on Stack Overflow and some of those sites.
That engineer might have clicked eventually through LinkedIn, but then the first click has a lot of value because the job discovery happened on the first click. The user might have gone to LinkedIn to research the job and eventually click on that job. And without the first click, without searching to Indeed, the journey of the user would not have even started. So the point I’m trying to make here is, the source of hire, if you were to make it as a last click, that I think the industry has largely, I wouldn’t say solve, but at least are aware enough to want to solve it. They see the value of it.
But I don’t think the first click, even though people are talking about it and they’re… I’ve seen a lot of smart, really, really astute, top of the shelf, talent acquisition leaders who still think about it and want to solve it, but they have to work, there has to be forum where the biggest buyers in the world, the direct enterprise employers and biggest suppliers have to sit on a seat of table and say, “Let’s create an industry in IAB, let’s create an industry’s data exchange or data platform where all of you will share your data anonymously, and then everybody can see the user journey.” Otherwise, it’s just not going to be possible.
William Tincup: 16:52 Right. And what’s interesting is it further complicates once people make that application. I mean, first of all, just to know the data of source of hire to application would be great. Okay. So if we add total clarity into what’s working and the different steps would be great, but it would also be great if it continued through the ATS process and all the way into onboarding so that closed loop, we could actually know source of hire all the way to onboarding and future terms we’d want to know into performance, like what actually worked both and how to find those candidates, but also what worked in the organization.
So I wanted to ask the question of, people now, when they listen to this podcast are going to feel like, “Okay, I get it, I now understand.” What should they be asking what we would call programmatic performer or programing performers. But now let’s just say the pay per performance and just stick with it. For all the vendors that are in that space, what are questions they say should be asking of those vendors?
KJ: 18:03 Oh God. Okay. I’ll make it very simple answer. If I was the buyer and I would have four companies coming to me and pitching for it, I’ll just ask a simple question, “Do an audit of my programmatic media buy in and come and tell me where I’m doing a bad job.” Now, what I mean by that is, every single job I’m advertising out there is out there, it’s in the public view. Tell me where I’m sourcing the jobs wrong, whether I’m posting the wrong job board, or I’m posting the wrong locations or my content is wrong. And come to me and advise me not just about what media buy in should I do, or can I get it cheaper? Not that kind of predictive undertake, which I think is basically impossible to predict since we are not in the real time world.
Everybody can say, “But if tomorrow a large job bot changes algorithm, there’s no predictive media buying.” But see, there’s a lot of insight that exists about job seeker behavior. Let’s get deeper into that at a company level. Everybody talks about at a industry level, at a category level. Even I call that useless, get specific to a company standard level and say, “Okay. All right, this is the company you’re losing most of your job seekers to, this competitor,” or, “This is where you typically have a good brand recall in this state. And this is where you’re gaining your people from, and let’s go and target those people.” And then bring out strategies as an evolution of what the company is currently doing.
The other thing I can ask about is, what kind of a transparency are you giving? Can I see every single source in your ecosystem? Not just a blind exchange, every single source. The third question I’ll ask is, do you have an ability to stop sub-syndication? Because I don’t want to compete with myself. Now, I’m going to call it out. A large company, and I hope it doesn’t go the wrong way, but Amazon pays 3X to 5X because it ends up competing with itself. Different divisions have the similar jobs going to the same location and they’re trying to orbit each other. And that has, I think the bad part of this industry, stop the sub-syndication, control it. One of our customers did it, and they saw about 40% decrease in cost per hire, just by removing that sub-syndication part of it.
And as you go further and further into it, we should ask questions that, do you really, really know what time of the day, or what day of the week is the technology tuned to accordingly change campaigns and spend according to that? Could you control a job level budget so that the hard to fill jobs are the ones which actually get the clicks? Is your system completely automated, where based on sequence of complicated rules, or even for that matter, autobot as they call it. Right now, autobots in the industry are nothing better set of rules. And people are trying to fool everybody else saying that, “We have an autobot,” there’s nothing like that. It’s set of rules and a concoction of that creates it. But at least show me your proof of that. Show me through logs that if you are changing bids, it is done for a reason. Demand proof for everything that you’re getting. Sorry, if I just went on.
