1. Home
  2. Insights
  3. Podcast

Advancing the Skills-Based Workforce Model Through Data and AI

As more and more executives are exploring the advantages of building a skills-based workforce model, the need for reliable data and next-level AI technology has moved to the forefront. Brightfield Co-Founder and CEO Jason Ezratty and AGS Senior Manager of Data Strategy Tim Johnson join host Bruce Morton to deep dive into the future of work through the crucial aspects of data and technology. They also delve into the immediate cost savings that can be found with greater transparency and visibility of your extended workforce, statement of work (SOW) creation and needed skill sets.

Transcript:

Bruce Morton: Allegis Global Solutions (AGS) presents a Subject to Talent podcast, a hub for global workforce leaders to unleash the power of human enterprise. Thank you for listening in as we explore the most innovative and transformational topics impacting business today.

Bruce Morton: Hi, I'm Bruce Morton, the host of the Subject to Talent podcast. Today I'm joined by Jason Ezratty. Jason is the co-founder and CEO of Brightfield, a world-leading workforce data platform. I'm also joined by Tim Johnson. Tim's the senior manager of data strategy here at AGS, and he's a data and business intelligence leader with a focus on strategic data partnerships, enterprise analytics and data management. So, I'm very excited to welcome you both to the podcast. Welcome.

Tim Johnson: Thank you.

Jason Ezratty: Thank you, Bruce.

Bruce Morton: So those regular listeners will know, we always kick off each podcast by asking our guests, how did you get into this wonderful industry of ours? How did you start in the workforce industry? And what led you to data generation and analytics? So Jason, let's kick off with you.

Jason Ezratty: My entry to the staffing industry was very much by accident. I found myself in need of a job post 9/11, and a number of my colleagues from a technology strategy firm had gone to this VMS company, which I had never heard of the category. And so I joined my former colleagues. I figured these are great people to work with and who cares what the rest of it is? But it was about three months in that I realized that this was how the world of work was changing and that this was not just my next job, but the job and opportunity to really change things. There's no doubt that the percentage of the contingent workforce was increasing like crazy, people were having trouble finding talent after Y2K, and I just realized that this was going to change the entirety of the world of work.

Bruce Morton: Awesome. Thank you. Tim?

Tim Johnson: Yeah. Thanks. Back in the day when I started with Aerotek, I started out in back office operations. It seemed like every job that I took, I gravitated towards data. And so, I spent five years with Aerotek, then I went over to the Allegis Group IT department to help with an ERP upgrade and led and managed lots of different work streams in the course of that five years and found myself wanting to get back into the business. So I transitioned over to AGS, and for the past 18 years have been developing analytics and business intelligence solutions to help our customers, our suppliers and the leaders within our organization use data to drive decision making. So it was really a passion of mine all the way from day one.

The business intelligence platform we built is now a critical component of Acumen® Intelligent Workforce platform. And over the past two years, like you mentioned in my intro, I've been spending my time on data partnerships in order to gather the best partners in the world around global data and analytics capabilities and really ultimately to shape our industry and how they look at data.

Bruce Morton: Great. Thank you. So, Jason, the last time you joined us, which has been a hot minute, you spoke of the Economics of Workforce Intelligence, which at the time was a relatively new partnership between AGS and Brightfield. Can you tell us a bit more about the partnership and how it's developing?

Jason Ezratty: Yeah. I like to think of it as a best-in-breed relationship where there's only so much that we, as the stewards of data and generating AI that infer market relevance, what's happening in the market at large, and then to have that connect with the necessary actions to make optimization come true. And so thats really how we see the chocolate and peanut butter coming together in this relationship. And it matters not only for us internally for how we achieve efficiencies, but whats most exciting is how it shows up for the customer. Problems that were previously intractable, things that could not be accomplished, only talked about, like gaming efficiencies and SOWs are now our favorite case studies. So it's been a very exciting relationship, but one that's bearing the fruits of its labor.

Bruce Morton: Great. Tim, anything to add to that?

