Use Case Podcast – Storytelling about Celential with Brian Hollinger

Announcer: Welcome to RecruitingDaily’s Use Case Podcast, a show dedicated to the storytelling that happens, or should happen when practitioners purchase technology. Each episode is designed to inspire new ways and ideas to make your business better as we speak with the brightest minds in recruitment and HR tech, that’s what we do. Here’s your host, William Tincup.

Tincup: 00:02 Ladies and gentlemen, this is William Tincup and you’re listening to the Use Case podcast. Today, we have Brian on from Celential and we’ll be talking about the use case or the business case for why his prospects and customers pick Celential. Let’s just jump into some introductions. Brian, would you do us a favor and introduce both of yourself and Celential?

Brian: 00:23 Thank you, William. Yes, Brian Hollinger, VP of Sales here at Celential. We focus on optimizing and automating the top of funnel for our customers, leveraging AI, specifically natural language processing, and our proprietary talent graph.

Tincup: 00:38 I love that. For anyone that might not understand top of funnel, let’s do some definitions. These are people that have applied, and now, there’s a process in terms of trying to take them from application all the way to offer letter.

Brian: 00:57 It’s a little bit different. Essentially, a top of funnel is the people that are not applying. I guess there are two inputs to the top of funnel. One would be the people that do apply, but then recruiting teams and talent teams will often go outbound in sales. We call it outbound prospecting. In recruiting, it’s very much a similar motion going outbound, trying to identify the right candidates through, we’ll call, status quo channels. The biggest status quo channel would be LinkedIn. Many talent teams… LinkedIn, as everyone knows, multi-billion dollar business, and their flagship product is the LinkedIn Recruiter product, which is I believe tens of thousands of dollars per seat. We hear stories of people spending hundreds of thousands of dollars with LinkedIn, and it’s essentially a prospecting tool for them to identify candidates.

01:54 We come in and deliver a solution that’s kind of going a little bit beyond status quo, and just looking to identify candidates outside of those traditional methods through our talent graph, which takes into consideration hundreds of signals, thousands of signals across the web to identify really who are the best candidates, and then engage those candidates to fill the top of funnel. We do that in an automated way, reaching out to those prospects or the perspective candidates, I should say, on behalf of our customers. That’s what we’re delivering on.

Tincup: 02:30 I love that. How do we understand what a good… I love that you and I can talk sales and marketing language, which is great because it’s an easy transfer over to candidates and candidate language. How do we know it’s a good candidate fit? What do we need to do and what does the technology do to understand, “Okay, this would be a good candidate fit, as opposed to someone that maybe might not be a good fit”?

Brian: 02:58 That is a great question because it’s not just about the candidate, right? The candidate has all of its attributes, you could say, the technologies they’re familiar with, the companies that they’ve worked at, the experience that they have, as far as individual contributor, leadership. All of those different things kind of make up the candidate, but there’s also the company side of the equation, right? As with anything, I actually make the sales and recruiting analogy all the time, but I say recruiting’s harder because you need two people. You need two sides to say yes, whereas sales, you just need one side to say yes, and you can kind of move forward. So, the key aspect to that is that it needs to be a fit for the company as well.

03:45 What we’re able to do is we’re able to understand the candidate’s makeup and profile with a great depth of understanding of their technical skills and depth of experience with specific programming languages and infrastructure technologies, combined with the company and the company attributes, which is also a part of our graph. So, we have an understanding of all of the engineers at our customers in our talent graph. We understand their makeup, their experience, the companies that they’ve worked for, the level of expertise they have. We understand the company itself and how many employees are there. Is it series A, series B? How fast is it growing? All of those things make up the company profile, and then we’re able to understand, “Okay, who are the best fits from a technical perspective, from a skills perspective, but then also, who are the best fits as far as likely to work at a company at maybe series A or series B, or only a team of 10 engineers?” Someone that’s worked in only companies of several hundred engineers may not be a fit for a team of 15 engineers that’s maybe growing quickly.

