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: Don Weinstein, Corporate Vice President, Global Product & Technology, ADP
Episode: 339
Air Date: September 13, 2019

Transcript

Important: Our transcripts at HRExaminer are AI-powered (and fairly accurate) but there are still instances where the robots get confused (or extremely confused) and make errors. Please expect some inaccuracies as you read through the text of this conversation and let us know if you find something wrong and we’ll get it fixed right away. Thank you for your understanding.

Full Transcript with timecode

Today’s Show is brought to you by the Human Resource Executive Magazine’s HR Technology Conference and Exposition, held October 1st through 4th at the Venetian in Las Vegas.

Join me and thousands of your colleagues at the world’s largest exhibition of HR Technology. Act now using the code HREX and you can receive a $300 dollar discount on your ticket.

Thanks. We’ll see you there. And by the way, don’t miss the Women In Technology segment.

John Sumser: Good morning, and welcome to HRExaminer’s Executive Conversations. I’m your host John Sumser, I’m sitting here with Don Weinstein and Don is the head of product at ATP but that doesn’t come close to describing the span of his work in the span of his career. So I would ask you to introduce yourself very quickly.

[00:00:34] Don Weinstein: Well, thank you John and thanks for having me. It’s always a pleasure to come back. So yeah, I have responsibilities at ADP for all product and technology and I think I believe it within the human capital space where the biggest out there and that also means we have a lot of coverage and a lot of span so, you know organization in technology at ADP it’s 9,000 people and we’re in a dozen different countries around the world. And [00:01:00] we you know support 810,000 clients about 41 ish million at current count of paid employees every single pay period.

[00:01:11] John Sumser: [00:01:11] I haven’t showed up here. I’m going to give imagine that when you were a little kid playing in the sandbox you went, hmmmm, products at ATP
[00:01:20] That’s it. So, how’d you get here?

[00:01:23] Don Weinstein: [00:01:23] That’s a good question. I started my career. I don’t know if you’ve ever had this conversation. I’ve worked on telecommunication satellites for the General Electric Company. So things like Dish Network, GPS, So now the connection should be obvious to you from there…to know — I’m just kidding.

[00:01:41] John Sumser: [00:01:41] I’m just totally obvious. I was interesting story.

[00:01:45] Don Weinstein: [00:01:45] I was at IBM. The previous stop and I was actually in the in the strategy group there at the time when IBM was moving into business process Outsourcing in particular. I [00:02:00] was working closely with the folks who are working on HR Outsourcing within within the Global Services arm of IBM, and we’re really looking at the business and we’re looking at the model and hit some very smart people doing some very good work
[00:02:13] But the thing that I kept struggling with was everything was everything is custom and it was the classic.

Lisa called your mess for Less lift and shift will take whatever you’re doing. We’ll use the same software that you’re operating on in many cases the same people because you would rebadge the people but we’re going to charge you we’re going to do it for less money than it’s costing you and make a profit
[00:02:37] It was a risky Endeavor.

But so and at the end of the day, the customer is the one who Bears the brunt of that risk, right? Because if I Outsource it to you and you don’t do a good job. You going to let 20% of the people go and maybe more because you need to make a profit on top. I’m taking a risk
[00:02:54] I don’t think people really appreciated that that risk. So I was in the business.

I was really interested to say [00:03:00] where can I find a better model for how to do this where you know, it’s obviously needs to work for the company, but it really has to work for the clients where I can deliver you consistent repeatable high-quality service
[00:03:14] That saves you money but takes you know, but it takes risk out of the equation. So I literally I was looking around us studying everybody in the industry and I became enamored with ATP from the outside looking in. So much so that I was sort of advocating within the IBM corporation that time that we need to make our Outsourcing model
[00:03:34] Look more like a tiki right? Let’s bring the best practices.

Let’s run it on our technology not the clients technology so much so that at one point one of the very senior Executives at IBM said to me I don’t hear about ATP anymore, you know come up with another model and then I got a call from a recruiter who was recruiting on behalf of ATP? I said watch that’s interesting. I mean I’ve spent a [00:04:00] lot of time being fascinated by this company from the outside. So let me go and take a look at it from the inside. So, you know, obviously I went through the interview process and really met and super smart people that I really admired him
[00:04:15] It was delighted when they gave me the offer and it was interesting having looked at it from the outside into turning to the inside some things that I didn’t really know or come to appreciate until I got inside the organization will talk and we can talk a little bit about the culture that you and I were just chatting about before about an organization
[00:04:34] That’s that’s fairly big and fairly successful. Very humble.

