HRx Radio – Executive Conversations: On Friday mornings, John Sumser interviews key executives from around the industry. The conversation covers what makes the executive tick and what makes their company great.

HRx Radio – Executive Conversations

Guest: Stacy Chapman, CEO and co-founder of SwoopTalent
Episode: 363
Air Date: May 1, 2020

Transcript

Important: Our transcripts at HRExaminer are AI-powered (and fairly accurate) but there are still instances where the robots get confused and make errors. Please expect some inaccuracies as you read through the text of this conversation. Thank you for your understanding.

Full Transcript with timecode
 
John Sumser 0:13
Good morning and welcome to HR Examiner’s Executive Conversations. I’m your host John Sumser. Today we’re going to be talking with Stacy Chapman who is the CEO of a company called SwoopTalent. Stacy, how are you?

Stacy Chapman 0:26
I am very well, thank you, John, considering all the things that are going on in the world. And you?

John Sumser 0:32
Well, I like the fact that you are trying to set some sort of international model for working remotely.

Stacy Chapman 0:40
I’m here for you!

John Sumser 0:41
Yeah, there you go. Somebody’s got to be the outlier. Would you take a moment and introduce yourself?

Stacy Chapman 0:47
Sure. So I am Stacy. I’ve been with SwoopTalent for quite a few years. I’ve been in the HR Tech industry for a few decades. So I go back from doing implementations in insurance companies in Australia of payroll systems to worked at PeopleSoft for a long time did some consulting, worked as a practitioner. And this is my second company I’ve started. Always been a little bit obsessed by data and systems which is where my heart probably lies.

John Sumser 1:16
So when you were a young girl, I can’t imagine that you thought this is what you would do with your life, how did you arrive at, I’m going to spend time in and around HR with an increasing focus on data. That’s a pretty dirty conclusion.

Stacy Chapman 1:30
Oh, completely by accident. I mean, how many careers are really planned? But no, I just I just the jobs and next, the logical next step at the time in a very short term perspective, what seemed to be the logical next step. I’ve all grown there. So I guess you know, you eventually do what you’re good at perhaps.

John Sumser 1:48
Well, you know, I don’t know how to communicate this effectively to the audience, but you are good at what you do. So tell us a tiny bit about the company.

Stacy Chapman 1:56
Sure. Look, we are a fairly small company with very big clients, we focus on data and system connections and system relationships with each other. So we our customers with our platform end up with a giant, what we call a talent data lake. So the company has a whole bunch of systems that pool data and clean that data and connect that data and make it into this great giant useful data set that also is connected live to all of the various systems that our customers have. So they move data around between those systems. They report on that data, they analyze how various things across the organization and across a tech stack are working. So we bring together in a hub and spoke model connecting all of the systems and all of the data that people may need for their various HR Tech and talent needs.

John Sumser 2:44
So I lost you at data lake and I was thinking about swimming out to the platform in the middle of the lake and getting a suntan. Give me an example of something that a client uses you for.

Stacy Chapman 2:55
Sure. So, one of our clients over the time of our relationship had about 14 different major pieces of recruiting tech. And we’ve integrated to one of their HR Tech. Those systems have got all kinds of different data in them, right? Your ATS has got candidate data, it might have custom data, it might have all kinds of things. You have a chatbot that’s got all these conversations really unstructured, but interesting data, you have got a CRM or a recruitment marketing platform. And all those different ones are together plus a workforce management system over here, and maybe your core HR system over there. We connect all of those data sets for them. So there are paths for data in and out of all of those systems in whatever way they want. And in the middle of that all of that data is brought together. So a big combination of structured, unstructured, private, public, or normalized, raw, all the different ways sources and structures that data can be are all sitting together in the middle in this sort of living, growing set of data. Now that’s a bit different to a data warehouse in that there is no any data can be in a data lake, Blog posts can be there, conversations can be there. So the customer will have all of that. And the use cases on that could be from the very simple like, make sure all of my ATS records get pushed to my recruitment marketing system in the right way. Or they could be more complicated in terms of, try to do natural language processing to assess what the emerging themes in conversations are. And the use cases are all about the customer and what the customer needs to do and what they’re trying to achieve. What we do is make sure all the systems connect and all the data works really well in the middle to as flexibly as possible support whatever they want to do. And John, some of them do really amazing things like predict the future capability of their workforce, and some of them do really necessary and slightly less sexy things like make sure that their chatbot data is not lost, you know? So, it varies hugely what they do we just make sure that they get to do what they need to with data and with automation across systems.

