Bryce Hoover: Well, I originally got hired to do something completely different, actually, but by the time I joined the company, the guy who hired me was no longer with the company. But this was right around 2017 when the Model 3 was just being introduced, and vehicle logistics needed analytics to measure and manage the rapidly scaling vehicle delivery operations. So I slotted it with that team, and have been leading analytics functions ever since, and have rotated through many different departments in the company, from vehicle delivery, vehicle service, finance, and then HR.
Leading analytics functions can be quite challenging. Seemingly every business problem could benefit from data and analysis. What's your perspective on how a people analytics team should identify the scope of work to prioritize?
Bryce Hoover: Yeah, that's a great question. This is a perennial problem that centralized analytics teams often face. There are endless problems and priorities that need the attention of analysts with dozens of stakeholders asking for their time. So it can be really difficult to understand where you fit as a team and what's expected of you, and whether or not you're even being successful. So to make a real difference, analytics teams need clarity on their purpose, and they need enough runway to do deep work that produces step change improvements in an organization. Ad hoc, shallow, urgent, unfocused work is definitely necessary at times, but will definitely not make a lasting impact. And if this is the modus operandi of the team, it will definitely leave everyone, both stakeholders and analysts, frustrated. So one is no one is satisfied with the experience. This is why the org structure is really crucial to get right. And I want to highlight the great work that the team at insights, 222, have done to emphasize the importance of the org structure in their research.
So the people analytics capability cannot be buried somewhere in the HR org, or they'll just be out of touch and behind the curve on rapidly evolving priorities. They'll be on the periphery, rather than at the center of the action. And also, the team cannot sit as an independent entity, equally at the disposal of multiple departments throughout the company, which requires the analytics leader to be the arbitrator of who gets the team's attention. So to really be impactful, the people analytics function needs to report directly to the HR leader, with the analytics leader, acting essentially as a Chief of Staff for the HR leader when it comes to data. So in this operating model, the objective is first and foremost to do everything possible to ensure that the HR leader is successful in their objectives. Now someone might object and say, 'But wait, aren't we here to make the company broadly successful, and don't we care about doing what's best for the business as a whole?' Yes, of course we do, but that's too broad to be a useful guidepost day to day.
Making the company successful can mean more than making the HR leader successful, certainly, but it definitely doesn't mean anything less than that. So we first need to ensure that the HR internal affairs are in order before trying to help other teams. So if the HR function is not well equipped with robust analytical solutions, access to data, and reliably able to generate insights on its own questions, that needs to be the team's primary focus.
And I guarantee you, if this is not yet in place, your HR leader is going to feel very uneasy. They will want to know that their team has strong commands over the data and has the capability to access, report, analyze, and deliver results that make the HR team successful operationally and also polished on upward and outward reporting. So this is really the first step in your Maslow's hierarchy, stabilize access to the data you need, work with the HR leader to set up baseline reporting and operating tools, and build your underlying data model that enables you to be nimble with ad hoc requests and deep dive analysis within HR. This is your beachhead. Once you're firmly planted here, you'll be able to expand your circle of concern, but the process won't change. You're not going to go rogue. It's all going to be driven by the analytics leader maintaining a strong relationship with the HR leader, and aligning on what's most important together, and maintaining a high degree of focus in an aligned direction. Focus equals impact, and maintaining the alignment with the HR leader is what enables you to stay focused, and not delete your effort.
Bryce, you said so many things that I wish came out of my own mouth, because all of those things are so critical and important, and I want to just really help our listeners understand, what does it mean to make the HR unit successful before moving outwards to the business. What does that mean to you at Tesla?
Bryce Hoover (05:41): Sure, making the HR leader successful starts with a strong relationship with that HR leader and understanding their needs. So I can't say exactly what it means for everyone, but to start in that direction, strong connection with the HR leader and the analytics leader, that's where you start. And then you understand their problems and priorities. What's keeping them up at night? Where are they trying to take the business? Where are they trying to take the internal or the HR function? Understand their intent, and then harness the technical resources on the team to maximize the business facing value of the data assets that you control that's high level.
And do you see any pushback or angst from the business when you're so focused internally on the HR team? Sometimes that doesn't always translate to other business units as valuable.
Bryce Hoover: That's a great question. Often the priorities of your HR leader are going to be aligned with broader business objectives, but you want to be driven towards those objectives by the HR leader, so that you are being pulled into it through that leader, rather than feeling like your team is being stretched by this leader and that leader and this leader and that leader. That's the delusion of focus that is going to naively affect your team's ability to deliver impact, especially if you're a lean team. If you're a team of four analysts, six analysts, you need to be realistic about what sort of impacts you can have and where you can have it, and the timeline of that impact. So trying to deliver impact everywhere all at once is not a good strategy.