Tim Johnson: Yeah. As far as the partnership, it's been a six-year journey between Jason and myself, trying to unlock the value of the data and the documents that are hidden sometimes in file cabinets, sometimes they're embedded and stored within various systems. And working over the past years to figure out the key to unlocking all that and getting those nuggets of gold out of them has been a really fantastic journey. It's been challenging at times, but we're really starting to see how these prediction machines and machine learning models can help us move fast and uncover things that we had envisioned. I'm not going to lie, we envisioned seeing those in there and getting to this point, but it's now way more efficient than we have been in the past several years. So really exciting times.

Bruce Morton: Great. And anybody that is in our industry or associated with it will have heard the term skill-based economy that we're now in or we're now entering. So in an ideal world, we're looking at the future of work and how work gets done through that skills-based lens. So how does data collection play a part in that, moving the industry forward? Because there's a lot of talk about it, but how do we actually operationalize that? I'd love to hear from you both on that point.

Jason Ezratty: Timothy?

Tim Johnson: I'll let you go first, Jason.

Jason Ezratty: I guess the starting point, just launching off of the SOW point, if you know that your deals are being priced and salespeople, their job is to price as high as they possibly can, they're charging for the most expensive skills. So you should know what that premium cost is for the most expensive skills, and you should know when are those most expensive skills in play versus not.

The problem of it from a data perspective is it's on a long list of things of, "It depends." It depends on what database languages there are. It depends on how much experience they need in this industry. It depends on which of these programming languages they have, and do they have them together along with a certain degree or certification or something else like that, or top secret clearance for certain customers. And understanding what is the contribution of all of those pieces is a natural AI data science question. It's where all of these many parts have some influence, at least sometimes, to some degree, and how to make sense of that.

So the recursive nature of machine learning that constantly checks do we understand the pattern to see how each of these individual skills play is exactly what we've been nurturing and maturing over the last several years. So we can differentiate not just the cost impact, but the time expectation impact on every geography basis. If you're not at the skills level, you're missing the mark by at least 20%, because you're in the middle of something, but you're not sure what you're in the middle of.

Bruce Morton: Right, yeah. Tim?

Tim Johnson: Yeah, couldn't agree more. The visibility and access that we need to have data in order to have a confident opinion and perspective on things is key and critical. And between AGS and Brightfield coming together, it really is that answer. It's more data being pulled together, we're feeding it into the machines, it's giving us more confident results. And in this case, more is more as long as you really trust the data that's going into the machines. And so confidence between our two organizations that we're getting that right and we're feeding it into the machines is giving us greater confidence.

It's also doing another thing for us. It's really helping us improve our processes and shining a light on the areas of the inputs and the outputs where we focus attention. The machine learning models that we have in place is giving us a chance to interrogate transactions on a daily basis. And so, that's giving our operators the chance to say, "Hey, I think there's an opportunity here. We can improve a process." Or "Hey, I need to take action here because there’s an opportunity," and ultimately that ends up being an advisory position for us. And so that's really a goal of ours at AGS is to be that strategic advisor for our clients. And again, a combination of having more data, having machine learning models that make us more efficient, and then having people that know how to use them, that's the special sauce.

Bruce Morton: Right. So I guess it's that combination, as you say. In this case more is better, but only if you can actually analyze it in the right way and truly understand, "What is it telling me to do and how can I act on that?" Would that be fair?

Tim Johnson: It puts it into perspective. And if we're talking about skills, are we extracting the skills from our acquisitions? Are we extracting skills from the candidates that are being submitted? And are we matching and merging them properly? So if we're all capturing that and we're all looking at it, we all speak the same language. It's absolutely critical, a foundational element of this all.

Bruce Morton: Right. So I guess that's about paying the right amount of money for the right skill, but perhaps not paying for skills that you don't need. Just that too high a level, as you said, Jason. If you're looking at the job title, do you really need all those elements? So drilling down deeper.

