Building a Talent Framework That Works

Raj DevHead of Talent

Credit Sesame’s Head of Talent Raj Dev & long time colleague Matthew Rafaelson present their Talent Framework: a new outline for attracting, developing, and retaining talent. Rob asks the specifics required to implement an advanced “high band” compensation strategy. Everyone drinks fancy whiskey.

Episode Transcript


00:05 Rob Stevenson: Hello again, my wonderful recruiting and talent acquisition darlings, Rob Stevenson here at the helm of your favourite recruiting podcast, and tonight, today, whatever time of the day you’re listening to this, it’s tonight for us, I have a humdinger lined up for you. In the studio, makeshift studio/beat lab/conference room where I have erected these sound dampening devices in a somewhat janky fashion, I have joining me Raj Dev, the head of talent at Credit Sesame. Raj, how are you?

00:37 Raj Dev: I’m doing well, it’s great to be here.

00:39 RS: It’s great to have you back. You are probably the fastest turnaround repeat guest I’ve ever had, so you get that distinction.

00:45 RD: Well, thanks.

00:46 RS: Yeah, I’m glad you’re here. And alongside you is Matthew Raphaelson, who is consulting with you at Credit Sesame. Welcome, Matthew.

00:54 Matthew Raphaelson: Great to be here.

00:55 RS: Great to have you, great to have you. In somewhat stark contrast, Raj, you brought friends with you, not just Matthew, but also two of the most delicious and fancy whiskeys that ever graced my lips. I’m no whiskey connoisseur, I’m not an expert when it comes to drinking whiskey, although I do know how.


01:12 RS: And so far, so good. So if you out there in podcast land hear some errant clinking of ice against a glass, just know that we’re having a good time over here. So I appreciate your mastery in all things whiskey, you as well, Matthew, and your mastery in terms of this framework you’ve put together, mapping… Well, why don’t I just let you explain it? How would you describe this talent framework that you two are putting together?

01:42 RD: Yeah, thanks a lot. So one of the challenges of talent acquisition, talent development in any company, is to move quickly and make good use of resources. And good use of resources in this case are, sure, there are the tech resources, there are some hard resources, there’s time, but the critical resource that’s visibly scarce is money. And so how do we take those dollars and use them to attract talent quickly, and then leverage that talent effectively in the organization? So for startups, who are, by nature, cash constrained and have to move quickly, this is essential, but this is valuable for any talent organisation that wants to attract, manage, develop, grow the best talent.

02:38 RD: And what we mean by a framework is the set of processes. We have an attraction model, bringing people into the company; a development model, how you grow them within their jobs; a promotion model, which is how you choose people for the next roles as the company progresses, more senior roles; and a compensation model, which is how you pay for all of this. These are not four independent models, they have to work together in an integrated way, and that’s what the framework does.

03:20 RS: Yes, and it’s also important for recruiters to realise that all these things downstream from you, your initial sourcing and your initial conversations, they impact your role tremendously, and the organisation’s success tremendously, right?

03:35 MR: Well, that’s right, you can’t improve upon the recruiting. The people who come in are who eventually get developed and promoted and paid. So starting out with the best possible resource at the beginning is critical, because you cannot just, all of a sudden, promote a great person who isn’t there already.

04:04 RD: And I would add, is that you have a great resource and you’re paying her the appropriate paycheck, the appropriate cash compensation. That’s essential, not just to attract, but that has a follow-on effect downstream, not downstream the recruiting funnel, but downstream in the employee life cycle, because if they start off with a cash compensation that is powerful enough to attract them, but then the compensation framework isn’t really synchronised with the way that talent is developed and promoted in the organization, then that can lead to a retention issue.

04:48 RS: Yeah, yeah, definitely. This is really fascinating. I’m glad you brought that up, because there is all this focus on equal pay for equal work, as well there should be. You get to a point where you want to compensate people in accordance with the value that they’re adding to the company. So you could have two people in the same function, one of them is doing fine, they’re doing an okay job, and the other person is just an absolute all-star who’s blowing it out of the water.

05:15 RS: And you used the example earlier, Raj, of someone’s cranking out patent after patent after patent, and they’re a once-in-a-lifetime kind of talent, they should be compensated along with that, right? And it shouldn’t be fair to just, this is the level for this role no matter what your output is, and by the way, here’s our incremental, merit-based yearly raise that may not be in accordance with the actual progress you’ve shown in your role.