Establishing a beachhead that you can grow and expand out of is much more realistic, especially when you're first starting out, you might have a young, developing team. So creating success somewhere is extremely important, and that's somewhere, since you're the people analytics team, starting with your HR lead, making them successful, and then working out that platform will enable you to start working with your HR lead to see what other leaders in the business can benefit most from our efforts.
And so really leaning on your HR leader to help other business units understand the scope of work that you're doing and how that relates to them. Leaning on your leadership is basically what you're saying.
Bryce Hoover (08:06) Totally well, and to be clear, you will be engaging with leaders throughout the business even as you're maintaining the priorities of your HR leader. Focusing on your HR leader's priorities does not mean you won't interface with other leaders in the business. If you are creating a headcount report at the enterprise level, you can be sure that every leader in the company is vested in that report. They're going to want to understand the definitions. They're going to want to understand the cadence of that reporting, who's getting it, and where the accountability lies for those numbers.
So you're going to be interfacing and solving/ addressing concerns and problems with other leaders along the way to create alignment for these solutions that you're delivering on behalf of your HR leader, because those solutions are going to be at the enterprise level, and they're going to be likely executive facing because that's the upward and outward reporting that your HR leader is responsible for. So yes, to your point, you will absolutely be interfacing and engaging those leaders, but you will be doing it in a way that is focused on the priorities of your HR lead.
And when building the foundations for people analytics reporting and tools, what are the challenges that leaders should expect to face, and what are your thoughts on how to overcome them?
Bryce Hoover: Well, great question. I mean, just like we were talking about related to definitions, it's really important to align on definitions at the outset with your HR leader. That might sound intuitive, but it's shocking how misalignment on definitions can create problems down the road if they aren't clearly expressed and articulated and agreed on right at the outset. So how headcount is defined, for example, will have huge downstream effects for how leaders in the company are held accountable for growth or accrue targets and attrition, or how you calculate attrition rates over what time period. How many headcount data points to use to determine the average headcount for the period. What segment of the headcount population counts towards attrition? These are all extremely important to flesh out, and the HR leader needs to be engaged at that level of detail to understand and agree before any reporting is distributed. And all those definitions need to be clearly documented so the leaders can follow along with the math. Otherwise you're going to get bombarded with questions or leaders just won't trust the reports because they can't back into the numbers, because they don't understand the definitions.
So that needs to be front and center at the outset of especially any enterprise level reporting. And then the other thing you really need to pay attention to is trending data.
So one of the easiest ways to lose trust in your data, especially at the executive level, is to show trending data where the numbers you're reporting on shift retroactively. So data for the prior week in this week's report doesn't match data in last week's report. That sort of scenario. So taking the time to understand all the operational nuances that can lead to retroactive changes week over week can be really tedious, but it's time well spent. It's really important to understand and account for that in your logic so historical data remains unchanged. Along a similar line, as leaders changed throughout the organization, maintaining continuous headcount and attrition trends through leadership changes can also be very tedious. Who's getting- whose headcount is getting redistributed, to who, and where, and at what time? That logic will continue to build up over time and can be complex to maintain, so you need really good documentation in your code, and someone on your team with the historical organizational knowledge to explain the reasons for how headcount was allocated, where, historically, as you're looking at long term trends.
Another point would be creating foundational tables. So many companies have employee change detail tables where they only create a new employee record whenever there's a change to the employee's details, like a job change, or a location change, or an employment status. So if this is all you have, then trending reporting, aggregated by employee attributes will be really tough, if not impossible. What you need is a daily employee snapshot period. You just need a record of every employee every day. And don't let the data engineers talk you out of it, because it's computationally expensive. Putting in the effort to establish this table, as well as other tables, like standardized date tables, for example, to easily tag dates like start with month, end of week, beginning of quarter. Setting up your data model, this will significantly simplify your reporting and make it easy to maintain continuity across the team. So take time at the outset to think through what sort of data model will make your life easier and enable standardization across the team. And then finally, write access. So this is probably the most pivotal thing to get right.
So if you're an analytics leader at a company, and you don't have a database environment where you have right access, you need to make that your number one priority. You don't want to be in a position where you have to rely on data engineers to implement your business logic in their data mart or keep queries offline or embedded in Tableau workbooks.