Jason Ezratty: And you could also have a long list of skills that don't necessarily imply a greater cost. So it's really making sure that you're paying the right amount for the right things. And the flip side of that is true as well from a supplier's perspective; how to defend your price based on the empirical information, the true data like Tim is talking about. So in a world where good information is available, people should leverage it and conduct their business off of it. It's really that simple.

Bruce Morton: Yeah. So to bring this to life, can you share any examples? Obviously don't mention any company names, but can you share any examples of how this partnership is starting to impact our client's businesses?

Jason Ezratty: I'm happy to. Go ahead.

Tim Johnson: Yeah, I'll go first, Jason. So over the past year, we've been sharing anonymized data between our mutual clients and putting that to action on training machine learning models. And when we shared that data over the past year, we had a lift in our Canadian-based business. We have several clients in Canada, and through the sharing and working with Jason's machine learning team and their data teams, we added 17% increase in the amount of titles that we can cover in that region. So that really showed up strongly for us. And you say, "17%? That seems relevant, but is that massive?" When you're talking about a mature market, it is really relevant. So, in a very mature MSP staffing market, we were able to increase coverage by 17% between the data that we are able to bring to the table about that. So outstanding for us as far as coverage, it improved our confidence in many areas. And so what does that mean? That for our clients, we're able to represent more confidently the rates that they should be paying versus what they are paying today, and it informs our rate cards. And so overall, it gives us a chance to be confidently in an advisory position with our clients.

Over to you, Jason.

Jason Ezratty: I think it also gave us an opportunity to better understand the AGS way and how to reflect that in the data so that AGS programs are able to be set up in ways that are natural to the true expertise of the AGS market analytics team as well as the program operations teams themselves. So it's not just a generic representation. AGS has an opportunity to really tune that to the AGS way, which is exciting for the customer and for us.

Bruce Morton: Right. And I guess back in the day, I remember a time when you could update a rate card annually. I'm guessing it's more common than that now, I know it is. But what I'm trying to get to, is this something that we're creating through this partnership that we're excited about, or is this being customer-demand / solving the customers’ problems”-led? What's been the driver to elevate the partnership in the way we are?

Tim Johnson: I think ultimately, yes, you're right. Traditionally there had been annual rate reviews. Now we're going to get to a world where it doesn't require the rigor or the scheduling of an annual rate review to give our clients advisory around the positions that they're staffing. So as I said before, there's a chance to inject that into the processing of our daily requirements and then identifying which ones have the greatest opportunity for being filled, whether it's the fill-ability model that we have with Brightfield, whether it's the rate benchmarking capabilities. We're giving our clients a sense for, "Hey, this is an area of concern. Markets may have shifted. We may be able to source this in a different location. There may be a better place to fill this position." So all these things come into play in arming and equipping our people that are doing staffing, the staffing specialists or even our market analysts that are setting the rates for our rate cards, giving them early warnings or early indicators as to where they may need to pay attention. We're not waiting for an [annual] rate review to have an impact on a client.

Bruce Morton: Got you. Got you. I like that. Anything to add to that, Jason?

Jason Ezratty: Yeah. I think when building a rate card practice off of granular global scale rate benchmarks, more data is always going to be better. And so in the case that Tim is mentioning, in some cases, because we have a variety of rigorous criteria that you have to pass on a threshold basis, bringing a lot of that data in enabled us to get above some of those additional thresholds. And what that means is not only geographic precision or breadth, but also adding greater precision to all the other ways that we look at the information. Are rates on the rise? Are more suppliers coming into these marketplaces? Are things protracting and taking longer? Are companies having trouble being compliant with their rate cards? It's not just what is the central tendency of the rate and how does this company compare on an average-to-average basis? It's all of these things that come together to equate to a picture of healthy spending, not just price, but the whole picture. Because if it's only about price, we all know customers are unhappy and spend goes away.