Like sometimes we hear a lot like you guys don’t talk enough about what you’re doing. And you’re right and we should do more of that. But you can see the culture of the organization is is very humble. The other one that I became instantly enamored with was at the time an underutilized asset was the data
[00:04:56] You know coming out of the engineering background. I’ve always just been [00:05:00] fascinated by analytical applications and there was so much data that we were sitting on and at the time I don’t think we were massively underutilizing it and I kind of made that a personal mission to say no we’re going to be known for something
[00:05:16] We’re going to be known for our data for the quality of our data. We have that it’s interesting about it is we have both quantity of data because of the breadth of our. Ridges, I mentioned 810,000 clients 41 million people that were serving right now.

If you look back we did over the last 10 years
[00:05:34] We’ve touched more than 90 million people but also the quality of the data because there is a lot of bad data floating around our industry and the one thing that we know if there’s anything that’s high quality data. It’s going to come out of the payroll system, right? Because if it’s wrong somebody will scream at you Earth, I think was underappreciated those how much data we actually get out of out of payroll Beyond just compensation
[00:05:56] For instance one of the best things we know is turnover and retention because I know when [00:06:00] you started and I know when you stop sometimes better than the nature are better than benefits or anything else that many many more things like that that we could reverse engineer, right?

We know who’s taking retirement benefits
[00:06:10] We know who’s taking health benefits. We know who’s taking other types and now we can start to cross-correlate. So that was probably the biggest aha moment for me was. The opportunity we had in data and the path of Journey. We’ve been on ever since to say how do we turn that around and put that into something that’s going to be useful for the

[00:06:29] John Sumser: [00:06:29] clients social learning directions to go
[00:06:31] I don’t know if you know that I was

[00:06:33] Don Weinstein: [00:06:33] an engineer. I didn’t lie. No. No,

[00:06:35] John Sumser: [00:06:35] you know, no, you know. No, I think I think that’s I think that’s where our conversations will go in the future. I want to start with 810 thousand clients. And so and so ATP is as close to a household name as yet in this industry and the challenge of presenting the complexity of ATP to [00:07:00] 810,000 individual clients
[00:07:03] Your marketing department just amazing. But but there’s so much clutter that conversation that must be a perennial

[00:07:11] Don Weinstein: [00:07:11] headache

[00:07:12] John Sumser: [00:07:12] detritus to try to sort out. How do you streamline that so that you get the message clear from

[00:07:20] Don Weinstein: [00:07:20] sender to receiver

[00:07:21] John Sumser: [00:07:21] without making it so simple that you lose the value that’s that’s available
[00:07:26] So that must be that must be several mind for you.

[00:07:31] Don Weinstein: [00:07:31] Yeah, absolutely. It’s a great question. I would like to describe it as at the good problem to have first-class problem. The good news here. I think are the interesting side as we actually just recently launched a brand refresh campaign. And so to your question, how do you cut through the complexity of that is you stop worrying about us and you can stop trying to make distinctions about the the client companies and you [00:08:00] go to the end of the chain and it’s the workers right
[00:08:03] Because you’re right small business is very very different from a large Global Enterprise. But a person who’s working in the marketing department of a small business and the person whose work in the marketing department of a large Enterprise they have more in common. Then you realize and you know at the core what do people want, you know, they want to clear they want to pay they want to be paid on time in the right amount

[00:08:26] You know, they want to have a good experience at work and they want to have simple technology that they can use and so if there’s if there’s one common thread running through not only everything that we’ve done really what we talked about yesterday during our our analysts a event was a t-piece focus on the workers and stepping out of the corporate office for a second

[00:08:46] Let’s go to the. Go to the place of where the work is happening who is doing it and focus on them. And so to that point the brand the new brand that we unveiled if you saw is like a cute little alliteration there the ATP they turn it to [00:09:00] always designing for people kind of a people and worker first message and we started running our first television