John Sumser 5:06
So, just before we got started, you were talking about a big question that you’re thinking about which I would spit back as, how and why people change and how to make the most of that. But why don’t you elaborate? What’s the big question that you think about?

Stacy Chapman 5:18
Well, at the moment with what’s going on with the economy and the workforce and the population of the world, customers are needing to make really serious direction changes and gear changes. So for example, we have one customer who has decided to do field jobs only from internal mobility. Historically, internal mobility was a real political challenge for them because they weren’t allowed to hunt from other managers aren’t allowed to hunt from other managers workforces, and for some organizations, that’s a very real obstacle to internal mobility. But right now, because of the way the economy is, they can only recruit internally. Now from our perspective, that’s just a look at a different part of the data lake but for many companies, that’s a really huge transition to have to make, you know. So it’s been nice for us to be able to say to customers Oh, you know, you just have you put this flag in and you’ll be all of your stuff will be connected and we’ll limited to your internal workforce, it won’t, you won’t have to worry about sorting that out or going to look for more data or worrying that your data is stale or any of those things because we have all the data together, away, you got it. And so it’s been a really nice example for us of letting customers be able to quickly implement changes in strategies. And we’re also seeing some quickly implement changes in system. Like I’m sure you’ll be noticing, there’ll be a lot more video conversation. There’s a lot more remote type technologies, people either trying or using much more than they ever did. And we can make those changes, collect that data and make sure that things that you have are integrated because customers have built those connections over time. It’s not like a massive effort for us. It’s just right. Yeah, just turn that on. And here’s the API. Or, yeah, that works great. And that’s been that’s been really good for us for our customers in In terms of their ability to make those policy and strategy changes very quickly, and it’s been quite rewarding in that it just lets you know how much having the systems brought together in the data brought together does give you an ability to make changes in the direction you want to go the tools you want to use the strategies you want to put in place. So it’s been stressful. It’s been really nice that they’re able to do that with just the pulling of levers.

John Sumser 7:25
So you’ve got this great big pile of data in this data lake. And it was it was an incredibly aspirational goal. And then all of a sudden, this virus thing happened. And you’ve got this what looks like an anomaly, but in some cases means that the business has to entirely switch its business model. And so a question that I keep kind of shape up is you’ve got that pool of historical data, and it may or may not be relevant to going forward and you’ve got this accumulating set of present data that may or may not be relevant your way forward, and you are trying to make data driven decisions. How the hell do you make sense out of that?

Stacy Chapman 8:08
It’s a very real challenge and trends and aberrations over time are always a challenge, they become less of a challenge, the longer the period of data you collect is, of course, that’s a mild, that’s a bit of a problem with privacy regulations. So anonymization comes in play. So yes, that’s a really big deal. But I think people are also connected also connect these two external data sets. So what happens is if you start to look at whether it’s an employment rate, your price, there are all kinds of external data points, and they can be industry tied and they can be relative, they can be fairly granular, too. And I think you’ll find that a lot of people, especially in the people analytics world have tied their internal things to external data points. And so sometimes that helps you determine leading indicators of change and things that are relatively that may be our external forces that drive Your workforce. So that’s one thing that when you have relatively extreme swings, it can be easier to identify those sorts of external points. But otherwise, I don’t think there’s a formula to answer it. I think this is going to be ironically, where human experts and thinking about context become more able to help people understand how to drive that or when not to pay attention to those things. So ironically, I think this is a point where human interpretation and the understanding of contexts and external factors becomes really important. Now, you know, I always have believed that you need to have the human context to understand why things happen. I don’t believe the argument that a roomful of monkeys will produce Shakespeare was ever true. And so I think that you what you have is an opening here for people with human expertise to be able to guide the use of data in those decisions.

