Joveo's founder and CEO Kshitij Jain (aka KJ) turns the tables asking some profound questions on Chad & Cheese.
What is this programmatic magic you speak of?
Data point + volume = accuracy
Focus on the 80/20, not the 20/80.
The process is irrelevant, it's the outcome that matters
Programmatic distribution vs. Programmatic targeting
and so much more...
PODCAST TRANSCRIPTION sponsored by:
Voices Intro (0s):
Voices. We hear them every day. Some voices like mine, smooth and confident. While on the other hand, the Chad and Cheese podcast is like listening to a Nickelback album. You'd rather stab yourself in the ears with an ice pick. Anyway, y'all now listening to Voices, a podcast series from Chad and Cheese that features the most important and influential voices within the recruitment industry.
Voices Intro (31s):
Try not to fuck it up, boys.
KJ (35s):
We didn't do any rocket science about a year and a half back on that. It was easy.
Intro (39s):
Hide your kids! Lock the doors! You're listening to HRS 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.
Chad (1m 9s):
Welcome back. We're continuing the conversation with a veteran of the recruitment tech industry Kshitij Jain aka KJ, founder and CEO of programmatic platform Joveo. Enjoy! Back in the days when SEO was like magic, it was like wizardry, right?
Joel (1m 32s):
And witchcraft.
Chad (1m 33s):
Nobody knew what the hell it was. And that's why they really didn't do it. In most cases, I knew it was really slow to adopt the same thing I'm feeling is happening with programmatic because people just can't understand what programmatic is. I know that if I go to Monster.com and I post an ad, or I go to Job Board, a College Recruiter and I post an ad, or what have you. I know my ads there. And I don't know it's there for 30 days. And I know what I know the thing is you're asking for kind of like a trust value volume that's hap happening here that you just don't get what talent acquisition ...
Chad (2m 11s):
how do you get beyond that? To be able to help them understand that there is this magic wisdom called Joveo that will help them better target. And how many data points do you guys actually go against?
KJ (2m 23s):
Two questions in this fight? I'll start with the second one and then go to the first one. So, data points. We focus on on one side that your quality is defined by you like your hires to find the quality. Of course not hires could mean that you want to have a lifetime value. Sometimes, a person who you hired stayed for three months versus the person who stayed for three years is a totally different ball game. So what are the line you draw? Right? So, so getting to that point is equal to do the match, right? So you define the quality.
KJ (2m 54s):
Once you define the quality, then the match has to happen between the job and the candidate. And I always look at the world. If you look at it noise, literally if you take it off three or four data points, you get the quality of the 70 to 80%. Is it 20/80 rule, right? 20% data points make an 80% impact on reduction. Reducing the, the lack of relevance. I'm not called quality for example, right? And I'm going to start with a very small thing. Like if you are looking for a nurse, let's say an ICU nurse in New York City.
KJ (3m 27s):
Now you're not looking for nurse from Philadelphia or Boston or Phoenix or, or like San Francisco for that matter, right. Even though the nurse will be completely relevant in terms of our quality applicant, it's not relevant. Or if you would get, if you were to get a staff nurse or a travel nurse in New York City, that is just a title location for the jobs where the title itself means the resume. The factor, itself signifies that if you're working as a nurse, you have a certain qualification, certain certification that you have.
KJ (3m 59s):
Get these two things Right? Right. So these are the two data points. Let's start out saying that. Let's get that right first. Right? If you were to take baby steps, can you get into me tomorrow? And I'm talking to the practitioners in the space and everybody out there is that at least your job is being shown to the right person in the right location with the right title. Right. That's how, hard? Difficult is that's two data points for you? Get that right! Right? Now, let's go further down the line right now, you're talking about key words. You're talking about latest job descriptions. You talk about the meaning of words, right? A business urban professional is closer to a sales manager, then a marketing manager, and then as compared to a business analyst, right?