William Tincup: 21:22 No, no, it’s wonderful because I’ve always thought of programmatic pay per performance in the sense of here’s it helps you with the where and how much. So if you don’t know, if let’s say you’re new to the job and it’s a new job for you, you have no idea what you should set aside. So that’s the, how much, and then where should that be put? Pay per performance programmatic can help you if that’s start you on that journey. What you’ve just added is another layer, which I think is beautiful is, when. So it’s the, where, when, how much, probably even goes further and deeper than that, but this idea… Because we can do this in social. We know when people are going to be on Twitter, we know this in Facebook, we know this in Instagram or TikTok, et cetera. Why don’t we know this about full stack developers?
KJ: 22:18 Absolutely.
William Tincup: 22:19 They’re looking at jobs at a particular time on whatever site that they’re looking at. So I love the fact that we’re thinking about when. Let’s deal with kind of the next things that we obviously… We went from child to adolescence and in an adult and understanding, what do you see just this year, maybe even into next year is just kind of the next evolutionary step for us.
KJ: 22:51 Next evolutionary step. I can say, I can hope this would happen, I’m nobody to predict. But I can really hope that the consortium of the biggest data employees in the world and the biggest suppliers in the world and the biggest platforms in the world will come together and say, “We agree with one standard rule.” Because it’s an easy win, it’s a low hanging fruit. Much more for which I hope happens, but I doubt it will happen because I’m being realistic. I’m not being pessimistic, 90% cases, I’m a die hard optimist but in this case, I can say, let me be realistic. Is, open up your traffic insights and data and share, anonymize your clicks and share it with each other so that we all know the user journey. We all know the first click and the last click.
We can all, therefore the platforms can work with ad service to enable real time bidding, then the best job would win. Then the best job seeker eventually win, that is what it is all about. So that I hope will happen next three years or four years or five years, maybe. The third thing I would say that I hope happens in industry is, and this is, I have seen that about 10% to 20% of the talent acquisition professions are extremely astute. They’re constantly investing in learning, building their tech stacks ahead of the curve. But there is a lot of work that is required in the general recruitment marketing world to try and learn about how ad tech is because the real revolution will come from them. Couple of people holding the banner up will not change the country, it has to the masses that have to rise up and say, “I’m going to change this company,” or, “I’m going to change this country, and I’m going to change this industry.” And everybody has to start demanding it, the revolution will come from suppliers trying to get a revolution going, or the buyers.
It’s been 20 years, 25 years, suppliers have been there in our industry and the revolution hasn’t happened. I think the buyers should get educated and somebody should take it upon them. And maybe a third party industry forum like this industry’s IAB or for that matter, something else that would undertake the effort, show that from these places should bring very deep down recruitment marketing courses, which will educate the masses on this.
William Tincup: 25:19 Yeah, because as the audience is listening to this the incentive for job boards to get together and to create standards, isn’t there for obvious reasons, they’re going to have to show traffic and they don’t want to show traffic. But if the pay per performance programmatic, all the large branding agencies, et cetera, if they all kind of got together and basically demanded standards, then standards would then exist.
KJ: 25:52 Correct.
William Tincup: 25:53 I love it. This has been fantastic, KJ. Thank you so much for coming on the show. And also thank you for educating us with ad tech.
KJ: 26:01 Thank you so much, William. Thanks for having me today. And I’ve always enjoyed the conversations with you. So I appreciate that.
William Tincup: 26:07 Vice versa. And thanks for everyone listening to RecruitingDaily podcast, until next time.
Announcer: 26:13 You’ve been listening to the Recruiting Live Podcast by RecruitingDaily. Check out the latest industry podcast, webinars, articles, and news at recruit-
The RecruitingDaily Podcast
William is the President & Editor-at-Large of RecruitingDaily. At the intersection of HR and technology, he’s a writer, speaker, advisor, consultant, investor, storyteller & teacher. He's been writing about HR and Recruiting related issues for longer than he cares to disclose. William serves on the Board of Advisors / Board of Directors for 20+ HR technology startups. William is a graduate of the University of Alabama at Birmingham with a BA in Art History. He also earned an MA in American Indian Studies from the University of Arizona and an MBA from Case Western Reserve University.
Please log in to post comments.Login