04:59 So, really understanding the company attributes, combined with the candidate skills, as well as that experience that they have, really those are the factors that go into making the match, in addition to understanding the depths of skills and understanding, which are very important of course. But it’s truly understanding what’s a good match because if you say like, “Look, this is a seed stage startup that’s hiring its first engineer,” it doesn’t make sense to go after, “Oh, I only want A plus engineers that have been at Google for 10 years.” Well, someone that’s been at Google for 10 years, that’s going to be a really hard sell, and for lack of a better term, you’re going to be barking up the wrong tree. So, who are the best fits that maybe have had that experience and maybe started at a company, series A, series B, and been with a company through to series C or D and kind of been part of that growth story? Those are the types of things that we’re able to do at scale that…

06:02 Really good recruiters do this. The best recruiters understand all of these elements that go into understanding who are great fits, who are great candidates, but doing it at scale is incredibly hard. It’s impossible, you could say, because it’s not just identifying the candidates, but the second part of it is engaging the candidates. So, what we do for our customers is that after we’ve identified the best fits for our customers, we then reach out to those customers on behalf of a member of the team, typically a hiring manager or CTO or a member of the leadership team. So, that it’s in an automated yet still personalized way that is able to say like, “Look, because of your experience working with this technology at this company, and actually a few members of our team might have went to the same school as you did, about the same time you did, we think you might be a good fit. Are you open to a discussion?”

07:02 So, being able to generate those messages and generate that outreach in an automated way, coupled with the identifying of the candidates in an automated way just allow us to really move quickly and move at scale for our customers where… Like I said, really good recruiters could do what our technology does, but it would take them much, much, much longer, and they wouldn’t be able to do it at nearly the scale that we’re able to deliver on.

Tincup: 07:32 Two dumb questions alerts. One is, is the data that you mine from more than just LinkedIn? Yeah?

Brian: 07:41 Oh, it’s really anything that’s in the public sphere. We’re able to… Part of our secret sauce is taking things like a personal website or taking a college dissertation, and tying that back to the individual in a way that is accurate. There’s millions of signals, billions of signals across the internet, across all of these candidates. And then, it’s truly taking that signal, and applying it and associating it with the individual, all of these people. To make up the graph is very, very difficult to do and something that we are able to do. Like I say, it’s taking just things like Stack Overflow, GitHub contributions, personal websites, college dissertations, academic research papers that might have been published. All of those things, we’re able to bring back to the individual, and then leverage that to really identify the best candidates and make the best matches, as well as enrich the messaging, right? Because we want to be able to engage those candidates in a way that’s meaningful and personal, and all of that information feeds into that ability to do so.

Tincup: 08:55 The second dumb question was, do people… I, first of all, despise software categories, but I also understand that HR and recruiting’s budgets are built in Excel. So, I would’ve put you in kind of that smart sourcing technology category, but again, because I hate categories, I want you to tell me what your prospects and customers… How do they view you?

Brian: 09:18 Typically, we replace traditional external agencies and recruiters. Oftentimes, our customers, they’ll have talent teams themselves, but talent teams typically for roles that they may be constantly hiring for or specialty roles or maybe they’re just understaffed on the talent side, will leverage agencies and recruiters. So, typically, if you were to categorize for lack of word that you say, “Budgets are made in Excel. What column or what row in Excel do we typically dip into?” Let’s reallocate budget that we would’ve used for external agencies and recruiters, and then apply it to Celential because the economics of external agencies and recruiters for software engineers, technology hires, which we specialize in is incredibly expensive.

10:18 So, a typical technology hire, software engineer, member of the technical staff is going to be between $150,000 and $200,000. And then, a typical agency is going to be 20%, 25%, 30% of that first year salary. So, you’re looking anywhere between $40,000, $50,000, $60,000 for one hire. They’re able to spin up these agencies and recruiters at no cost, but then, if you make one hire, it becomes prohibitively expensive. Where we’re able to come in and deliver a consistent quality pipeline of candidates typically between 20 and 40 candidates per month, and our customers typically would need about 20 to 30 candidates to make one hire, the economics make sense pretty quickly for our customers, and that’s typically why they come on board with us.

