Big Data Brings Big Change to the Recruiting Table
HCI’s Nine-to-Thrive HR Podcast featuring Mike Guglielmo
Mike Guglielmo, Assistant Professor of Human Resources at Temple University, shares much-needed knowledge of big data, including how handling data has changed, where data can be found, data in decision making and the most ideal technology to get your hands on. The full transcript is included below for your convenience.
Hello everyone, I’m Holly Pennebaker, and you’re listening to a new episode of HCI’s Nine to Thrive HR. This podcast features experts and practitioners in the field of HR and brings their knowledge of the most pressing issues facing talent management straight to you. We talk about current industry problems, but most importantly solutions you can use in your own organization. When it comes to talent acquisition, today’s top candidates know that having the right talent in the right roles can be transformative for your organization. If companies want the best talent, they should employ a customer-centric approach to talent acquisition that emphasizes effective candidate management, and clear alignment with business goals.
For example, SkillSurvey’s cloud based technologies help you make better hiring decisions faster, and experience a more efficient, and effective recruiting process with their online reference checking, and talent sourcing solutions. Here to share some expertise on the topic is Mike Guglielmo, Assistant Professor of HR at Temple University.
Mike, I know you’ve got quite the history, so why don’t you tell us a little bit about where you’ve been, and what you’re up to now?
Hello everyone. Mike Guglielmo here. I am on a non-tenure track position, as Assistant Professor of Human Resources at Temple University. I’ve been an adjunct there for over 30 years. Recently, in the last two years became full-time. And before that, or during that 30 year period, 35 years actually, I’ve been in various industries from chemicals to manufacturing, service delivery, publishing, financial services, and healthcare. In a variety of positions from accounting to finance, to IT, and HR, and as well as operations.
I’ve seen quite a bit around data, as well as ran recruiting for two major organizations. One in financial services, the Vanguard Group, for a number of years. And then Genesis Healthcare, in the healthcare space within the last ten years. Anxious to provide some information to our listeners on this podcast, and hopefully my background will add some color to what I have to say.
I’m sure it will, Mike. We’re lucky to have you today. And so, let’s just start out with a broad question on what is big data in recruiting?
Well, that’s… I guess, a question that many people ask in many different aspects of life, let alone the business world. In recruiting, it is data that I think many people may associate with large databases. Quite a number of folks are looking to LinkedIn for information, and there is lots of data there that can be found. What I really prefer to say is that many folks that believe that they understand big data, think they can just data mine information in these huge vast databases, and it’s much more than that.
You can’t just data mine your way to a great candidate. You really need to understand alignment to your business, data aggregation, which I know we’ll talk about in a few moments, discovery of the data, data visualization, and then insights into the data. You want the best data that is going to provide the most informative information back to you, if you’re in a recruiting department, that you can then turn over to your customers, most times your business customers, to say here are your best candidates, and here is why.
That entire process sounds relatively simple, and it’s not. And then there is lots of hidden pockets of data that I think are being missed, that are part of that big data conglomerate, that I think would be beneficial for our audience to hear a little bit about as well.
And so, it sounds like using these big databases is not exactly a one size fits all, and that it really comes down to finding the data that’s most relevant to you, and then that helps you specifically find the best candidates, depending on who your company is, and what it does. I’m sure that has also come with a lot of change. Mike, in the past 20 years, how have you seen things change when it comes to being able to collect and handle data when it comes to recruiting?
Well, as a bit of a dinosaur myself, I worked in several jobs that required that we did things with paper and pencil, and 11 x 17 columnar paper. Was lucky enough to be part of the first PC being introduced into an HR group. It sort of transformed me, in that I was going for my Master’s degree at Temple University in the evening, and was a finance major when I started, and this meeting of data, and PC, and technology with jobs got me very, very interested in the application of technology to HR. From there it went to databases, and spreadsheets, and then to client server applications, and applicant tracking systems that have been used for many, many years. Probably some are still being used, where the company is housing the hardware and the software locally, as opposed to Cloud based services.
And then there is add-on programs that have come along. One of which has been SkillSurvey. More recently I became aware of them. I was asked to be a reference for someone, had never seen or heard of SkillSurvey before, did the reference, and at the very end it said, “Would you like to be contacted for a job?” And I said, “Wow, that is a terrific way to capture information from a passive candidate pool, the boss, or the peer of the applicant.” And as soon as I wound up in my last role in corporate America, in recruiting and healthcare, we reached out to SkillSurvey, and started to use it right away. And that was a terrific way to find new candidates, many times at a managerial level, who weren’t actively looking for jobs.