[00:09:08] That don’t remember ATP ever being on TV before so we ran a television advertising campaign and in the campaign, we featured the employees of our client companies. We had everybody from the Tabasco company maker Roscoe sauce big fan we had we had Air France was one of the clients that we had the cupcake shop and these are very very different businesses

[00:09:31] But what we showed was well here is a worker who’s making Tabasco sauce or. Flight attendant is getting ready for a trip and that became that common unifying thread that we could pull through the marketing and The Branding and here’s the good part is that The Branding actually lines up with the product strategy highly useful when that comes together

[00:09:53] So everything that we’re doing and everything that we’re working on with that with our latest generation of products is really focusing [00:10:00] on on workers and managers. It really started with started with our mobile application. We’ve talked about that. We wanted our mobile app 2009 and now it’s still one of the the most popular mobile applications in the entire world of business not makes yeah, that’s number one in HCM

[00:10:16] But in the entire universe of business applications is typically runs in the top five 1Mobile app used around the world by all the workers. Now the folks who are running the Enterprise’s we’re using and we’ve got different solutions is Carlos Rodriguez our CEO mentioned yesterday, you know, like we don’t sell the same technology to a
[00:10:36] Our two-person entrepreneurial company as we’re going to send to sell to a large Global Enterprise to run kind of the back office. But the face to the worker is all the same and that’s how we kind of Incorporated that into the brand campaign as well and its early stages, but I will tell you, you know, we got some really great coverage out of that, you know, you do a lot of paid media, but the one of the ways you can look at is also the earned media that came out of the [00:11:00] campaign some good coverage and the clients loved it our Associates loved it and made them all feel really proud

[00:11:06] Out of the work that they did, you know helping people in their work lives

[00:11:09] John Sumser: [00:11:09] server go down the next Road and think what’s interesting about what you just said is the way that it parallels what you’ve had to do to turn data from a great big pile of information into something useful. Yes. So it’s almost it’s almost the inverse problem where you going to start with the individual worker data and roll it up and I have a chance Dozen Years
[00:11:35] Go to consider the value and the extensive kerning adt’s data in from a bunch of cuttings on The Cutting Room floor and do something super valuable and it was a Herculean task should talk a little bit because that’s your work, right? Oh, that’s that’s that’s that’s that’s the trajectory that you’ve been on
[00:11:57] That’s how that. But

[00:11:58] Don Weinstein: [00:11:58] wasn’t easy that’s for [00:12:00] sure and we started even we started building out a data science team before we even knew what that was. Actually today. It’s numbers in the hundreds people because it is it is a Herculean task and you get into the universe now and you see there’s all these different sort of titles

[00:12:15] One of the popular ones. I love New Age titles. I’m just fascinated by them. So the one that you’ll hear now is data Wrangler who are now. Oh, yeah. She’s just that’s a that’s a good one. So there’s a lot of wrangling to do first because we had all the data. And we knew the data was right but a couple of things one
[00:12:32] Is it existed in different pockets in some cases that the underlying meeting wasn’t always perfectly clear. Like I said, I knew it was correct. But what did it actually mean and sometimes when you create data for one purpose and then you try and repurpose it if you don’t exactly understand not just the data point itself to the context in which it was gathered

[00:12:49] You could end up committing data data malpractice right to be quite candid about. Built out a huge team. I will say the technology has evolved a lot. You know, if you [00:13:00] look back at where we were 12 years ago to where we are today. So things like Technologies like Hadoop that enables us to pull all the data together in a giant cluster and start to run machine learning and self-discovery algorithms against it

[00:13:14] You know, that wasn’t there, you know when we first started. Don’t the thing that you’ll see that we’re doing is we’re always very cautious about not overstating what we know and that’s important to us because I think we want our brand to stand for if we say something you can count on it. We’re not just you know, pushing out a bunch of stuff and nonsense

[00:13:36] So we started out very simply with I got a lot of data. It’s in different formats. I can’t make sense out of it yet. But here’s something I do know. I know how many people we paid this week. And I know how many people we paid that week and I know that for sure and if I take those two numbers in context I can tell you something about the economy

