Big Data for HR Interview

The following is a transcript of the interview between Total Picture Radio producer Peter Clayton and Andrew Gadomski.  Andrew will host a monthly podcast to discuss issues related to Big Data and HR.

To listen to this entire interview, click here.

Hi.  This is Peter Clayton.  Welcome to Big Data Channel podcast here on TotalPicture Radio.  I am in New York City today with Andrew Gadomski who is the founder of Aspen Advisors.  Aspen is a consulting and HR technology company that specializes in providing businesses with workplace analytic tools and processes they need to succeed.  Over the past couple of months Andrew and I have been talking with Anna Brekka(Recruiting Trends) about launching a Big Data Channel here on TotalPicture Radio and turning the microphone over to Andrew and having him interview a number of the thought leaders in this space because as you know if you’re attending any of the recruiting or HR conferences or events this year, big data is always on the agenda and it’s something that a lot of people really don’t understand.  It’s a term that is very confusing to a lot of people.

Workforce analytics, big data for HR and predictive HR analytics are being discussed as necessity for the future of HR organizations.  But what is big data for HR?  Who is using data smartly?  This program that we’re going to be launching this year is dedicated to that discussion by talking with both leaders and recruiters in the talent acquisition space on how data is affecting their organizations, their performance and their careers.

Andrew, thank you so much for speaking with us today here on TotalPicture Radio.

Andrew:  My pleasure.  Thanks for having me, Peter.

Peter:  Let’s start at the beginning talking about big data for HR.  Again as I said in the introduction, this can be a very confusing topic for folks.

Andrew:  It’s not a very well-defined topic and that’s part of the problem is that I think recruiting, especially in HR, there’s a lot of terminology that floats in the marketplace and there’s not a place to necessarily have definitions or a glossary, and big data for HR is certainly not absent from that list.  Really because there’s no definition what you’ve got is people struggling with trying to get a hold of their metrics, understand what metrics to track and then realize that big data does not equal metrics.

Big data, for lack of a better definition, is data that you can’t easily necessarily get your hands around or a score card and it’s constantly changing.  So you need powerful analytic tools or intelligence tools to have you sort through what is not thousands of lines of data, but potentially millions.  That’s what’s, I think, a little scary to the marketplace is they don’t know where to start because everybody’s starting at the same time, and it’s well related to the cloud in that we know that we can store unlimited amounts of data now and there’s so much noise that you can’t quite get your hands around it.

 Peter:  A lot of the research that I’ve been doing for this interview I was surprised to learn that there is so much what they refer to as dirty data out there and that is a real challenge in just getting into this stuff. First you’ve got to clean this stuff to make it usable.

Andrew:  Part of that comes from really over the past 15 years human resources has spent a lot of time implementing systems of record, whether it’s a performance management system or it’s an applicant tracking system, CRMs have been the rage for a number of years now.  And then with all the social media, what we are is we’re engaging a lot of data and a lot of systems at once but because we’re not necessarily looking at that data, there’s really no one checking on what’s going in or how often it’s going in.

And so now what we have is you’ve got analytic tools, business intelligent platforms that allow you to look at that data relatively easily and then you realize how dirty it is.  You realize that’s what’s been put into it is old or should be archived or isn’t effective and now you realize that I need to build processes with my teams such that they don’t put in dirty data because I can see the data all the time and that’s a real problem for a lot of HR organizations right now.

Peter:  In the brief you sent me for this interview you talked about that there are really three things at its core which define the whole big data business around HR and recruiting, which includes things like predictive analytics.  Can you talk a little bit about that?

Andrew:  I think I’ll quote or steal some ideas from other thought leaders in this response.  There’s been a lot of talk about big data and one of the things that other thought leaders are talking about is how there’s data that you look at transactionally or in terms of execution and it’s a little bit more reactive.  Then the next stage is linking it to strategic HR organizations and then using it to make some other types of strategic decisions.  Then the next step is talking about more talent management and workforce analytics and then using it for predictive analysis, and I think that’s correct.

Predictive analysis using big data is one of the keys and that’s really where people who are doing workforce analytics or workforce planning or in HR really need to spend a lot of time so that they can make different types of decisions about the workforce itself and what they’re going to add or how they’re going to resource that workforce using big data.

That leads us to the second part, which is making sure you’re fusing in data from other functions within the business.  HR is spending a lot of time trying to figure out its own data but it’s one of the few functions that permeate throughout an organization.  So you need to add customer service data, sales data, IT data, customer data and that’s a new place for HR.  They haven’t been asking for that type of data ever and now all of a sudden they’re going to start asking for cost and quality and information about customers and the organizations, other functions are going to say “Wait a minute, why am I giving you this data?  Why is it related to HR?”  That’s going to be interesting problem for a lot of different companies and there isn’t a roadmap.  This is a problem that’s going to be unique to each organization because it’s politically loaded.