All these little touch points along the way, data that you get from assessments, from screening, from sourcing, search engine optimization, and all the data that’s out there. I guess the biggest opportunities for companies is to figure out which ones help, you have to figure out what your company’s strategy is, and if you’re in a downsizing mode, it’s probably not something that you’re going to spend a lot of time on, if the data is all about getting people in. However, if you’re in a growth mode, something like that probably will align to your company’s strategy, and you want to capture all the data elements you can, and then put them together in a way that you can try to make sense of them. And if you’re lucky enough, and this is for more advanced users, many of the data scientists do some of this work, there are ways to look at connections, and correlations, and predictions based on all that data you have, to help you make better decisions moving forward.
Like, how do you retain people? What’s the warning signs that we should look for, of people who are at risk of leaving? If our best performers have these things in their background, how do we target that based on the way that we’re recruiting right now? Just to a name a few instances of that. But it certainly has changed… And the Cloud, and making this information available on a relatively quick and cheap basis.
I remember back when I first started in corporate, we had mainframes, and mid-range computers, and data storage was so expensive that you had to have codes for everything. You couldn’t write out the full name of something. That’s why you may see codes for certain things in legacy systems today. The ability to basically make it into a commodity, and then to get it at it, and analyze it using tools that have been around forever, like regression analysis, and structured equation modeling, things like that. Really help you take the data once it’s aligned to your company’s strategy, and figure out can I do something with this, and make better decisions?
So, it sounds like technology has been a big factor in helping us get better when it comes to data. I think maybe we would want to address the source of that data. Could you tell folks where is the best place to find the most helpful data?
Well, most companies I would expect would utilize their applicant tracking system. Since we’re talking about recruiting, their applicant tracking system is usually the best source of your initial dataset. It is… Should link into your advertising sources. And when candidates come into your system, you should be capturing as much information as you possibly can, and not just the sources of where they’re coming in, but also how they found about you.
We used to have this challenge all the time, did you find them through a particular ad that you placed on the web? Or did they find you through word of mouth? Why not capture both, and figure out if there is any correlation between those, and if that gives you more insight into where you ought to spend your advertising dollars?
And then there is data around screening, which you ought to be able to automate. You’re gonna probably hear a lot more about that in the next few years, about automation in HR. I know a number of recruiters are a little terrified of losing their jobs, and I don’t see the relationship side of their jobs going away. But I think a lot of the transactional, and monotonous work that has been done, can be automated with excellence. And with each piece of automation, it makes it easier to capture data.
Information that a person may have captured and logged potentially, sometimes they didn’t, through a screening process, you can capture, and figure out if there is some knowledge and insight that you can gain from that information. Then during interviews, I’m guessing that many people do a five point scale on the appropriateness of the candidate for the job, maybe a 7 point scale, it doesn’t really matter. But whatever it is, there is a way that you ought to be capturing that information. Once again, you can go back and take a look at if there are any insights into who’s made it into the final rounds, and are you wasting your time on that if they don’t ultimately succeed in the job?
There is also a heck of a lot of information if you use an assessment tool, that many times looks at cognitive ability of the candidate, the behavioral traits of the candidate, the job fit of the candidate, and all of that can be captured. Then there is references, like for SkillSurvey, as I mentioned passive candidates can be captured, so that you’re getting an additional candidate pool, but also we were able to capture information on individuals through the reference checking function of SkillSurvey, that allowed us to take a much better look at whether we wanted to hire someone or not. I can tell you after being in the business world for 35 years, HR for 25, recruiting for 20, the number of people that we actually ever declined based on a phone reference was very, very small, almost negligible.
What we were able to capture through automation, and through a system like SkillSurvey, allowed us to take a much more cautious approach to who we were going to hire, and it really improved our ability to cut down on 90-day turnover, one-year turnover, and even I would say two-year turnover because we were able to find better fits for what we were looking for.