Bruce Morton: Right. Great point, great point. And as organizations are moving towards what we today just call this Universal Workforce Model, in other words taking a holistic view of the work that needs to get done and looking at all the different ways and types of talent to do that, automation and the freelancers, gigs, employees, service providers, et cetera. I know a couple of the challenges that that is creating, in a good way, [are] the problems we're solving. We're getting asked regularly by our clients, "How do we expand globally and how do we get that holistic view so we can give our business the options of the different ways of getting work done?" So, a big piece of that, of course, is how do we bring that services spend under the umbrella of an MSP program, as an example, and get greater visibility and flexibility? Can you just expand on some of the things we're doing in that space to help our clients out of those challenges? I'll go to you, Jason.

Jason Ezratty: Yes, happy to. I think given the idea of how to understand how to wrap our arms around the complexity of total talent, which was the old form; universal talent. So all the many different forms that can come together, it's such an enormous topic. It tends to either die or be forestalled under its own weight. And so if this is now the 15-year conversation, what has made it go so slow? In my belief, in my understanding of it, having studied it so closely, it's just the overall weight of it that creates this inertia.

And so what we've found to be most successful is to size it much more appropriately, shrink down the opportunity so it doesn't have to be the enterprise and everything. The old saying, "Think globally, act locally," applies here perfectly. So we're working with an organization that has 10,000 SOWs that they want to have evaluated, but going after all 10,000 at once is not going to be very helpful. Saying, "Bring me the next 200 and let's show you what those opportunities are," and making those opportunities real in real time, working with the AGS counterpart to show how the action comes across. It's not just deal criticisms of your SOW, it's things that can be changed and result in realized savings or other efficiencies in real time. That's the action. You have to start the flywheel so that inertia is in your favor, not something that feels impossible to overcome.

Tim Johnson: Yeah. For AGS, over the last 10, 15 years, it started out with collecting data around the resources that are the extended workforce that's coming through services engagements and really having some visibility into the headcount. So we call it headcount tracking resource management. We really wanted visibility into that. We quickly gathered information about those people; what suppliers had them, maybe what role they were playing. And so as we got that information and visibility, it became apparent to us where are the key pockets. So as Jason was saying, where should we focus our attention? Rather than trying to tackle them all, where are the greatest areas of opportunity?

The next question that we started getting was, "Well, what is their job title and how does that compare to our staff augmentation business? What rate are we paying for them?" And so we expanded from just counting heads to gathering more intelligence about these workers that are on site and working through services engagements. That was the expansion of that. The visibility it gave us was fantastic. What we could do to action it is where we're really sitting right now. So how do we gain way more intelligence about it? So we run people through SOW admin. So they're running through, we're capturing project level details, we're capturing more resource level details, we're tying them together and we're understanding rates, we're understanding milestones, and we're starting to just gather more and more information. And that's the phase we're in now.

The next generation of that, which we've been working on the last five years, six years with Jason and team, is, "All right, there's all this data that's in these documents. How do we merge that together with all the transactional elements?" And now we're completing the picture and it's giving us that insights. The keys to this is how do we leverage the services business that we're understanding now through transactions, and how does that compare to our contingent workers, our staff augmentation? Can we start calling them similar things or the same things? These are all talent roles. They all have skills underlying jobs and occupations underlying that. And can we come up with the ability to connect all these things through a taxonomy, whether it's occupation, job titles taxonomy and the skills taxonomy? That is the underpinning of all of this so that we have the visibility we need in order to do comparisons, to draw analytic insights, and to actually have actionable things [to create an] increase for clients. Those are all the things that are working in concert together and that's the journey we're on.

So it's exciting. It's all coming together. I think this is a super relevant conversation for that in knowing where we're heading.

Bruce Morton: Yeah, and I think it does mean, for the first time, I guess really, we can compare apples with pears. It's always been the challenge, right? The black box of services. It's no longer a black box, it's a glass box, I like to say. But talking of buzzwords, what do you say to those prospects that say to you, "Well, can't I just ask ChatGPT what I should be paying for an app developer in Palo Alto this morning?" What's the response to that?