I think the third one is then you’ve got to use the first two.  So you’ve got to use this concept of predictive analysis and having competencies around doing that and then gathering data from all the different organizations and then making decisions about how you’re going to structure the workforce in the future.  When you understand what’s going on with your customers or how engagement is going on with your customers because of HR processes or you understand how quickly you can recruit because of recruiting processes as an HR process, you can make decisions on am I going to develop people, am I going to acquire people, am I going to outsource, am I going to hold off on this project, are we going to develop this product, or oh wow we have a lot of different skill sets we didn’t realize we had across an organization, we should develop this type of product.  And that’s really where HR wants to be and it really has the ability to get there but you have to take the first few steps.

Peter:  Something that I’ve been concentrating on over the last couple of months is around the candidate experience and employer branding and that is a key element I believe to what we’re talking about here today and how HR and recruiters need to work with the marketing and advertising departments who are out there perhaps managing the brand on the social networks, the Twitter feeds and the Facebook and the LinkedIn so that they really understand and can participate in the conversation that the marketing departments are having out there with their customers and potential candidates.

Andrew:  It’s interesting you say that.  That’s absolutely true,  especially if you’re a retail oriented brand.

Peter:  Yeah, like a Verizon.

Andrew:  Exactly.  So if you’re like a Verizon or if you’re a bank or anything like that where you have lots of customers most of those organizations have field operations where not only are they staffing tens of thousands of people, they have hundreds of thousands of customers all of which who could be employees and vice versa.  All of them could be customers.  So the connection between marketing data and HR data has the opportunity to be very, very tight.

In fact just in the last two weeks I’ve had a number of senior talent leaders call me and say “We’re thinking about restructuring how marketing and recruiting or our employment brand people interrelate.  Should I lead that effort?  Should marketing lead that effort?  How does that relate to our data?  What should we give up?”  I appreciate the question because it’s very relevant right now that even I get questions like that from senior leaders on how they’re going to engage with marketing in regards to social media but really in regards to data.

Peter:  Back to this whole conversation about predictive analytics there are organizations out there now building software that can go out and scour the social networks and predict when an employee is going to leave their current role and perhaps may be looking for a new job.  A lot of the things around the predictive analytics that I’ve read about are if there’s a particular function within an organization where there’s a lot of turnover you can use predictive analytics to go in and really analyze that and figure out why is there so much churn in this particular department and put some metrics towards it so you can realize what you need to do with your pipeline to be able to fill those positions as they become available.

Andrew:  Well it’s interesting how the predictive analytics are probably going to challenge a lot of the HR processes that we have now.  If you look at analytics and you know that there’s a tendency for people to leave after a certain amount of time or you know that there might be a group of people who are going to leave, the natural reaction is we’ve got to keep our pipeline up.

Well the thing is with big data, actually that may not be the way you think about it.  What you might want to look at is how does this impact us financially, how does this impact customer service and then how does this impact service or product development?  Maybe what this means is you change your projects, you change your initiative to accommodate that natural turn rather than trying to actually thread the needle of lowering retention.  If you actually knew that your retention on an associate for an example is normally 24 months and people say well we need to know month 21 through 24 we have to be pipelining I challenge that and say if you really know these things then why not build a project plan around initiatives that are based on 24 month cycles, rather than just focusing on the fact that you need to replace the team member, maybe you change the initiatives so the team members when they leave the projects are over.  That’s where big data really comes into play.  It’s almost impossible to do that if you only look at HR data, but if you look at business data you can start doing those types of things.

Peter:  Andrew, this whole conversation around big data, it didn’t creep up on us; it just dropped on our heads, right?  It’s sort of like the conversation around mobile, where two years ago nobody was talking about it and today it is the topic of conversation.  How has this happened?  How has this evolved so quickly to be a concern and something that HR and recruiting specifically really need to pay attention to?

Andrew:  I think how it manifested is believe it or not I think the recession had a really good push for data because it forced major corporations to squeeze their resources and they had already been operating in a somewhat lean environment and they squeezed it out and they realized that they have all this data lying around, they have to get a hold of it and of course then people started using it and they get a competitive advantage.

Right now talent will continue to be a shortage.  Talent will continue to be tough to get people, but we have lots of data.  So it’s a natural resource to start investigating and with things like the cloud and better security and quite honestly the fact that you can control it, it’s a lot easier to control data than it is to control people, by the way.  It just is.