Having write access to your own database environment will give you the ability to fully own all business logic, stored procedures, views and tables used for HR reporting. Now you're obviously going to have clear guidelines from managing sensitive data, but that extra effort is well worth it to ensure you can directly control the evolution of your data assets.
Felicia Shakiba: I love the point about write access. That's a minor thing to think about, but so important,
Bryce Hoover: Yes.
And so I want to zoom out and get to the real tough stuff, which is, how do you recommend analytics leaders partner with different parts of the business to integrate people data and really gain buy in.
Bryce Hoover (17:07): That's a great question. So I would say if you're just starting out for the first 6 to 12 months, just stay razor focused on two things. Build your underlying data model which we've just been discussing that will enable you to deliver business facing results quickly and accurately. And then build all the reporting the HR leader team needs to report up and out. Now this work will never go away, but once it reaches cruising altitude, you'll notice the conversation with your HR lead will begin to shift. Their answer to the question, 'What's keeping you up at night?' will change from questions like, 'How can we best communicate how headcount is evolving across the organization?' It'll change from that to higher level questions, like, 'How can we ensure that this company is the most rewarding place to work for the best people?' So you'll notice a shift from very targeted questions to like, higher level enterprise impacting questions, and that will shift as you and your leader feel confident that you have the foundations under you to be able to answer these questions. So once you start fielding these sorts of questions, the stakeholders you need to engage will expand dramatically.
You'll need to accurately wield a broad array of data points. So the first stakeholder groups to engage are the functional leads. So who leads your training department, who leads benefits, who leads leads, who leads comp, who leads HR claims? You're going to find these functional leads, and you need to expand your understanding of these functional domains and create full alignment with these leaders on how to measure their work.
That's step one. Then step two is to consolidate and organize and synthesize the data from these functions in a way that can tell a holistic story. So once you've done this, then you're ready for step three, which is to get the data into the hands of your business leaders and demonstrate the power of that data for decision making in their day-to-day business to improve the employee experience.
So step one is to get alignment with the HR leaders, and how is it that they would measure their work? The second is to consolidate and organize, and then the third is getting data into the hands of the business leaders to help make better decisions.
Bryce Hoover: Exactly, and not just getting the data, but doing the work up front to show how the data can be used to make the decisions. These solutions don't have to be perfect, but they should demonstrate to those leaders what is in the realm of possibility, and then they become thought partners on how to iterate on that.
Felicia Shakiba: Absolutely and I'm sure that iterative process is for everything.
Bryce Hoover: Exactly yes. But that's the first step to get in front of them with something that's meaningful and valuable that you've put thought into that is synthesizing the results of this data and makes decision making off of the data that you're providing intuitive, at least from your perspective. And then that's when you're ready to start the conversation with the leaders, to really hone it in to exactly what they want. And once they see where they're starting from, from what you've provided, they're going to be great thought partners on how to direct it from there.
How do you think about managing the employee experience with all of this data, and knowledge, and partnering?
Bryce Hoover (20:21): Great question. I'm a firm believer that great people want to be compelled by a mission that's worthy of their time, energy and emotional investment. They do not want to be coddled, but they also don't want to be constrained by obstacles that are not inherent to the mission they're trying to achieve. So delivering on your company's mission is going to be hard, and that's why we're all here. We're all employed because we're tackling hard problems that haven't been addressed yet, that are worthy of us addressing, but that's a hardship that talented and motivated people are inspired by.
So what they don't find acceptable, and rightfully so, are constraints that get in the way of this mission that are not inherent to the mission, and these are all the constraints that arise from just people organizing themselves and working together, and the frictions that arise from inefficiencies in your structure or cultural dynamics. So you want to surface these issues so that your teams can maximize the energy that they dedicate to the mission, and they're not slowed down by the frictions that related to the core innovation or operational solutions that will deliver on that mission. That's just very high level.
So let's double click into that. How do you think about getting the right data in front of the right people?
Bryce Hoover: This is a tough one, especially in HR, because you're dealing with such sensitive data, you have to ensure that the right people are seeing the right data and only the right data, and not extra data. So especially when we're reporting outside of HR, this is a top level concern. Scaling access to business leaders actually shouldn't be too difficult, because the org hierarchy is an attribute of each employee. So as long as you bring in data to the reporting layer at the record level before you aggregate it, you should be able to permission control the data within the reporting layer itself, but scaling access to sensitive data internally to HR stakeholders, ironically, can actually be much more difficult, because HRPS have their own structure, and each HRP is responsible for supporting multiple leaders, and their responsibilities often overlap with HRP to leader coverage. So that poses a challenge, and this relationship is rarely documented anywhere, for most companies, much less in a format that can be ingested and used to control permissions.