John Sumser 9:53
So I’ve been thinking a lot about the hospitality industry recently. You know, all of the great hotels are closed and their workforces are laid off. And when they reopen, they’re not going to be opened as the same hotel they’re going to reopen, you know, the business of the hotel business up till now was to provide a certain level of experience as a primary object going forward, it’s going to be to provide safety and then the level of experience right and so that means the buildings have to be reappraised the the replacement staff has to be trained entirely differently. All the cost models are wrong, all of the efficiency models were wrong. And so the question of do you hold on to the historic data? And is it useful? This is one that I think a lot of people are going to have. What do you think?

Stacy Chapman 10:42
I think it’s going to be less useful in some areas than others. Certainly there will be major change and and there have been major changes in society before and in the workforce before and I think being able to segment to have a look thoughtfully at what those areas that are most changed. We’ll be because some of them the history will, will be completely different. I think remote work is another one. And the structure and design of work is another one that’s going to work really differently. And the way that you interpret or use data might be very different. Or you might literally decide that for parts of the workforce, you will throw it away. Equally though, if you think about you and I had jobs and all that hugely changed. So there are parts of the workforce which continue regardless. However, the other force at play, aside from yes, history might be changing purely by the expectation and also the level of fear that human beings have for engaging in some of the activities they used to engage you will take a while but also this in its by its own style of event, a pandemic will trigger and push innovation in other ways too. And I think automation is going to be the big one there. So the hospitality industry is going to have pressures around the provision of safety and hygiene and cleanliness, and they’re going to be big deals for a while but there is also The only way for an organization to really avoid costly impacts of human pandemic is to have fewer humans, unfortunately. And so I think that there will be a lot of investment in automation where it is possible. So we will see one of the medium term results of this being the escalation of automation in places where it can be. And I think that will come to play in hospitality, it will come to play in lots of areas of the workforce. So I think that there are the obvious ones, the short term ones where there are big sets of data that veel wildly from what the history has been. Now, don’t forget, many people do have data from sort of 2000 and 12th, which is a recovery as well. So there are data in various phases of time people have, but also I think the really smart people will be talking to executives and talking about whether we will make decisions to make other sets of data also less relevant by changing the way that we do things. It’s going to be a big deal.

John Sumser 12:54
Yeah, I think we’re in the process of learning a whole lot. So let’s briefly talk about the technology at SwoopTalent. Do you think of it as AI? Are you still using that language? What’s the core technology?

Stacy Chapman 13:09
There’s certainly quite a bit of AI in there for sure but no we’re not using that language. I just I think that people what will happen with AI and also with machine learning is it will become like people never worried whether you were using Java or some other operating system to build functionality, right. And I think to a certain degree that will happen here as well. It doesn’t really matter that we’re using natural language processing to find things out of unstructured text, it matters that we’re finding things out of unstructured text and the way that we’re doing that. So I think they increasingly become just tools, you don’t build them at night. Now I see when you start using them to do create models or build historical projections, probably organizations need to pay much more attention to what the underlying ways of doing that are. Because right at the moment, to your point earlier, if you’ve got machine learning models that are built on historical Data only, and they are trying to predict things for your next six months, they’re very likely to be wrong. So you do need to understand that you can make changes and tweak them. But we try not to talk about like what the AI is only the functional and design decisions that we put into things in order for people to have how much control people have at any outcome. Does that make sense?

John Sumser 14:21
Yeah, that makes all the sense in the world. But something that I’ve been thinking hard about is all of the technology in an operation like yours is about assessing and appraising people or things or data flows. And all of those things are subject to an extraordinary amount of change. And so right now, it’s a moment where the very meaning of the data is in flux. And so I wonder if you think you have some sort of insight about how big this change is going to be on operations over say the next year.

Stacy Chapman 14:58
Let’s talk first to that about what we do. When we say the meaning of data, and I think that is a point that is really important here, I think that this situation like assessing things is definitely a part of what we do. But trying to understand what the meaning and the connectedness of data points are is a really important part of what we do. That is the part of what we do that I think becomes increasingly important. Like how are you able to have data structures that let you connect a fairly messy set of workforce data to a fairly messy set of external data? So you can try to see yourself in context, that is something that becomes increasingly important, you know, knowing that that is an apple and that is an apple become more important when you’re trying to get a sense of your situation and your short term future situation in the context of other things, therefore other datasets so I think that’s increasingly important, but I think that the assessing things I let me give you an example, you know, some of our of our functionality is about matching talent to job and we use machine learning for that and do a whole range of do that now that is something which is too early a conversation going to get really strange what people are using it for is not to try to find the perfect talent in the talent pool as they were six months ago. What they’re using it for now is to help to automate the winnowing of a surge of application. So the technology is still there, but its application in a work context is really different. And those are the things that are emerging where to answer your question, I have to say, eh, don’t know yet.