Peter:  That’s true.

Andrew:  So you can look at it and you can say let’s shape it and let’s mold it, okay this is a resource, let’s use it and we just haven’t been.  So it doesn’t take a lot in a downward spiraling economy to look at your resources and try to squeeze everything out of it you can.  So that’s how it kind of dropped on us.

What was the second part of your question?

Peter:  The second part of my question is what are some of the things that recruiters and HR professionals need to do proactively now to start implementing some of the things that they need to do to get their heads around this and to get resources in-house to manage this data?


Andrew:  I think the first thing is don’t make the mistake of immediately transitioning to different systems.  That’s a red herring and I think that it’s easy to hear that and say ‘we’ll change your system of record and then thereby will change your reporting problem.’  Well I’m here to tell you that that’s not true.  That data could be exported and pulled into any number of intelligence platforms and moved around and all kinds of companies have internal organizations that can help weave that data.  They can use the cloud.

So don’t do the silver bullet thing, which is we’ll get a new ATS and that will solve all of our data problems.  We’ve tried that before with other problems and that’s not necessarily the case.  What we need to do is we need to first allow for time, or you invest in time where we’re going to look at analytics, we’re going to make decisions based on data and we’ve got to slow down enough to actually stop, look at the data, analyze it, make decisions, have conversations about it.  And the problem is we don’t have a lot of competencies out there that the people have invested in to do that.

So the first thing is slow down, see what you have, start making decisions based on it, see how valid it is and start using the right kind of governance on a regular basis to do that.  Start making the decisions.  That’s the first part of, I think, the step.

The second one is get the rest of the business involved.  If you only focus on HR data, you’re going to have a lot of deaf ears in the organization because they’re saying “Well, that’s your data.”  Even though I own it and it’s in my function what’s going on in my business, if you start linking in financial data, operational data, safety data, those types of things and say “Look, we’re doing this HR process and we’re seeing these things over here improve,” you’re going to have tremendous credibility much more than you would without this and you’re going to get a better audience.

And so that’s what we need to start doing.  And those are two really big things to do.  I mean that’s probably enough.

Peter:  Right, absolutely.  You brought up something here that certainly has been going on for the last couple of years and that is the Oracles and SAPs of the world are buying up everybody.  There’s just a tremendous amount of consolidation going on with the ATS providers and the big data base systems, the IBMs, the Oracles, SAPs in an attempt, I guess, to try to be able to bundle everything together in one nice little package and market to their clients.

Andrew:  It’s a very smart move.  I like what SAP is doing from an on-premise perspective and how they’re thinking about keeping things safe and secure.  What Oracle is doing makes a lot of sense with their HCM Fusion products and those types of things, these are all highly valuable.  Having a system and having it all work together is highly valuable.  Again, though, you have to still use the governance – and I’m not a big fan of that word but I’ve used it twice – but you have to have that discipline to make sure that you’re not making dirty data with that system.

These are very, very expensive and you have to realize that even if you go to those types of systems or if you consolidate to a handful of systems, recruiters especially use all kinds of other systems and methods to engage data.  Just social media alone, you have to aggregate that data.  That’s not going to necessarily aggregate an Oracle for you very quickly and not to pick on them.

LinkedIn is another example.  LinkedIn has really been playing around with who can pull data out of their system and pull it into another system.  They’ve been struggling with trying to figure that formula out.  And so just those two things – social media and LinkedIn – recruiters spend as much time doing those things as they do in an ATS.

And so if you can’t integrate those key things for recruiters anyway, that’s why I caution this whole concept of moving to a different system of record is use what you have and if those things aren’t working because you’re not inputting good data, or people aren’t using it, well, then get a different system of record that’s more innovative.  But the consolidation makes a lot of sense for companies that want to make this move.  It’s a significant investment, the chief executive can really wrap their hands around it, and he or she can say to the organization ‘we’re going to spend X million dollars to get this done.  So here’s that investment.  Don’t mess it up.’  And you can hold some accountability because of the spend.

I think it’s a strong move what SAP has done, I think what Oracle has  done in combining and acquiring these businesses and success factors and everybody else consolidated, it makes a lot of sense, but you’ve got to have leadership kind of say if we’re going to do this, we’re going to do it right.  What’s the plan?  How are we going to translate that down to the individual contributors, how are we going to monitor it,  because it’s easy to spend a few million dollars and say it’s “working” (as I throw up air quotes on a podcast) but how do you know?  You really have to know.

Peter:  And certainly, one of the big dilemmas I’ve heard from HR leaders for the past 5 years is that data is all over the place.  It’s on Excel spreadsheets, it’s on systems that don’t talk to one another and they’re just trying to figure out how they can somehow get everything to talk with one another and play nice together, right?  Is that a right assessment?