So how do you get sensitive data that needs to be permission control into the hands of your most influential HR partners, who can actually drive the cover, who could actually drive the conversation with the business leaders if they had access to the data. So how do you give them access? The quick answer is to hard code, but it's not scalable and it's hard to maintain. You'll just, you'll just be able to manage access to your top level HRP leads. So you'll give HRP lead for all engineering access to the engineering org, but you probably won't be able to manage that for all their direct reports. So the lead HRP will have access, but you probably won't be able to scale it down any more than that and maintain it. What you really want to do- so actually get that hard coded solution in place, but then what you really want to do is digitize the relationship between HRP's and leaders.
So there needs to be a mapping maintained in your HR workforce system defining each employee's HRP, not just some org level SharePoint, but each person's HRP, and every employee record needs to have the corresponding HRP. Once this is established and you've got that, really that data flowing into your database, you can use a combination of the HRP employee mapping and the org structure to ensure that the HRP's can see the data for leaders that they support and all the people below them in the employee work structure. So that's the general roadmap.
So the data is specific for each client group.
Bryce Hoover: Yes, and this has not only will this help you on the data side, but it will force the HR function to get much more organized around who supports what, and to see that mapped out clearly, will probably help them as an HRP unit rather efficiently.
So then, how did you work with leaders to understand and apply people analytics insights?
Bryce Hoover (24:52):
Fostering a data driven culture with HR is really important. So back to our point about getting data into the hands of those HRP's, that's essential, because the analytics leader can't maintain a relationship with each department head at an enterprise level organization.
So your HRP's are essential partners, because they need to be able to translate this data, they need to internalize it, they need to understand it, and then they need to be your partners in managing that the insights and the data that's generated, they need to be able to use that to drive penetration of that data into the business, so that the business is aware of the resources, but they know how to use them, and it becomes part of the decision making cycle. And the HRP's are excellent at that, because they understand their business units so well, and if you come alongside them as a leader to help them internalize the data and the tools and the capability of the tools, then they're the ones that are going to be best positioned to find the most useful application of those tools in the orgs that they support.
So you'll find that your HR partners are going to lean into this change big time, because it's going to give them a competitive advantage on being more useful to the leaders that they support. So really lean into that- those relationships with the HR partners, and make sure that they know how to be successful with the tools so they can elevate their contribution with the business units they support. That's a win for everybody.
Felicia Shakiba: I completely agree, and I think that having that relationship with HR partners are really important, and helping them collect the data and make sense of all the qualitative data that they collect throughout the year is very powerful.
Bryce Hoover: They're going to be able to make connections that nobody else can make, because they will see the data points, and they'll be able to connect those to the qual- all the qualitative conversations they've had over the past month, and they'll be able to help the leader piece together a much richer picture of what's going on, absolutely.
Bryce, what advice do you have for other companies looking to solve business problems through people analytics?
Bryce Hoover (27:13): First thing you think need to think about is leadership that you need. If you're a startup pre Series C, you probably just need a good analyst. But if you're past Series C through enterprise, you need a data leader to partner with the HR lead. This person is really key. If you're looking for a hybrid leader who sits at the intersection of business and technology leadership, they need to be a trusted thought partner, engaged and contributing to the strategic direction, to address problems and priorities, but they also need to be effective at solving complex data engineering problems, upstream product design, leading and coaching analysts in effective data modeling and business logic, and ensuring the quality and effective delivery of those findings.
So the person you're looking for is an end to end data leader that the HR lead can trust to ensure that the data is harnessed for maximum business value. And then the second thing I would really think about is operating independence. You need to be able to deliver results at the speed of relevance. The people analytics team needs to have full control over delivery of the HR analytics capability. So what does that mean? It means that once the raw data from applications has been made available in whatever reporting environment your company uses, the people analytics team needs to have full control over the development and subsequent tables, views, stored procedures, business logic, automation, processes, all of it.
So if the analytics team owns the business logic but doesn't have rewrite access to its own database, you're going to have a hard time being successful. There'll be too much friction in the process working with adjacent engineering teams to make that happen. Again, a good leader is essential here, because they'll ensure they hire the right analysts who are nimble, database management, data engineering and automation, in addition to the basic analytic skill set, so that the team has the skills that they need in fills to manage this broader scope that enables them to have the level of operating independence that you need.
Felicia Shakiba: Bryce, thanks so much for being here.
Bryce Hoover: My pleasure.
Felicia Shakiba: That's Bryce Hoover, Former Head of People Analytics at Tesla.