KJ (4m 33s):
That, that relationship of a job with each other, which is more relevant. You're talking about. We look at about 30 odd data points on the job description side. Because we know the quality, right? The resume is already there. You can go deeper and deeper. It can be like as many data points, but I would say about 10 to 15 data points in resumes are good enough. It's easier to solve 90% of the problem. It's like in my mind, I would say 20% effort is required to solve 80% of the problem. 80% of the effort is required to solve 20% of the problem.
KJ (5m 4s):
So if you, if you were to look at just these 30 to 40 data points, I think you've solved 80% of the problem. And this is, these are the data points you look at. When we, you know, when we have a higher data, we have the application data. We have the, of course we have the job data, right? We have deep job data, right? That itself solves a lot of these problems. In fact, I remember there was a customer for us, which is which, and this is one of the largest companies. Like it's a company that has more than $20 billion in revenues. And in our space. They came in and said, okay, we tried everything.
KJ (5m 35s):
It ain't working. It is working. We've got probably 25%, 30% better results. You guys can do any magic on this? I said, I don't know. Maybe you did everything. All we did is just two data points. My friends, we basically said, okay, a certain job description of a nurse can be done as intensive care unit nurse, or ICU nurse, or intensive unit nurse or whatever it is, right? Or they have their own terminology just knowing that all of the jobs belonged together. And the campaign is run for that one single title, which basically all these titles meant that how can I assume those job going to the same campaign as an accountant that is either you're getting a right click for nurse and a bad for the accountant or the other way around that's doing that.
KJ (6m 16s):
Improve the results overnight. But 25%, I'm sorry. I'm not, we didn't do any rocket science about a year and a half back on that. It was easy. It was just again, common sense. Yeah. Few data points. Do a good job of it. We do about 40 data points right now.
Chad (6m 29s):
Okay. Okay. So one thing that has always driven my conversation when I'm, whenever I've been in sales has been outcomes, right? It's all about outcomes and that's what the organization should be looking for. And during this entire conversation, you've talked about hiring outcomes from the onset, right? So how hard is it to actually engage some of these big brands in the conversation of, instead of just eyeballs, we're talking about hiring outcomes, it seems to be a pretty easy transition.
KJ (7m 6s):
So how hard it is to explain this concept. And when, I go on customer calls with my team, so one thing the sales team talk about is holding back, right? He's going to, he's going to go so far ahead in terms of talking about tech and the message has to be simple outcome driven, and I think you hit it right there, right? Define the outcome, right? If you can define the outcome, people don't really, truly need to know the processes and the nuts and the bolts and how that works.
KJ (7m 38s):
Right. Listen to their needs and say, yes, we can deliver it and prove it. Once that is there, then you can start talking about what are the nuances of programmatic that, Oh my God, this, this has to happen in programmatic. And are you bidding it tight? I'm sorry, but bidding it right as good for maybe a person who is a creator. Right. I don't think our common manager even understand what bidding means right now? Now I started hearing people talk about real time bidding and dynamic bidding. I get it. I get it. I understand that. So I think the idea is to be able to always work with the talent acquisition professionals and address the needs.
KJ (8m 14s):
I had a customer which said that, Oh my God, I don't know how this works? And they're like massive customers. I want to see and know that your system knows how to do it on their own. And like, we do do that on our own, but we don't have to show it. They just need to know this. So, so, so to be able to, there's a, still a lot of learning curve. And as you said, right, a talent acquisition and HR adoption is the key. I think we need to do a better job of even communicating that message to the CEOs of the company.
KJ (8m 45s):
I bet you call any CEO in the world of large corporation. Getting the right talent is among the top three.
Joel (-):
For sure.
KJ (8m 51s):
So I believe if you, if you bought from the right and seeing the right things, you will get heard. I just think that we try to over complicate the messaging and industry.
Joel (9m 0s):
I want to back up a little bit and I love, I love Chad's example of SEO and, and being able to understand that, I mean, we're still, we're still making the joke of most employers can't even spell SEO, which still gets laughs, which is great. But for those who listeners out there don't know what programmatic is. So I want to like, just go to the basics of that. And then I want to know why Joveo was different or how any of these programmatic solutions differentiate themselves. Is it just the distribution, the technology, or both?