And then there are also other information that you get after people are onboarded, how well they think the recruiting process went, how well the manager thought the recruiting process went, both at the… Let’s say the 30-day, 60-day time period, and get that feedback, as well as six months. In a prior role, we used to ask both the hiring manager, and the candidate, if you had a choice to take this job all over again… This is six months later obviously, would you still do it? And we had great scores, but every once in a while it would dip down, and we’d have to try to figure out what are we doing differently, or not so well that we need to adjust? And likewise we would ask the hiring manager, if you had a chance to hire this person all over again, six month later, would you do it? And again we would scrutinize that data to make sure that we were really getting the best candidates in, and did our process go as well as it could?
Ultimately, you want to make sure that the people that are coming in are staying, are productive, bring back a reward, and a return on the investment that we’re making in them, and the recruiting process to bring them in. And most importantly, if all of that is not improving customer and financial outcomes, then you really need to take an even closer look at your data to figure out what you can do differently, because there is nothing that we would be doing that doesn’t address that type of outcome, customer and financial outcomes. It’s got to be aligned to that, or else you’re really just going through activities, as opposed to aligned outcomes for the business.
All right, thanks for all of that information, Mike. And I bet when it comes to this data, and how there is just so much out there, it can be really overwhelming to people. I know myself, when it comes to looking for different things, or finding answers, I might not always know where to start. With their being just a tremendous amount of data available, how do you know what to pick and choose from to help with your decision making?
That’s a great question. And I’m going to borrow a term that’s being used, and probably overused right now. There is a process called Agile, and there is… Dr. Peter Cappelli talks about Agile HR. And it’s an approach that I’ve taken in creating dashboards for most of the roles that I’ve been in. If I have limited data, there is data that I want to start to track, and we haven’t yet. We may have disparate data sets somewhere tracked in a spreadsheet, in an Access database, in a SQL database, and none of these things are connected back to the company’s Human Resource Information System, and / or the applicant tracking system.
But I’m less worried about the absolute aggregation of that data, as I am the ability of someone in HR, or a metrics group to get at that data, report it on a regular basis, and again make sure that it’s aligned with the outcomes that you want. If you’re not worried about retention, then really why are you tracking retention data? If you have people lined up at the door, or the virtual door as it would be, like Google does with… I don’t know, they have a bullpen if you will, of people that are dying to get into the company. They don’t really have to advertise as much, and build an employment brand. It’s already pretty well known. In that case, why would you track your branding efforts and your sourcing efforts?
Knowing what your business strategy is, what your human capital needs are, drives a lot of what you’re going to track. Get what you can, track what you can. Measure it well, make sure that it’s valid and reliable in your measurement techniques. And if it is, we just started to display it on a dashboard, and through an iterative process with our business partners said, “Hey, this is what we think that we need to show to address your needs, business partner. Tell us where we’re off. Tell us what you think you want us to dive into, that we need to segment even more?” And they would do that very readily once they had a data visualization, or some type of dashboard that they could put their eyeballs to, and say, “Yeah, I don’t think I really need this, but go ahead and track it anyway.” Or, “Hey, you forgot about this. Can you add this in?”
And that’s when we would go back and try to figure out if we could get the data, number one. And if we could, to start to track it on a regular basis. But there is a lot of data visualization tools out there that put all this together. Some are free, some are expensive. And that helps in the ability to communicate this information out to the field. But, hey, I’ve used the spreadsheet in Excel, and built… Along with a small team that I had in both of my last two organizations, Vanguard and Genesis, to basically capture the data in separate tabs in a spreadsheet and display it out. In some cases we built little access databases to make the display capability easier. But in many cases we just gave it a stoplight indicator. We’re red in this one, and we need to do better. Here is the initiative that we’re going to kick off to make our numbers turn around. Yellow, it’s cautionary. Green, we’re good. We’re on target.
And we would typically measure against our goal for the year and keep that in an ongoing basis to make sure that people could visualize where we were against those goals, and also to give some historical context. Usually just the year before, or the same time period, like March year to date this year and last year, or June year to date this year and last year, and sometimes we would go two years back. Just start somewhere.
I’ve given this advice to anyone that I’ve counseled on this. Start somewhere, work with your business partners to make sure that you’re tracking what they need, and you certainly can track what you need to get your job done, but you want to make sure your business partners are in agreement with your tracking, and it helps them with achieving their business results.
And then once you have accomplished to some level of gain for your department, and your business partners, you can start to take a look if you need to bring in somebody from the outside to do some of the real analysis, where you can do regression, and structured equation modeling, to figure out whether there really is correlation, or causation between different variables. That’s where you really can say, “Yep, that actually does cause this, and here is your percentage of certainty that it’s actually in alignment, and correlated.” That’s a little more complicated. In some cases, it’s a lot more complicated, but it’s still beneficial to start somewhere, and get the data aggregated, and put onto a one page dashboard that you can communicate to both your team, and your customers.