[00:13:53] So we created the National Employment report and that was an enormous success for us and we’ve been doing it, you know every month for more than [00:14:00] I think we’re in our 13th year now of the the any are the National Employment report and so we just methodically moved down the path saying, okay. So what else do we know for sure we came out with this Workforce Vitality index because the other thing we do is I know how many hours people are working so we can

[00:14:15] Hit not just for people employed, but are ours trending up or down the one that’s been very popular lately is Wages right people getting paid more or less like there’s growth in the economy. We see the job growth, but is that good growth or back rub are these good jobs or not as much and now we’re we feel like solid and stable there and we’re starting to push past it into some really additional exciting use cases the K10 my personal favorite, so, Because and I think only ATP can do this because we have such a breadth of data that we can see people when they move from one ATP client to another ATP client and we were talking we’re talking about ethics and privacy around this before everything [00:15:00] I’m talking about

[00:15:01] It’s aggregated. Its anonymized. We have complex hashing algorithms in place. We have separate teams. So we’ve got the data wrangling team that’s responsible for getting the data. We have a different team data science team. We don’t cross paths on those that had to head to head to put that caveat out there

[00:15:17] But true. Here’s the analytic that just fascinates me so I can see if somebody goes from company a and Company B, but I can also see when somebody goes from Company B to company a. And I can look at the position in the. And now we know and we’ve shown in our data we publish this when people move jobs, they typically get an increase in compensation when they moved so now I can tell you if people are moving to your company and we would do this we wouldn’t do it at for privacy purposes

[00:15:48] We wouldn’t do a single Aid of the movement, but I could Define a basket of 20 companies and say well somebody from one of these 20 companies comes over. What is the premium I have to [00:16:00] pay to attract them? And and if one of those companies takes an employee from me, what is the premium that they are aiming to hire from me

[00:16:11] And if I can take those two ratio two numbers and put them into a ratio. I believe that is the best most quantifiable metrics of a company’s employment brand that exists. Well, there is not a good metric out there that exist today and again, I wouldn’t say that that’s a perfect metric but that’s

[00:16:27] Light years better than anything out there if I want to hire an employee from you. I have to pay 20% premium. If you want to hire an employee for me, you’re only paying attention percent premium that tells me something about your employment.

John Sumser: [00:16:40] That’s really interesting. Can you get granular enough to punch that by

Don Weinstein: [00:16:45] profession absolutely so by by location by that, so that’s where we were joking about it
[00:16:51] But you know, if you want to do big data, he got to have big data very got to be and and in some cases, you know ATP were big and some cases being big can be a

[00:17:00] disadvantage know you you not able to move, you know, kind of as nimbly as a small start-up but we’re at really is a tremendous Advantage is having that kind

[00:17:09] Scale of data that we can look at it by location by by profession or job. I think we’re at about 3200 different jobs. Now in our jobs taxonomy that we’re able to we’re continuing to look really because it’s funny as many as we have the clients like no, I want more granular. I want more granular. I want more granular

[00:17:27] So we’ll see how far we can get with that. But that’s the key to life is having the rich underlying

[00:17:33] John Sumser: data set. I can think of an armload of companies in the. Data analytics space who would like that as an input. Yeah. Are you licensing

[00:17:45] Don Weinstein: [00:17:45] out to other data manipulators? We’re having some conversations to see where that would go It’s always important to think about the use cases, but your instincts are spot on and what’s interesting if I think about a number of those those types of [00:18:00] organizations, you know, what distinguishes us again is the quality of the data when I when I know something I know it for sure, right
[00:18:08] Your location compensation organization. Those are just incontrovertible facts and those are not as easily rendered in some of the other organizations. I’m not going to name names here, but I would just posit that again. There’s a lot of data malpractice happening in our industry and there are a lot of folks who have data but they won’t tell you about the quality of it

[00:18:32] I know because sometimes they’ll come and try and sell their data to me as a customer and tell me what they think they know about my organization. When I had

[00:18:40] John Sumser: [00:18:40] a meeting where I just laughed out loud

[00:18:43] Don Weinstein: [00:18:43] to me, that’s great. Yeah, so good about ourselves after that. That’s

[00:18:49] John Sumser: [00:18:49] that’s that’s interesting. So that gets pretty complicated though
[00:18:53] The thing that the thing that are seeing about as you were talking is this old joke data science has two [00:19:00] components 80% of it is cleaning data and 20% of it is bitching about giving and it sounds it sounds like you. Found some ways to be effective even in the bread because because the data that comes in while it is type of factual
[00:19:20] It’s also dissonant because company a things about itself in these ways of Abby thinks about solving these ways and the overlap between the two circles that makes the Venn diagram. We have to work to make that

[00:19:34] Don Weinstein: [00:19:34] that’s right were.