John Sumser 16:35
Thank you for that. No, no, no, I think you’re you’re in a large club, although not everybody is as quick to admit it in public.

Stacy Chapman 16:47
There’s some interesting observations that we’re making right now like that one, the use of matching technology to automate the winnowing of candidates. You know, what that means, in a bigger broader sense? I don’t know. I feel like it might mean that relate to more automation, which leads to fewer humans, which makes me sad, but sort of is also where I live.

John Sumser 17:06
Well, one of the things that I’m seeing, it’s not just winnowing of candidates, but I’m seeing succession planning take on a different meaning. And so imagine that you just laid off a third of your workforce. And what you know about the remaining workforces that over the next year 15 to 20% of them are going to get sick enough to require time in the ICU, and three to 5% of them are going to die. And so when you think about your organization, and you think about that risk for each job, it means that you need to have succession planning and some access to who’s the most likely replacement down at the down at the traction level of the organization at the strategic level. So you want to know is there somebody in payroll in the payroll department who has a kind of knowledge that nobody else has, and how do you identify it? And how do you train somebody to take that over if they get sick. And so I imagine that that matching technology will be applied internally for internal mobility reasons, but internal mobility reasons that were designed to produce business continuity rather than promotion tracks for people. What do you think?

Stacy Chapman 18:18
I would agree and there are two, I agree with you. And there are also two external things that have come that change the way people perceive that. One is, Boris Johnson being in intensive care, which I think many senior people in positions of power, and with money and wealth, up until that happened probably felt they were inured. And so I think you’re probably looking at senior parts of the organization paying more attention to more traditional succession planning in a different way. But the other big social thing that’s been really interesting here is the definition of what is essential. You know, as we look at what happens here, and you start to think payroll, as you say, but also grocery and things which as a society and as organizations we maybe didn’t consider to be critical and pivotal roles in the past look very different now don’t they?

They sure do. They sure do. So, what are the big ethical issues that you run across in your work?

I think that over automating the selection is always a big ethical issue. And we run a run across that quite a lot like what is quote unquote, the best candidate for a job, whether that’s internal mobility, I think there are going to, and also there are huge ethical issues about data privacy, and its interpretation in use. And I think as people, we probably, you know, we need to work to help everyone be more educated about how that happens, and what data you are exposed to, you are exposing to the world. But I think that is a bit of a minefield, like if you think about going back 10 years in time to try and do anything that looks like or how does our organization look after an upswing? What variables will be different? How will we run this differently? And you don’t have data from 10 years ago because privacy says you must not. Then that is a challenge from a business perspective. But equally, it’s right to keep data private as it is required. So there’s an ethical issue around, for me, the de-identification of data and the protection of employee privacy that I think people haven’t got their heads around yet.

John Sumser 20:17
So there are other people out there who claim to do similar things to what you do what makes you different.

Stacy Chapman 20:24
Look, I think for us the sheer flexibility of what we can do look, we have, and this is “nerdy as,” but our database self-adapts to data, so and our API’s is self adapt to data. So what we give an organization is the ability to have an API that lets them push and pull any data from any system that they use anywhere, and I’m pretty sure nobody else has that. So we don’t if you have customizations, we don’t care. If you have systems you built yourself, we don’t care. If you have got systems that are so old, you can hardly even remember who you bought them from we don’t care. And I think that gives us a lot of a lot of a leg up in, especially in larger organizations who have got systems stacks that can be quite insane because of how much compliance they have to deal with. So I think the outrageous level of flexibility that we offer is something that just other people don’t have. And I think also people are still crazily thinking about point to point integrations as a way forward. And I think where we might be the only one that’s doing a very serious hub and spoke version of integrating everything in talent.

John Sumser 21:30
So what you mean by that is rather than focusing on having an API relationship for every system to every other system, so it looks like a spiderweb, when you lay them all out, you’re talking about having a central place where all of the data is and if you want to get that data into your system, you pull it out of there, is that?