Jason Ezratty: Yes. But before I criticize ChatGPT's ability to execute our business, I think part of the apple versus pear problem is that in reality, when you look at the organic nature of it, not just what we want to label this thing as an apple, it turns out there's such a thing as an appear, and an orapple. And that's actually how things exist in nature. So we've now developed the ability to understand that inherent diversity of talent. There is no [one] such thing as a software developer. There's many, many such things as software developer. And we stick them in a box because that makes our lives easier, not because that's an actual representation of nature. And so what we've had built is the ability to better interrogate the actual dimensionality, all those many attributes that make up the reality of nature; one entity, one organism at a time, one job at a time, as well as looking at the entirety of the ecosystem.

And so ChatGPT, number one, I love. And all of the large language learning models that have come with the generative AI wave of this year has made my life a lot easier from a roadmap perspective. We can do a lot more, a lot more quickly. However, it's important to understand their limitations. Number one, they're not trained with our transactional data that we all get to see. So no matter what, they're not going to be able to see what's actually happening in the transactions. They're only gleaning from what's read upon across the web, in that respect.

Then the other thing, if you actually ask the question, what you'll see coming back are all of the great conditionals. So it's smart about what are all the conditional things that you need to be worried about in order to precisely answer the question of, "What is a rate for a blankety blank and blankety blank?" All of the attribute type questions that it takes that it would not then be able to render an answer.

And so I think large language models are good at expediting some of the steps that we do in data management, but not replacing. And especially anything that's in front of the human being where it's trying to create that human interface, something that's more conversational, that's fabulous. But if you're looking for discrete business answers that relate to economics, a much greater degree of precision is required, as suggested by ChatGPT itself.

Bruce Morton: Great, thank you. So AGS and Brightfield, we both have very busy months coming up, collectively as well as individually. But we've got obviously the SIA's contingent workforce summit in Dallas in September, I think 17th and 18th, or 18th and 19th. But we're also partnering on a workshop in September for our joint clients, helping them to avoid the pitfalls of misclassification. So on that topic, what are the key points you want workforce leaders to take away from this session, plus those coming up, particularly around the misclassification

Jason Ezratty: Sure. Misclassification is an abstract term. So basically this workshop is trying to say not only where there's misclassification and why you should care about it, but also what can be done about it. And I think in the past several years when people have addressed the topic of misclassification, it's remained an abstraction. It's not an, "And therefore, here's how to make things better." What has me so excited about the content in this workshop is it's the first spectrum all the way to realizing the results of benefit, whether that's properly classified workers in the moment, in the future, better reconciliation of processes and systems to make that happen in the first place, and just better deals, taking out the unnecessary premium costs that tend to be a part of a lot of these. And it's that tactical win that helps pay for the strategic wins. So for me, that's the game-changing difference of why this workshop is exciting. It gets to those realities. It's not an academic discussion.

Bruce Morton: Great. Well, thank you for that, and as we look to wrap up this podcast, we normally ask our guests the crystal ball question. I'm going to ask that to Tim, but I've got a special one for you, Jason, so hold that thought. But Tim, let's come to you with the crystal ball. So what do you think, three, five years ahead, you pick a timeframe, what impact has this partnership had on the industry?

Tim Johnson: Yeah. Thanks for that, and I appreciate the opportunity to answer it. I think we're going to have the forward-thinking clients that are going to come together and they're going to be willing to do the hard foundational work that's going to give them the visibility they need into their internal workforce, their extended workforce. And that hard work is really finding that common taxonomy for occupations and skillsets and allowing them to be embedded within their foundational systems, their business tools and platforms, their VMS technologies, and that common language is going to be the bridge for us. And once they go through that journey and they do the hard work, that's going to give them the perspective they desperately need. So, "How does my internal compare to my extended workforce? The extended workforce that I have, how does that compare to the outside world?"

In the end, if we do this right and we get that visibility together, that workforce visibility, we're going to make business leaders have more confident financial decisions. And then ultimately, I think you're going to find that these forward-thinking customers and clients are going to tap into talent pools that others aren't going to be able to quickly identify. So, it's going to give them a strategic advantage as an organization on acquiring talent, making better decisions, and ultimately hitting all your goals that you have as an organization.