05:41 RD: That’s right, that’s right. There are some limits, though. That connection between value delivered and the discrepancy in pay, that dispersion in pay, has to be conspicuous. It has to be understandable to everyone.

05:58 RS: Right.

06:00 RD: Where it is, I would assume that no one would really have any sort of problem with the dispersion of pay, because you can see that it reflects a dispersion in talent. And one case is in the NBA, for example, where no one seems to really be all that bothered with these huge pay contracts that are actually 10, 20x that of the…

06:26 RS: The sixth man, yeah, yeah.

06:29 RD: Exactly. But talk about conspicuous. You can see that someone’s talented, and you can see that they’re tall, you can see how they’re…

06:37 RS: Performing, yeah.

06:38 RD: Performing like a ballet artist on the court. And so people appreciate that, they look at that and they say, “I can’t do that.”

06:46 RS: Yeah, yeah. And in the event they did chirp up about it, it’s because they’re looking at their own numbers. They’re like, “Wait, I’m just as good,” right?

06:53 RD: That’s right.

06:55 MR: And one of the things that Raj and I talk about is that that is, the NBA example pertains to certain types of jobs in a typical company, but not all jobs. And you get into a false paradigm of saying, “Well, in order to accommodate the 10x coder, the star employee, we should have wide job bands so that you can pay at the high end more than you pay at the average.” And that is true for certain types of jobs, it’s not true for all jobs.

07:34 MR: Some jobs fit a bell curve type of distribution better, and a narrow band, and it’s really important for companies to understand that and to set the compensation structure for the jobs appropriately. And I would say most companies, maybe even all companies, have both kinds of jobs, even in the NBA. So a basketball team, the players probably exhibit this wide band compensation structure, but what about people in the front office? Maybe less so, I don’t think that in the marketing department, or the accounts payable department, they have that same 10x type of performance disparity.

08:26 RD: That’s a great point. And that’s not disputing in any way that, the more value you produce, the more you should make, but the magnitude of that discrepancy, it changes depending on where you are at the function.

08:42 RS: So are you saying that perhaps they should have that sort of pay disparity in the front office, or it’s a result of the lack of scarcity relative to the, like a LeBron James?

08:53 MR: No, I would say that we’re saying is that the job design in the front office is gonna be fundamentally different than the job design of the players, and that you’re in the accounts payable department or the promotions department, your pay ranges are probably going to be in a tighter band. And one of the concepts that we want to describe is when you’re creating the job bands, there are three parts of a job band. There is the entry part, which we call the time to proficiency. So you’re new to role, you’re learning the job. How long for that job does it take to get to market proficiency? Then there’s, when you’ve reached market proficiency range, and then, beyond that, there’s the mastery range.

09:47 RS: Right.

09:47 MR: You’ve been in the role, you’ve learned it, and now you’re mastering it. Different jobs are going to have different ranges. So for example, an entry-level job, a more clerical job, the time from start to market proficiency might only be a year or 18 months, and then your mastery period may be also a short time, and beyond that, you really can’t be, even if you’re the world’s greatest accounts payable person, in most companies, at a certain point, that’s not gonna be adding much incremental value. And if that’s the case, the job band should be fairly narrow. An example: You would not want to have that job band to be ranged from 75% to market to 125% of market, ’cause you bring somebody in at 75% of market at 2, 3% average merit increases every year, it’s going to take 10 to 15 years to go from 75% proficiency to 100%.

11:09 MR: It makes no sense for somebody, for a job where it takes 12 to 18 months to get to proficiency. And I have seen that, lived through it many times where, because people didn’t think of the proper way to structure these jobs, bring people in at 75% because oh, it’s a good deal, we’re bringing in… Saving money, and then, every 18 months, they go through a fire drill of re-leveling the department because people are being paid too low, and that’s a structural flaw, and it’s resolved if you design the job bands properly, recognising that some jobs should have narrow ranges, while others can have much wider ranges.

12:02 RD: Matthew, you’re articulating these concepts really well. I’m wondering, as we talk, if there are some simple rules that might be helpful for folks that can help people distinguish between these narrow band jobs and the broad band jobs?

12:21 RS: Yeah, yeah, that was what I was thinking. Because you look at this sort of comp structure does exist, in sales, for example. The best-selling salesperson is gonna make a lot more than someone who just barely nudged up against their quota, and that’s just sort of the way it is, but that doesn’t really apply to other areas of the business.