Andrew:  That’s a very good assessment and the fact is that from a big data perspective, it’s likely not all going to sit in one system anyway.

Peter:  Right.

Andrew:  You’re going to have data coming just simply from a vendor, and that would have to integrate with your system of record or it may or may not.  And so the thing about data is it really separates into two pieces.  You have fixed data – let’s call it that – and these are things that are in columns and rows.  You can basically export it.  It might be tens of thousands of rows and tens of thousands of columns but we’ve input data into a field and that could be exported into you’re a fancy Excel spreadsheet, basically.  There’s lots of that.

And then there’s the variable data which is data that’s usually connected to that system of record like a résumé.  So we attach a résumé and maybe it is parsed and then maybe it is put into fixed data but many times those are just documents sitting in the tool.  The thing is you actually can go through variable data and you can pull it altogether and you can scan it, you can pipeline it, and you can move it around and you can do the same thing with fixed data but you have to realize that this is a straightforward process, very similar to climbing a mountain, though.  You can kind of see it but you have to have the right tools and the right equipment, the right training to get there.

I think the concept of we can never do it and I’ve heard that a few times in the last few weeks from some of the HRO providers we talked to and they say “This is impossible.  We’ve got 300 companies working with us and how are we going to see all their data?”  You can see it but once it’s there it’s great but you’ve got to make the effort to get there.  You can do it.  It’s just a process that most HR organizations say we can never do, but of course you can.  It’s just data.

Peter:  Let’s talk a little bit about outsourcing and the contingent workforce which with some companies now is up to 20% of their workforce.  That’s another trend we’ve seen over the last few years with this recession is companies aren’t hiring FT employees.  They’re bringing people in on a contingency basis and if they don’t know who their FT employees are, they sure don’t know who their contingency employees are.  So I think that’s another reason and another excellent way of looking at setting up these data systems so you can really understand who it is you’ve got working for you and what their role is and what their expertise is.


Andrew:  Absolutely.  When you think about big data for HR, you certainly don’t want to limit the data associated with just your employees.  You really want to look at what data do we have for our contingent workforce, for our vendors, or services that we outsource to in general and also our technology that might be replacing human capital, right?  So those are our resources that are doing work.  And if you can cull data from all those places, that’s when that predictive analysis, pulling data from other functions and then allowing for making workforce decisions, those three things we talked about earlier – if you absolutely should pull that data.  In fact, that’s how you’re going to really make decisions as an HR leader in the foreseeable future is I’ve got all this data from who worked for us in general and here’s how we should then reinvest in that workforce and here’s how we know because we’ve got the data from the contingent labor group.  We know what their capability is.  Let’s not go ahead and hire 25 people or 2500 people to do this.  These guys are doing a better job than we are.  The price is a little bit of a premium but on the 5-year horizon, it looks like we’ll make more money.  So we’re going to go ahead and we’re going to outsource it deliberately.

We’ve used data to make that decision because we have daily performance of that group.  And that’s where HR is really going to move to.  We’ve always talked about let’s have a seat at the strategy table, right?  HR is everywhere; 70% of a company is SG&A and there’s a lot of people working at companies and HR is going to be in a really good spot to make those kinds of decisions soon.

Peter:  Talk to me a little bit, Andrew, about your applications.  You not just do consulting work but you have developed applications to help companies organize this data and give them meaningful reports.

Andrew:  A few years ago as we’ve been doing efficiency consulting within HR and within recruiting, we would go back to our customers and we said how are things going, and they were really struggling monitoring their progress after we had put in an initiative, because we leave. So they have to keep running and they’re trying to keep it running and they kept on fumbling on their data.

What we decided to is make it very easy for them to pull the data both fixed and variable out of their systems and into a location that they could dashboard it and scorecard it, look at key performance indicators and that’s actually one of the first key steps is having an application that allows you to feed in your ATS, your HRIS, customer service data, sales data – whatever kind of data you need to make those HR decisions and have what we call data visibility, right?

So you really can see it.  Now, it might be dirty but at least you know it is and so then you start to clean it because you can see it, right, not just hope it gets better.  So we have an application that does specifically that and we have to create that solution not only to operate in the cloud which is very popular but we also had to make it operate on-premise which is an older term that people have been talking about SAS or software as a service for so long, we had to make it on-premise as well.

So the cost structure is the same whether you do a hosted solution or you do on-premise but there are customers who are going to say “We’re not going to throw any sales data or customer service data or customer data through a firewall.  We’re going to keep that in our own security.”