[00:19:36] John Sumser: [00:19:36] To talk

[00:19:36] Don Weinstein: [00:19:36] about them, but for sure so job title and position has been probably the hardest me and then even within that lets say I can get the job title in the position right being able to look at, you know level of his
[00:19:49] Okay, you’re an accountant. Are you an accountant one or you an account to or you an account 3 because that’s how organizations. Just kind of think about it. And so that’s where the locus of energy is going right now again this [00:20:00] countable data that we have out there and then it gets interesting if I can start to pull some of these other factoids together
[00:20:06] And so that’s where we’re looking at our own internal information. We’re also looking at trying to mash up with some external sources publicly sources. Look there was a big case that that just got resolved earlier this week going to incur the LinkedIn kids and a lot of us were kind of sitting on the sidelines watching that one
[00:20:24] And you know, we’re on were on two sides of that equation because on the one

[00:20:28] John Sumser: [00:20:28] right

[00:20:30] Don Weinstein: [00:20:30] on the one who can I you know where I’m going with that but on the one hand, yeah, there’s some external sites that we would like to be able to tap into that. We think could reach our data same time there a lot of folks who would like to tap into us and so we need to be mindful of that
[00:20:41] But at least in the context of this conversation, I think that’s where the whole world is going first and foremost in terms of how can I just triangulate so I’ve got my data I do my best job. Of wrangling yet, and I’m going to start to Ping a bunch of additional sources and see if I can’t paint a [00:21:00] richer profile out of it getting people involved in the problem as well crowdsourcing that effectively I think is a good thing to do, but it’s also a false trap that you can fall into if that’s your primary source, right
[00:21:11] And again, that’s the difference of what we’re doing versus what some other folks are doing where if you’re exclusively crowdsourcing you have a bye. Stilton to your data set where as we have a core data set and we may use the crowdsourcing to test and look for outliers.

[00:21:26] John Sumser: [00:21:26] So this is going to end up falling in your lap as heart of Market education that you’re talking a lot about data quality
[00:21:35] And you’re talking a lot about these sort of layers of data quality, you know search so you get the perfect

[00:21:44] Don Weinstein: [00:21:44] clean stuff. Yes,

[00:21:46] John Sumser: [00:21:46] and then and then in order to make some prediction, yes need to supplement it with data of a lower quality and right. I don’t think there are very many people in Ur out of data science who actually [00:22:00] know how to think about and evaluate levels of data quality
[00:22:04] So think about that a little bit because because that’s an area in need of serious engineering.