12:42 RD: That’s right. So one framework, one rule of thumb to use is, whether the job is constrained by systems or processes, mechanical or otherwise. So, if you’re on an assembly line, that assembly line only moves at a certain rate. And I don’t know if you’ve seen the Lucille Ball episode…

13:08 RS: Eating the brownies.

13:09 RD: With the candy, exactly, exactly.

13:12 RS: Totally, it’s a classic.

13:13 RD: Where the assembly line is moving too fast, you can’t create value anymore. So you’re really bound by the speed of that conveyor belt, and maybe the quality of wrapping, or something like that. A secure…

13:26 RS: What is the conveyor belt in this metaphor?

13:28 RD: Candy?

13:29 RS: No, no, no, no, no, within the talent function.


13:31 RD: Oh, oh, oh.

13:33 MR: It could be process. It’s where… Raj is correct, there are many jobs that are process-bound, where the process really greatly narrows the possible range of performance.

13:47 RD: So if I’m a receptionist in charge of answering calls, I can only create value when a call has to be answered, and then I start the shift and I end the shift. And there are a finite number of calls that can be answered. There is certainly some variability, because there’s the quality of your answers, and there’s some variability because there are some high-volume situations where, maybe one particular receptionist has to answer a high volume of calls, and irate people, so they’re tougher to deal with. But for the most part, the range of things that you can do that would give you the opportunity to create value, that range is not that large.

14:38 RS: I see.

14:39 RD: For example, someone that codes for a living. That person’s laboratory is their head, their mind. They’re not bound by much. They’re not even bound by their computer being on. They can close their eyes and they can imagine new algorithms, or uses of their algorithms, and all of those things can result in a patent, or two, or three. So that represents the other extreme, where there is very little boundary on that job. I’m gonna interrupt myself. These are terrific questions, and they’re questions that aren’t asked when a lot of these fads occur in HR, talent management, talent acquisition in the comp world, and some people listening will remember, or have seen this term called “broadbanding”.

15:35 RD: And sometimes it’s in favour, I think in the ’90s it’s in favour, then it goes out of favour, then it goes in favour, then it goes out of favour. The need for it always exists, and the need for it to not be so broad always exists. They’re both correct. It’s not one-size-fits-all, and that’s part of what Matthew and I are trying to do right now.

15:55 MR: It’s not an either/or, and it can be very company-specific, and each company is going to have to determine, based on what it does and its own processes, which functions, which jobs, really do need these broad bands, so that you can accommodate the star performer who’s four or five times more productive, or 10 times more productive, and which fall into the more process, limited types of jobs where performance is naturally going to have a much narrower range of performance. And as I was saying earlier, I think all companies will have a mix of those jobs. And so, going with a fad and broadbanding everything, or narrowbanding everything, misses the point.

16:53 RS: And the risk of not doing this would be losing your top performers to another company that maybe pays above market, because in a world where you’re not doing this, probably what you are doing is going out and buying data on what is a market salary for this role. And then you’re gonna maybe, you’re gonna give that, or maybe you’re gonna tweak it a couple of percent to sweeten the deal, but then you’re just gonna lose to Netflix, who comes around and drops 130% of market on the table, right? So is the idea with this plan is like, you are less liable to lose top performers in that case?

17:27 MR: And remember, when you’re getting the market data, it’s likely just telling you what 100% of market is for a position. In other words, how to pay people who are at proficiency.

17:44 RS: Okay, I see.

17:45 MR: It’s not really telling you, “Well, how do you pay somebody new to the role in a job that might take them four years to get proficiency?” And it’s not telling you how to pay your stars. I think there is ways to look at the data to get some hints that a market might be telling you that these are wide band jobs. There might be some variance in the data, but you have to apply a little bit of data science to it, in my experience, to glean that, but I think, in reality, each company is gonna have to decide for itself what the range around average is for a job.

18:24 RD: And if you don’t do this, if you don’t have this process around why… To understand exactly what is the appropriate dispersion, and how do I leverage that dispersion in pay in order to attract high-level talent, you could just say, for a particular function, “We understand that there is a variance in the amount of value delivered. Great, let’s have a broad band.” And then you stop right there, and you don’t think about who’s market proficient and who’s not market proficient. Then you have a situation where you’re under hiring, and you’re making your numbers because you have this latitude in order to pay the… You use Netflix, but that could be the right example. Pay a higher salary in order to get market proficient talent. And so that’s a sloppy use of money.