Right. Awesome, Mike. I think a big important point that you made there is that you have to start somewhere. And I think that, that can be the hardest part for a lot of people. I also think a lot about the good’s and the bad’s. You know, you’ve mentioned a few of these already, but maybe we could just provide a simple list for folks to walk away with. What would you say are the benefits and limitations of big data in recruiting? What all have you seen?
The benefit is… With very, very simple tools, and I’ve been using what I would say is the simplest visually correlating tool that does a crosstab matrix. Microsoft’s PivotTables. They’ve come a long way, but I’ve used them since 1993 when they first really made available, and people became aware in my mind. It may have been around before that, but that’s when it really struck me that, wow, I can do these associations between different variables really simply. And with Excel now taking a million lines of data, the limitation before used to be 64,000 data records, and it’s always had the capability to reach out to different databases. And literally as far back as 1993, we were connecting PivotTables to SQL data, and crunching millions… In one case, I think it was 16 million records that we put together, and did analysis based on that.
Minimally, get yourself familiarized with Microsoft Excel’s PivotTables. There are additional things like Power Pivots, and there is Microsoft Business Intelligence, or Microsoft BI, that lets you visualize it easier. But the tools are so simple now, and you can really create meaningful charts, and data tables for your clients, that as long as you know that the data is clean and reliable, that’s one of the best things you can use as a tool. The insights that you get from it, again it’s worthless if it’s not going to improve customer or financial outcomes. Ultimately, you’ve got to draw some type of alignment all the way up to customer and financial outcomes.
Professor Dick Beatty talks about that often in a lot of his writings. He is one of the co-authors of the Differentiated Workforce, and he talks about this all the time, in looking at engagement data. You can survey people on just about anything, but if it doesn’t align to customer and financial outcomes, you have to ask yourself the question, why?
There are lots of tools. Lots of benefit of being able to communicate these insights, however small they are to your business partners. And then the downside is it takes a lot of time sometimes to get all this information together, to ensure that it’s reliable. Because ultimately if it’s not being captured by a machine, or technology, then you always have the chance that there is user error. And even sometimes the way that technology is set up, it may be capturing information that is skewing your data. You need to be really, really careful about what you’re gonna record. And one piece of advice I give to folks listening is if the numbers don’t look right, if there is something that just seems off, your instincts are probably right. Dive into that information, and go look more carefully.
I can’t even begin to tell you how much embarrassment I’ve been through, and how much I’ve avoided by looking at data. Or not looking at data as carefully as I should have because I was under a time crunch, but then saying, “Hey, I’m not gonna have this report ready. I need to look at it because something doesn’t seem right,” and finding, yes, ultimately there was something wrong with the underlying data.
There is a time commitment, there is… I think people that are afraid of data, or of the analytics around it… I teach an HR metrics class for undergrads, and graduate students, and there is very little statistics in it. There could be in probably a part two of the class, but really what I’m trying to tell people is that the data… The human capital data we’re capturing from recruiting to engagement, to retention, to culture, whatever it is. You need to make sure the data is reliable. You need to make sure that it’s aligned up through the company’s operations, it’s customer loyalty components, as well as its financial outcomes. And then you want to make sure that you communicate it on a regular basis, and you’ve got buy-in both from the people on your team, and the business partners that you serve.
It’s not always about understanding statistics, it’s just as much about understanding where the data exists, how to put it together, how to aggregate it smartly, and then how to display it. Data visualization I would say is as much an important part of big data, and decision making, as correlation analysis for lack of a better example.
Great information there, Mike. Thank you so much. Great points. And so, let’s go ahead and give people a little something more to takeaway today, and come full circle about, what would you say are the top three technologies that you think people need today?
All right. These are gonna be in no particular order, but artificial intelligence, and office automation wave is not gonna stop. And someone told me… I honestly don’t know if what I’m going to say is actually true, but the person that told me was convincing enough, so I’ll take him at this word. That back in the late ’70s, when the ATM was first introduced, there was much written about the death of the bank teller. We weren’t gonna need bank tellers any longer, and this individual told me that there are more bank tellers now, today, in 2018, then there were in the late ’70s. Because they changed what the people are responsible for. There is much more customer service. There is much more salesmanship, and selling of loans, and other financial instruments. It’s a good lesson for us to take a look at.