Tincup: 11:13 I’m so glad I asked the dumb question because I had you miscategorized in the sense of, I had you as a LinkedIn killer, and maybe even closer in my head over to something like SeekOut or hireEZ in that way that you can help people find talent, but this is far more robust in the sense of you can actually displace some of that really mismanaged staffing funds that have been there. We all know that the situation that’s been going over there, money is being spent, and they’re not necessarily getting their best return on their investment. This is a much better way to get a return on that investment, so I’m glad I asked the dumb question.

Brian: 11:59 Yeah. I think that’s part of it. I think that another part of it too is when you talk about other technologies, that although those other technologies still need to be run by people, and those people need to be hyperefficient, and those people need to know how to use the tool and use it appropriately, so no matter how good the tool is, if the person that’s using it… We all know the phrase user error. If the person isn’t totally dialed into that technology, maybe they’re using it wrong and there’s some inefficiencies there and some slack in the system there, whereas we’re coming and delivering. Our customers aren’t necessarily logging into a tool and doing the searches and doing the messaging. It’s all happening in an automated way. Our customers don’t need to change anything about their process for candidate experience and only just engage with the candidates that respond positively, and say, “Yes, I’m interested in a 30-minute call. Yes, I’m interested in learning more. Yes, I’m interested in kicking off the process with you.”

12:59 There’s no tool that the customer has to spin up there, so we hear this from our customers all the time and prospects, the concept of tool fatigue. “Well, we tried using this tool and the team didn’t adopt it,” or, “We subscribe to this technology and we used it for two months, and then it just sat there for the next 10 months.” None of that happens with our customers simply because they’re only subscribing to the pipeline of candidates, and there isn’t necessarily anything for them to miss or underutilize or use wrong with us.

Tincup: 13:30 I love it. Let’s pivot to the other side real quick because I want to talk about prospects. First thing is, let’s go into the demo. I call it the aha moment, but what’s your favorite moment in the demo?

Brian: 13:48 Part of our demo process is kind of exposing the customers to the matching process, and because there isn’t necessarily anything for our customers to use, our demo process is the free matches, and then taking them through what the process looks like and how we deliver on those candidates. We liken the matches that anyone can go to Celential.ai and run a free match against one of their job descriptions, and the aha moment comes kind of one of two ways. Number one is sometimes customers or prospects will say, “Oh, actually we interviewed this person two months ago, and maybe they declined our offer, or maybe upon further review, they weren’t necessarily the perfect fit that we were looking for.” To us, that’s incredible validation. Out of the tens of millions of people in our talent graph across the entire Americas, North America and South America, we identified these five people because we deliver five matches. And for a prospect to say, “Actually, we talked to this person,” we feel that’s incredibly validating and they kind of have an aha moment there.

15:04 The flip side to that coin is when we present candidates as part of that match exercise, and they say, “You know what? I probably never would’ve found this person on my own, but now, I’m taking a little bit deeper dive in you suggesting this candidate as a match, and I think actually, this could be a person that would be very interesting to us.” It’s kind of, “We already talked with this person, and wow, out of the tens of millions of people, you found one of them. Very impressive.” And then, the other part is, “We would’ve never found this person otherwise.” That’s, to me, where I get the most satisfaction because…

15:43 And then, once prospects ultimately become customers, that happens frequently. Because, like I say, customers are using status quo solutions, we’ll say LinkedIn. LinkedIn data is so sparse. I mean, you can have people that have spent five and six years at… My favorite example is Stripe. Someone spent five years at Stripe. What kind of work have they done at Stripe for the past five years? It’s got to be the most incredible experience over these last five years at a company like that, but on their LinkedIn profile it says, “Software engineer, Stripe, five years.” Oh, my gosh, what are all the details there?

16:17 We’re able to paint a much more contextual picture of a candidate’s experience, and then deliver that to our customers as part of delivering the candidates over to them after they’ve responded positively. And then, we have those customers that say like, “Look, I would’ve never found this person, and all of my search queries across all of these different tools would’ve never identified this person because information’s so sparse, but you connected us with them. They’re open to a discussion. We dove into their experience and got a greater depth of it, and wow, perfect fit. We would’ve never found them otherwise.” Those are the stories I’d love to hear.

Tincup: 16:56 Oh, questions that you love to hear from prospects. You just know they get it because this is revolutionary, right? These are the questions that you would love if you could program them to say, “Here’s the five things that you should be asking me.” It doesn’t have to be five, by the way.