Stacy Chapman 21:55
That’s correct. So you push it in there, you pull it out of there at will.

John Sumser 21:59
Okay? That’s a different view. And I take it, you think that that’s the way of the future.

Stacy Chapman 22:04
Look, I do. But ultimately it only works with a data lake in place, right? Because the other thing about integrations is typically data there just gets hurled. And it just goes from one to the other. And nothing in the middle is ever kept. And so there’s no place where there’s a version of all of it. There’s no place where there’s sense to be made relatively easily of all of it. So a hub and spoke integration model only works because there’s a data lake underneath it that keeps the data. So whatever data you have received from a system, there’s a copy of in your data lake regardless of whether you send it to B, C, D, or F system. So that data that is how the hub and spoke model works. If you think about most integration systems like Zapier is a quite famous one. Very, very constrained, because it can only offer point to point where once you have the hub in place, it stops being constrained. You can have processes that run across multiple systems. You can know that you can analyze what’s happening all the time because there’s the central store of all the data.

John Sumser 23:06
So do you have an analytics platform that allows you to look at everything that’s in the lake and do things that you can’t do in any of the attached systems? I bet you do.

Stacy Chapman 23:16
We do, well we use a visualization system, we didn’t build one ourselves, you know, we’re data people. So, we have a visualization system, where you can look at all the data that passes in and out of all of your systems, you can see how much movement is happening how it varies across the different systems across points in time across geographies or subsets you can slice and dice anyway you know, I you and I have talked about before I have this sort of idea that your HR Tech stack is like an organism and the data that moves around it is like the endocrine or circulatory system or whatever. And so what we want to have is a visualization of how everything moves around in those systems. Whether it’s by business process, by geography, by system, you should be able to see what’s happening with your data and your products at any point in time.

John Sumser 24:04
This also implies that you could set up a sort of a people analytics satellite hovering over the data lake without ever having to connect to the rest of the systems and just pull the data from a single source sort of at will. That takes a lot of the complexity out of people analytics, and gives you the opportunity to do some interesting things instead of wasting your time looking stuff up.

Stacy Chapman 24:26
Yeah, well, we have, we certainly have customers who do that a lot. Yeah, it’s a good one to do people analytics on. And we also do some normalization in there, which I think is really helpful. So not only can you look across all the systems, we can also make sure that all the data points have got, you know, the same job title and the same company name and the same school and those data points which historically have been really messy. We, that’s one of the things we use machine learning to do, to normalize that sort of data and get to understand what is the same so there’s a lot of the basic people analytics things and that also applies to pushing to a machine learning endpoint. Like if you’re trying to do machine learning, it’s a single point where all of the data is curated and made ready for your analytics, via machine learning.

John Sumser 25:10
That’s awesome. That’s awesome. So we’re running to the end of our time, what’d we forget to talk about?

Stacy Chapman 25:16
Oh, gosh, I feel like what rolls forward into the future for HR and HR Tech and what is going to emerge and do extremely well. And I think that’s the other thing that is the data point what’s where I just don’t know yet. You know, we were I remain really optimistic that having a really solid infrastructure to be able to connect and manage and assess things remains to be something that’s really important, but I don’t know it’s too early for us to see yet what other what the impact on other systems and other parts of this tech industry will actually be.

John Sumser 25:49
Yeah, it’s gonna be an interesting time. Thanks for taking the time to talk and you know what, I know you’re in Australia, but it never sounded like you were in Australia. It sounds like you’re in the room next store.

Stacy Chapman 26:02
The technology works John. Can you believe it?

John Sumser 26:05
Yes, I can, I depend on it. So thanks. We’ve been talking with Stacy Chapman who is the CEO of SwoopTalent, a Bay Area company that allows you to effectively manage all the data from your HR Tech stack and do amazing magic tricks with it. Thanks so much, Stacy. Would you tell people how to get ahold of you? I should, I should have asked that sooner.

Stacy Chapman 26:27
Sure, you can find us at SwoopTalent.com is the easiest way to find us.

John Sumser 26:33
Okay, so thanks, Stacy. And thanks, everybody, for tuning in. We’ll see you again. Same time next week. Bye Bye now.