So we had to develop actual software with our partners Metric Insights.  So we had actually developed that with them and make it so that they can put it on a server that they control which is really unusual.  You don’t hear about that a lot right now but security is probably as important if not more important than just the analytics now.  I mean you can appreciate that these applications, whatever we build, if it controls all this data, you can see all this data, it’s an easy target and you’ve got to have the right kind of security settings around it.

Peter:  Well, aren’t there robust security systems for cloud-based systems as well?

Andrew:  Oh, of course.  I mean you can even do… It’s not said very often but the concept of a private cloud where you can have your cloud solution, you can have that account, those are very, very secure systems but the thing is is that some systems that are legacy are only on-premise and so security organizations have decided to not push that through the firewall and maintain it behind the firewall.  And so some cloud solutions don’t allow for that.  And it’s harder for an organization.  So we’ve just decided to make our application very portable because we’re not going to dictate to a major corporation with 25,000 locations or 5,000 locations what their security setting should be.  We’re just going to say what are your security settings, and you can put this any way you want.

Peter:  For companies that aren’t embracing what’s going on today, this is going to be very disruptive for them and they’re really going to be behind the 8 ball if they don’t start really putting some resources and some serious thinking behind how they want to manage all of this big data.

Andrew:  Yeah, I think it is.  I think it’s disruptive for a couple of reasons and I like that word.  The first thing that’s disruptive is the technologies themselves that can analyze this data.  They’re not necessarily because you can get data from anywhere and kind of move it anywhere, you may not need to change your system of record to see your data.  So that concept itself is disruptive to HR that you mean I don’t need to change my system in order to see this data and get everything I want?  No, you don’t.  That concept is disruptive.

I think why also it’s disruptive is this is probably as important as the talent shortage or the war for talent to really age myself that we’ve been talking about for years.  It’s really one of the last resources that we haven’t really squeezed on and companies are looking for that competitive edge are really going to get a hold of big data and understand their workforce, understand engagement, understand their processes, they’re going to see lots of savings.  The ones who decide to not do this are going to find themselves having internal questions around their pricing strategies.

They’re going to say we feel like we’re being squeezed.  Our profitability is suffering, because the competition is going to be able to pull those resources together, squeeze the data, pull their markups down and become more price-competitive in the marketplace and they’re going to get more business.  It’s just that simple.  Those who don’t embrace big data are… they might get the work and that’s great but ultimately, people who control the data control information and that’s always a wonderful thing.

Peter:  What haven’t we discussed today that you think is important for our audience to know about big data especially people who are in HR and recruiting who are trying to figure out the best way of approaching this and wanting to go to management with some concrete plans of how they’re going to implement this?

Andrew:  I think the first thing is this is not a leadership initiative – big data.  This is a competence that needs to be built actually more at the individual contributor level than at the managerial level.  If an organization understands that they’re using big data to make decisions, the individual contributors are the ones who are actually controlling the data the most because they’re inputting the data the most.  And so there has to be a push by leadership and respect with the individual contributors that what you’re putting in the system has to be clean, that you’re not wasting time, that you are being measured based on data.  So understand that.

We didn’t talk very much about that but that’s what’s important with recruiters is using their systems the way that people are telling them to use them or if they’re not getting told the right way, say to leadership “Hey, this is not the way to do this.  I’m an expert.  I use this system every day.  Here’s how we should do it,” because then it’s going to get revealed properly in the data.  It’s really easy to go into a boardroom and say “Oh, we’re going to do big data and we’re going to pull all our systems together,” well, and we can help you do that, right?

I mean, but if the data’s bad, well, let’s say if someone’s entering the data portal and you got 400 recruiters working for you.  I mean they’re all going to put it in right.  Otherwise, this all kind of falls apart.  I think the other thing is leadership has got to spend time building competence and analysis.  They’ve got to spend some time with – data scientists are a hot topic right now.  We got to get…

Peter:  And there are not a whole lot of them out there.

Andrew:  There’s not a whole lot of them out there.  I would say there’s a lot to be said about that from a very, very big data perspective but just understanding key performance indicators, how to make decisions, using analysis on a regular basis… We shouldn’t shy away from allowing leaders to make mistakes, experiment, talk about the data, and then start using that data on a regular basis to make better decisions, that competence has got to be built.  Once that’s built and leadership is going to start to permeate down to the individual contributors who are going to say hey, if I want to be a manager I have to understand data, I have to understand how to use it, how to make decisions with it, and then they’ll want to know more about it, too.

Peter:  Andrew, thank you so much for taking time to speak with us on TotalPicture Radio and I look forward to working with you on the big…