Brian: 17:16 Yeah. Yeah. Yeah. It’s a matter of… Oftentimes, customers or prospects will ask like, “Well, how do I know that they’re a good match?” Well, we take care to ensure that the candidates are matched. We go through a calibration exercise upfront. “How do I know? Look, okay, it seems to make sense, but how do I know? How is this truly going to work for me?” That’s a question that comes up frequently. Another question is kind of a success problem question like, “Oh, my God, how am I going to be able to handle the pipeline of candidates?” A lot of our customers, they’ll say like, “Look, I’m not-

Tincup: 17:57 It should be a good problem to have.

Brian: 17:59 That is exactly right. That’s what I call a success problem like, “Oh, my gosh, you’ll be so successful. How could I handle that?” Or you will deliver. It’s almost like, “Hey, I understand that you’re going to deliver and I believe that. I’m bought in, but how am I going to handle it?” So, oftentimes, we’ll walk through with our customers like, “Look, this is how we operate. We can scale up and down to meet your needs, as far as the flow of the candidates, because we can toggle the volumes to ensure that we can calibrate things in a way that…” We know our system pretty well, so we can understand response rates and levels of engagement of the outreach to ensure that it can meet those needs, but then also, there’s a human in the loop element too.

18:48 We have a customer success manager on our team. Customer success manager really is an extension of our customer’s talent teams in many ways and I think is a tremendous asset for us with our customers. In that, we’re going to take those inputs and say like, “Okay, based off of your needs, we can calibrate the search in these ways. We can calibrate the volume of pipeline in these ways.” So, it’s not just a total flip of the switch. There is this human in the loop that’s able to make sure that we can best meet our customer’s needs. So, when they do have that question around, “Oh, how am I going to be able to handle this,” it’s the customer success manager that’s able to toggle the inputs.

19:33 “How am I going to know that the candidates are the right candidates? How can we truly validate this?” Well, we’re going to keep calibrating. It’s not just taking the job description. Our first step is take the job description. We part the job description, and we’re able to identify the right candidates. But then there’s a human in the loop. A human in the loop could be… Things come up where we’ve had prospects or customers, and they say, “Look, we’ve had really bad luck with people that came to us from big company X, Y, Z like Verizon. We had really bad luck with people who came from Verizon, so we just don’t want anyone from Verizon, so let’s exclude Verizon.” That’s not something that you put in a job description like, “If you work at Verizon, please don’t apply.”

20:15 But in a customer success conversation, we can toggle those inputs. We can include and exclude and really dial it in. Those are two of the biggest questions that… I know that they’re bought in, that the system will work, and how do we know that it will work? How can we validate it? And then, the other one around like, “Hey, how do we handle the pipeline,” is another one. That’s just a matter of like, “Look, it’s not just a total automated process. There is this human in the loop element and that covers a lot of the things that might be concerns.

Tincup: 20:47 Okay. Last question, no brands or company names or anything like that, but your favorite customer story or maybe your latest and greatest customer story, if you like.

Brian: 21:01 Yeah, there’s a couple of them that come to mind. I would go back to something that I discussed earlier around identifying a candidate that a customer would’ve never found. There was an early stage company. There was a CTO and a team of about six engineers, and they were looking for an engineering manager. It was that. I’ll never forget it. It was just super clear that the talent manager just came on board and we were brought on a little bit before the talent leader came in, and she told us, “Look, this person, we would’ve never found them, and this person is exactly what we were looking for. They didn’t have anything on their LinkedIn profile that would have cued us in that they would be a great candidate. And then, after taking them through the process and meeting with the executive team, this is the exact candidate that we would want for this engineering leadership role.” Engineering leadership roles are incredibly hard to fill, incredibly expensive to fill, and for us to deliver on that in a way that was unique and that they would’ve never found this person was one of my favorite stories.

Tincup: 22:17 Well, that was fantastic. This was a great use of time. Brian, I know you’re busy. Thank you so much for your time and wisdom.

Brian: 22:24 Thank you, William. It’s my pleasure.

Tincup: 22:26 Absolutely. Thanks for everyone for listening to the Use Case podcast. Until next time.