[00:22:10] Don Weinstein: [00:22:10] It’s an area need of engineering and you also you also brought up education as well. So it’s all hit it hit it two layers of start with the engineering problems and data quality. Its you’re right. Most people don’t actually think about it
[00:22:21] It’s scary. You’ve heard me use this term malpractice a few times, but I see that running rampant through our you know, our domain in human Capital Management right now. We went from a period where nobody even thought or talked about data to a period now where people like I just stayed is important, but there’s just so much joy
[00:22:39] Being pushed out there. I seen it. I’ve talked to let’s be clear about this in my role. I get pitched by other everybody because they either want to partner with us or be bought by us. And so they share a lot of information. Obviously, we’re going to go down that pathway to get under the covers and look at what they’re doing 90% of it or more [00:23:00] is just garbage
[00:23:01] Okay, it’s just really really bad stuff and that that bothers me because on the one hand, you know, the old version of HR, maybe had a bad. Mutation because it wasn’t data-driven but I think would be even worse. You know bad data is actually worse than no data. So one of the things on the education fund we’ve been talking about we actually had you know, obviously this morning Marcus Buckingham gave a presentation for our group and he’s been leading the ADP Research Institute we’ve talked about do we need to do in education series about data and how to use it and put that out there
[00:23:35] Again, we could do it as like a public good and put it on the research institute site, you know, this isn’t this isn’t about a business model. This is about elevating the profession. So just ask a couple of basic questions with somebody shares with you some data are some analysis to say well, what does that mean
[00:23:49] What does it prove? You know, where did the data come from? How do you test it to know that it’s accurate or not? I don’t think we inspect enough. So and then if I could get at the second part of the problem, which is [00:24:00] okay. So then how do we engineer around? Some of these difficult issues a good performance is
[00:24:04] Favorite we talk about we talk about quality of hire for a second. That’s an interesting thing right and something that in my role. I literally the larger aren’t large organization. We hire over a thousand people a year. So I’m really interested in the quality of hire and I know we don’t have
[00:24:20] Great information about it. Especially if we’re looking for our performance rating system. That’s not a good place to look but are there other factors that I can look at? Like who’s getting promoted goes getting above Target like bonuses or raises go. Those are more tangible factors than a subjective rating that’s biased by the individual reiter, you know for all constrained by a certain pools of you know, bonus dollars available that starts to give me a little bit different insight about well, who are the
[00:24:49] After form is of an organization. One of the things I’m interested in is which of my managers and team leaders do the best job hiring because I want to be able to track people longitudinally through [00:25:00] their career and say I’ve got somebody here who’s a real star. Let me go back. We’ll who hired that person who do they work for now, but who hired them and can I start to find those folks who are the best Recruiters in the best hires in the in the organization and create data that don’t easily exist today,

[00:25:17] John Sumser: [00:25:17] so we should have
[00:25:19] Much longer conversation about how do you measure something? My theory about how you get it the question that you’re asking inches quality of hire. Here’s what we know about quality of 50% of recruiting decisions are understood to be Mistakes by month 18. There’s no other function in the organization that’s allowed to fail

[00:25:42] Don Weinstein: [00:25:42] that

[00:25:43] John Sumser: [00:25:43] Steely jokes
[00:25:44] So what do you know? You know that on day one. It’s a honeymoon. That’s right. If I Was Eighteen fifty percent of the time it’s a divorce. That’s so what you want to know first is is the hiring manager still happy with the decision. So that’s just a thumbs up thumbs [00:26:00] down every 30 days out of time and as you start to accumulate that data, You’ll understand by hiring manager where the fall off is now, let’s understand where the fall-off is
[00:26:11] Then you can ask the next set of questions so that instead of having to come up with a measure of quality or some assertion about quality. You actually track the evolution of satisfaction with the process and at the places where it becomes a defect.

[00:26:29] Don Weinstein: [00:26:29] Yeah,

[00:26:29] John Sumser: [00:26:29] that’s where you go to investigate next right
[00:26:32] So, I think that’s actually a pretty easy thing to

[00:26:35] Don Weinstein: [00:26:35] set up Folly you you intuitively went to a place where we like to go as well, which is. If I’m trying to assess like the managers you and the quality of hire I shouldn’t be asking. Well is John a good Tire? I should be asking and I happy with my decision exactly what I hire John again if I could write and what we’re doing there in Marcos talks about this as well a lot is I might not be a good assessor of [00:27:00] you, but I’m a good assessor of my own intent
[00:27:02] Right and I can assess fairly if I had to do it over again. What? And that isn’t that is a Fairly reliable answerable question much more so than will. It was John’s performance rating of 3.2 or a two point nine, which is almost useless.

[00:27:19] John Sumser: [00:27:19] It’s completely useless. It’s a measure of politics and exactly so we’re going to exhaust the time

[00:27:26] Don Weinstein: [00:27:26] I’d like to do some more. I would like

[00:27:29] John Sumser: [00:27:29] that is where the service let’s get scheduled for sometime after the big HR Tech conference and have a deeper conversation and let me just say thanks for doing this is a great conversation I really appreciate it.

[00:27:41] Don Weinstein: [00:27:41] Thank you for having me always a pleasure.

John Sumser: All right, great.