But for the recruiters, and the recruiting assistants, and the sourcers, and the recruiting coordinators who are worried about their jobs, start thinking about what else you can do. Because there are many things that can be automated like screening questions, you don’t have to have a person ask someone screening questions. You can have your applicant tracking system, or an adjunct system ask those same questions, gather the data, figure out if that person is gonna make it to the next level or not, and have legitimate cutoffs, and levels that the person would have to achieve in order to make it to the next round of interviewing. But the real benefit of having a person interact with the individual is as part of the candidate experience.
Well, if you really think about it, everyone that comes in is a potential purchaser of your service or your goods, so why would you want to upset them? You want to have interactions, even if it’s an automated interaction. “Hey, thank you so much for your interest in our company. We’re still looking at applicants, we’ll let you know.” Or “Your qualifications didn’t meet, blah, blah, blah.” You get the idea. But to not send anything back is reprehensible, and doesn’t improve your employee brand.
Those areas, the employee experience, how you’re gonna utilize artificial intelligence, and office automation to make that part of the experience better, so that you can have the human component work on the experience and tell people why they want to work for the company when you get down to your final few candidates.
The need for analytics around that data, we’ve talked about already today. I don’t want to belabor any longer, but understanding where your candidates came from. And the more data we gathered with our branding and advertising partner that I used in healthcare, I was able to figure out which products, and websites, and other ways of gathering candidates, and ultimately applicants for our jobs, which ones gave us the best return. And we were able to change our spending habits with some of those partners to spend more on the ones obviously that were giving us a better return, and more candidates, and more hirers. And ultimately hires that not only came onboard and stayed, but were promoted, so that we utilized very, very basic information. Nothing overly sophisticated when it came to analytics, but we were able to look at, hey, this is the way that we want to spend our money, and continue to bring in great candidates, that are in fact enhancing the productivity and the return of our HR practices, and ultimately for our company.
And then I guess the last thing is whatever you’re using… And I touched on this already, but I think it’s worthwhile to reiterate. Your systems really should be candidate friendly, and candidate experience friendly. Because those that are forcing candidates to go through your system, whether it’s in a kiosk in a store, or you have to use your PC. Hey, let’s face it people… Almost everyone has a phone, or a tablet, and they want to use that system that they do for so many things in their lives, to even apply for a job. And they want it to be simple. The old rule used to be, more than three clicks, people are probably gonna not continue the transaction.
Focusing on what the ultimate candidates want and getting technology that helps support that experience is probably gonna pay off big in getting the right number of people to come through, a diverse and inclusive group of people that are gonna come through, especially generationally. And then the folks that are gonna complete the transaction, not just look at your job posting, but also fill out an application, and hopefully ultimately get hired. There are a couple of the things.
I can’t say enough about my experience with SkillSurvey, and the ability to do reference checks in an automated fashion. It’s simple, it’s quick, it’s easy. And it really takes away a lot of the monotonous, and mundane, and repetitive tasks that we as recruiters used to do.
And so, I hope that’s helpful, and there are just a couple of the areas that I think would benefit our listenership with… From a technology, and more of an experience standpoint, then actual tools. But the experience from a branding experience, and a candidate experience I think is absolutely paramount. And your customer experience from your business partners that you’re trying to get this information together for, so that they can make better decisions, and have better people, and candidates, and hires, inhabit their open positions.
All right, Mike. We’ll let that talk on technologies take us to the end of today’s episode. And so, we’ll thank you one more time for spending some time with us.
Well, thank you very much. It’s been my pleasure. And I hope to be able to do it again sometime soon.
All right, everyone. Don’t forget that you can discover new talent and assess your candidates’ soft skills with the only solution proven to reduce first year turnover. For more information, visit www.SkillSurvey.com.
And be sure to subscribe to this podcast if you enjoyed your time with us. You can find HCI on iTunes, SoundCloud, Stitcher, and on the YouTube channel HCITalent. If you’re listening on iTunes, please give us a rating. Your rating helps other professionals and talent minded people discover the program. For Nine to Thrive HR, and all of us here at HCI, we appreciate you for tuning in.
Here’s what to do next
See how to harness big data to make better hiring decisions
Create a free trial account and access our solution for 30 days
Tour the platform that’s changing the way people reference check