Joel (9m 32s):
And maybe really quickly Chris Forman, founder of App Cast, one of the first programmatic solutions out there said that 40%, this is a year or so ago that 40% of job postings were done programmatically. Is that true? You, if you're talking about distribution yes. Yes. Of course. If you are sending a feed and putting a job out there and without having to post a job, perhaps yes, that is true. Okay. So let's so let's, so let's define it. And then how is it different per like provider?
KJ (10m 3s):
So let's say for your large company or General Motors, or you're a Pepsi, or you're a T-Mobile right? You would not want to distribute jobs manually. Each of the job boards, you post it to, right? A you'd want a way to kind of post a job and kind of have data to back up that this particular job should go to this job board right? So that is a program or machine deciding which job to post to. So there are two levels, right? One is just the job distribution. So if you say that, that BroadBean and Equest on the job were programmatically posting jobs.
KJ (10m 36s):
If you say machine was posting it, that is true. Right. But, but is it programmatically optimized to post? It's a different conversation. I wonder how many jobs in the world are actually optimized to be posted programmatically. I'll be still spraying the jobs, just saying that machine is doing it. I think that data has to be looked at. How many jobs are actually being chosen to be posted to a particular source based on data is the question that I think I don't have the answer to. I think we need to find out the jobs that we do. We do know that we try our best to use our metadata across all of customers in terms of apart from the job board for this location, for this title has a highest likelihood of delivering a hire or even a most relevant applicant.
KJ (11m 19s):
And so we use that insights and trying to optimize the programmatic distribution. And to that extent, yes, there are jobs getting pushed like that, but I don't really know how many of these jobs out in the world are getting posted like that because enterprises represent only an audience, right? SMB customer, right. U S is known to be a place where small and medium enterprises have created this country what it is as much as large enterprises have. I don't think they even know the PO programmatic. I don't think they even post anything in the programmatic fashion.
KJ (11m 51s):
So we don't really know that data. And I, at least in my mind, I don't know that data right enough through programmatic
Chad (11m 59s):
So what makes Joveo different from everybody else at this point?
KJ (12m 2s):
Sure. So we started out with the philosophy, right? We said, okay, we'll not be, this will not be this right? We'll never be a job board. We'll never be a staffing agency. We will be married deep with the concept of putting the right job in front of everyone. And then for that, we have to focus on getting the entire, we call it the "breadth and the depth of data" Breadth is every single source. Right? So now if you stay true to that, one of the very first things we did is we gave the power in the hands of traders or advanced users who actually have what social search, slots, postings, a PPC, the entire breadth of it.
KJ (12m 42s):
They have everything, all the data sources, even the career site and native traffic, organic traffic that can be distributed and tracked at the job level. Now that's where I'm coming to the depth of data, right? We can actually, for that matter, right? If you have a single ad on Google AdWords, we can actually attribute that ad for that location, for that particular job. And how many clicks you got for that in the system itself. And then the depth of right? Fact to hire it should be built into the system. It should not be you charge extra for that, or is it is the depth of data because only when you get the higher data, which we actually call is the intelligence, right.
KJ (13m 19s):
That intelligence then feeds back the sourcing strategy. And that is what makes us different, right? That's the feedback loop, the constant learning, the constant investment, right? We have a team of a program engineering team is about 65 people right now. What do they do all day long? Right? They're a lot more behind the scenes, since we're launching new products. We only do one thing to do programmatic, but all that engineering talent and strength and the engineers that came from the likes of Googles and Amazons and Microsofts of the world, all they're doing is trying to get this better.
KJ (13m 49s):
Everything is driven to the mission of three to one. Can we get to a point where companies can get three clicks, two applies, and one hire without having to do it themselves?
Applause (-):
Clapping
Outro (14m 2s):
Look for more episodes of Voices. This Chad and Cheese podcast series devoted the stories and opinions of industry leaders, subscribe on iTunes, Google podcast, or wherever you get your podcasts. So you don't miss a single show for more visit chadcheese.com.