19:20 RS: Right.

19:21 MR: There’s no question that if you do have jobs with broad bands, that if you execute poorly against it, you’ll get terrible results. And so, just because a job band might go to 150% of market or higher doesn’t mean that you pay people way above market just to make recruiting goals, just to speed up your hire.

19:47 RS: It happens though, right?

19:48 MR: Yeah, of course. So you have to have disciplined execution behind the concepts, or it doesn’t work. But that’s I think pretty much anything without disciplined execution could end up badly. That shouldn’t be the reason why you don’t do something.

20:07 RS: Sure, sure. Then this re-evaluation, this reframing how one looks at compensation, who is responsible for doing this? Is this a CEO task? Is this a VP of People task? Who do we… Because I can just imagine recruiters are saying, “Yeah, that sounds great, but am I gonna put together this whole new comp structure and try and convince my boss that the reason we’re not retaining top talent is because we’re only paying them what 100% of someone who is at proficiency would make,” as the example you gave, Matthew?

20:41 RD: Well, there are two pieces to your answer. Number one is, if I’m understanding correctly, where does the idea come from and who quarterbacks the process, right?

20:51 RS: Yeah.

20:51 RD: Number one. But then, the second part of this question that naturally follows is, how many pieces have to be connected once the quarterback initiates the process? So let me answer the first question. That depends who quarterbacks this process. In some companies it’s finance, in other companies, it could be talent analytics. In another company, they could have a compensation function, or compensation analytics function. So that varies.

21:22 RD: It’s less important who quarterbacks this, just the history, the legacy of the company, where the talent is, who knows what the reason is? It could be quarterbacked in multiple places, but not integrating this across functions is the real issue here. You can have a brilliantly designed plan, and then you initiate it in talent acquisition. You don’t have the appropriate guard rails, it seems to work because they’re making their hiring goals, but it’s sloppy hiring, and it really doesn’t follow along cleanly into the development and promotion aspects of compensation. So just making sure that you have an entire view of the process is important, whoever is quarterbacking this.

22:13 MR: I think, ideally, your quarterback would be somebody with the title of Head of Talent, Head of Talent Management, because it does require bringing together multiple disciplines, and there is a very strong financial aspect of it. It’s no mistake that Raj has a finance background, and is able to tie that piece into it. And I would expect that the talent manager of the future would have a strong quantitative and financial background to be able to integrate this model.

22:58 RS: This review of compensation is pretty advanced, I will say. And there’s a lot of I’s to dot and T’s to cross before one gets to this stage. And I think being able to attack it at this stage presumes that one has removed all forms of biases from the offer process, from the interview process, because it seems like this sort of variable comp spread out across the company would be ripe for biases to be involved. So, is there… Does one need to… Is this like a “remove the splinter in your eye before you remove the beam in your neighbor’s eye” type of situation, where it’s like, first do this, then you can go on to the advanced comp strategy? Where does that play into it, do you guys think?

23:49 RD: Yeah, that’s a terrific question and has to be addressed. And I would say if it’s not being addressed, it points to… One of the things that, Matthew, you and I have been skirting around, diving into, but not really addressing head-on, which is the need for discipline around implementing this type of a framework. So if you’re going to create these large bands, they can’t just be large bands and then that’s it, because if you have large bands, you will get faster hires if it’s undisciplined, but that’s at the cost of under hiring. So is it really worth it? Because at a certain point in time, they’re diminishing returns.

24:39 RD: Other ways that this lack of discipline reveals itself is bias. Let me actually fast-forward to something that the recruiters can just grab onto and implement tomorrow. Where the rubber meets the road, especially for the recruiters, is how the comp is inside the requisition form attached to the job description, okay? There are two things that I require from a solid job description, and there are two things that I recommend every talent profession think about when they look at a job description. Number one is, is the job description crisp and clear enough so that anyone can pick up that job description and conduct a screen or a non-technical interview with the candidate, number one. Number two, is the job description, does that job description describe the role clearly enough that a one-year performance evaluation can be done just using the job description?

25:57 RS: Right.