A key critical point would be, "What are my goals? What are my expectations of an MSP or a services program? And how do we align the work that we're doing to meet those goals?" There's a lot of targets out there, there's a lot of data out there; really honing the story and making sure that it's actionable and it's aligned in my expectations. And then ultimately, you're measuring the efforts, measuring the work that's getting done towards those goals. So that's the key and critical thing that we need to figure out, and those clients that do, they're going to have a pretty big strategic advantage in the workplace.

Bruce Morton: Awesome. Love it. Thanks Tim. So now for the Bruce bonus for Jason. So as I'm listening to this, and obviously part of it, but also listening in in a way, there's a lot here. There's a lot. So if I'm a listener in this space, buying into this, understanding the importance of data, where the heck do I start? If you're going to eat an elephant, eat it in bite-sized chunks, right? How can we make this very tactical for organizations to give them perhaps a proof point, something to walk in the boardroom with and say, "Hey guys, you need to pay attention to this."?

Jason Ezratty: Excellent question. It's a bit similar to my prior answer, which is you need an immediate result. I gave a similar answer just about people contemplating how to start with AI. I give very similar advice. Start with something that will have actual tangible results in a sub-six-month timeframe so you can walk into the board not with a theory, but with results. And then you have the theory that takes those results at scale, takes it further, maximizes the opportunity. But without that story of, "Here's exactly what happened and why we can believe in it," then in these times, surrounded by so much uncertainty, executives are less likely to pull a trigger on something that's so enterprise-wide impacting as what we're talking about. So leading with the results in hand I think is what's critical. So it's that scale thing. Whether it's five or 200 SOWs, that you can see that delta, show the difference, show the report, from Tim's point, of worker tracking. "Here's all the workers that we have been able to track. This is the small percentage that we know stuff about. This is the even smaller percentage that we feel good about in terms of what's known." Supplier relationships, prices, et cetera, all of those things.

And I guess one other way I would answer your question, Bruce, [and] I'd like just to take a crack at answering the crystal ball question as well, which is I think the way our partnership shows up perfectly is that we talk about it in a less technical fashion. The way that we all know what long distance calls were, when we stopped talking about fiber optics. We just expect it. Same thing with the spam filtering on our emails. Remember the disgusting mess that our email inboxes used to be, and now we have spam filtering because of AI and all these other technologies that make that work for us. So I see this similarly in staffing. We used to worry about the stigma of the temp being lesser than the employee. We don't talk about that anymore.

So to me, a lot of the best improvements where we can declare the win is because people aren't talking about it as much. They simply expect it. It's just part of the package. It's an expected table stakes, part of the value prop. And that's what I see. I think that we have worked so hard to make the maturity of this partnership come to the state that it's in that we can now start to turn the corner of customers just expecting these things.

Bruce Morton: Great. Perfect. Well, it's a wonderful way to wrap it up.

Jason Ezratty: I'd drop this mic, but I know it was too expensive, so I don't want to overdo it.

Bruce Morton: Yeah. So, thanks for that. I guess I'll wrap it up here. I think it was a really good way to end. I'm going to hold you to that. So, if an organization reaches out to us and says, "Hey, I've got a half a dozen SOWs," we're willing to show them what's possible, right?

Jason Ezratty: We'll even let them execute the savings on it, and we're very confident that they'll come back demanding more. Yeah.

Bruce Morton: Yeah. Great. So, we'll wrap it there. Thank you both so much. People know where to find us. I don't think I'm going to ask that question. People know who we are. So please, more interested, more questions, please feel free to reach out to Tim or Jason. Thank you very much for listening. Thanks guys.

Jason Ezratty: Thank you.

Tim Johnson: Thank you.

Bruce Morton: If you enjoyed this episode, please rate and review us on Apple Podcasts, Spotify, or wherever you get your podcasts. And if you have questions, send them to SubjectToTalent@AllegisGlobalSolutions.com. Follow us on LinkedIn with #SubjectToTalent and learn more about AGS at AllegisGlobalSolutions.com where you can subscribe to receive additional workforce insights. Until next time, cheers.