25:58 RD: And then, the third aspect that should be considered when looking at the job description is this: Does the language and tone represent the company and exclude any possible sources of bias? So what we do is, we use Textio as an application, and we use the application for clarity, but we also use the application to filter out possible biases, gender biases, for example. So one example is “determined” tends to bias against women and towards men, whereas “collaborative”, “caring”, “compassionate” tends to bias towards women but against men, and both are required for, say, an engineering role, but the second set of terms are frequently not incorporated in the engineering job descriptions.

27:03 RD: And then the next thing is, in looking at, not just tone, but the bullet points for either irrelevant bullet points. For example, “years of experience” could bias against a proficient young person who’s a superstar that you could get at a bargain, or could bias against one category of person or another category of person, but are irrelevant to the job itself.

27:36 RS: Okay.

27:39 MR: There’s also a set of examples almost on the other end of the spectrum, which I call “recruit-o-babble”, and it’s words that you see in lots of job descriptions that can mean anything, and as a result, mean nothing.

28:01 RD: As a result, mean nothing.

28:01 RS: Yeah, yeah, yeah.

28:03 MR: So…

28:03 RS: “Positive self-starter.”

28:05 MR: “Self-starter,” “proficient in Excel,” “professional,” “detail-oriented.”

28:09 RS: “Good communication skills.”

28:10 MR: Right. And… Exactly. And so, because they can mean nothing, basically, then how those become evaluated is almost purely subject to bias. And what I like do in my old world when I wanted to hire a lot of analytics folks, instead of saying, “self-starter,” “proficient Excel,” “professional,” those kind of things that meant nothing, I would try to come up with language that would speak to the person that I was looking for in a way that wouldn’t speak to someone else, and something like saying, “If you love data, building models, and quantifying the world around you, you will love our analytics team.” And when you paint that picture, going back a year down the road, you now have some specifics there to evaluate whether that person really did exhibit those skills and behaviors.

29:14 RS: Okay.

29:14 MR: Come up with something.

29:16 RS: Yeah, no, I love that. And I think the entire talent framework that the two of you are putting together, covering it in one podcast episode is a little ambitious, but we did get a good look at compensation review, we got a good look at job descriptions, and the initial part of the framework. So at this juncture, I feel like we’re winding down here, but I wanna end on a high note. Raj, do you have any parting wisdom that you’d like to impress upon the recruiting superstars of tomorrow?

29:49 RD: Sure. I’ll finish up with a reminder, those two questions to ask about the job description and remembering the tone in the job description, remembering that there shouldn’t be anything irrelevant, there shouldn’t be an over-specification in the job description, there shouldn’t be irrelevant specification, there shouldn’t be recruit-o-babble in the job description.

30:13 RS: Right.

30:15 RD: And then I’ll say this: The job description, as dreary and humdrum as that document may be, may appear, as administrative as it’s been historically, is one of the first opportunities for you to show off your talent brand, for the candidate to understand the culture of the company, how the company operates; all of that can be represented in the job description. And then finally, that job description is an advertisement.

30:54 RS: Yes.

30:55 RD: And that’s something that people don’t always take into account.

31:00 RS: Yes, it is an advertisement. And so, when Matthew explained, if you like to quantify the world around you, yeah, do that, get your Don Draper hat on and be compelling, speak to the person who is gonna be good at that job.

31:11 RD: That’s right.

31:13 MR: Exactly right.

31:14 RS: Alright, well, I think we’re at a good point here. So, at this juncture, I would just thank you both so much for coming in and chatting with me about all this stuff, it’s been fascinating. Thank you, Raj, for your whiskey and your amazing master chemistry, old-fashioned concoction, it was absolute, an absolute delight. Matthew, your wisdom was a blast as well. Thank you both for coming in here, this has been amazing, you were amazing. I don’t know what else to say. This has been great.

31:41 RD: Thanks for having this.

31:42 MR: Yeah.

31:42 RD: It’s been a really pleasure.

31:43 MR: This was really fun, thank you.

31:44 RS: Cool. And to all of you out there in podcast land, that just about does it for your pals here at Talk Talent To Me. Once more, I’ve been Rob Stevenson, Matthew Raphaelson has been Matthew Raphaelson, Raj Dev has been Raj Dev, and you’ve all been amazing, wonderful, talented, recruiting darlings. Have a spectacular week, and happy hunting.


32:11 RS: Talk Talent to Me is brought to you by Hired, a double opt-in global marketplace connecting the best fit active talent to the most exciting recruiting organisations. If you would like to learn more about how we can help you find your next great hire, head to hired.com/employers, and we’ll get started.