An Interview with David Hanrahan, VP of People Operations at Zendesk: Coupling Performance & Engagement
David Hanrahan is one of Silicon Valley’s HR innovators as VP of People Operations at Zendesk. After spending several years as one of Twitter’s Directors of HR, building the HR and Recruiting function at Change.org, he is now leading the vision and strategy around people at Zendesk. He has authored several works around strategic HR, including HR Tools Need to Flex Around the Culture, Manager Playbook, and several LinkedIn posts.
What are your current challenges with high performance today?
I think some big challenges that are on the performance side or identifying performance is this sort of traditional performance management model, you know doing an annual review... you get a rating from it, you put it in a tool and the manager tells someone how they did. It’s very top down and it doesn’t really understand or seek to understand what the employee’s motivations are.
I can’t tell you confidently that the traditional performance review system is actually going to identify high performance because of the inherent things like manager bias, nor is it going to help elicit higher levels of performance if you go through it. It’s a lot of energy spent on something where you don’t have any confidence that it actually ID’s high performance, or it fosters high performance.
I can’t tell you confidently that the traditional performance review system is actually going to identify high performance because of the inherent things like manager bias, nor is it going to help elicit higher levels of performance if you go through it.
One of the big challenges is that it doesn’t understand why the employee is motivated. There really isn’t an interest in that aspect. Engagement and performance have traditionally been de-coupled. You have a long list of tools out there that do engagement stuff, and then you have a long list that does performance stuff.
Is your current strategy then, to couple performance and engagement?
It is. So, we’re trying to couple engagement and performance at a macro-level first just by thinking of those two on the same timeline, instead of saying, “There’s one project team that’s just going to focus on our engagement approach, and they’re off over in this corner of the building, and they have their own timeline. Then there’s another team that does the performance thing, and they’re managing their timeline, and they’re just not talking to each other.”
Yes, so we actually have the performance and engagement model together in terms of the quarterly cadence. We seek to understand that information about employees basically in the same place, and so what we’re first instructing managers and leaders to think about is to understand what motivates their employees and how well their employees’ motivators are getting met - that’s one of the first key pieces of information.
Then we ask, “Why do we do that?” There’s pretty clear research that shows engagement and performance are correlated. So, higher engagement generally leads to higher performance; lower engagement generally leads to lower performance. And so we say, “Ah, I see, OK.” Then we go into a performance quarter where we have basically in the same place, the managers and leaders can see the latest trends on engagement for their team. Then they’ll understand from that same sort of mechanism, what the performance trends are.
I’m going to take a step back for a second. How do you define, or what are the factors, of engagement?
Great. Ya, so we use three questions in general that give us our overall engagement score at the company level, and those are:
Those three questions lead to an aggregated engagement score, and from that we can do a statistical analysis where there is a statistically significant variance on a team-by-team basis so we can show, typically, “This team right here is actively disengaged, versus the company aggregate, or average.” And so the way we do engagement surveys is we first ask, “What’s important to you?”
A very traditional model of engagement is you come up with your fifty questions...one of those is the eNPS question, there are some Gallup questions thrown in there, but the traditional model is that we’re just going to ask everyone the same question. What that kind of fails to recognize is that [everyone has a different take on what’s important to them]. So then, it’s more interesting to understand first is what’s important.
We then tell people…”Here’s twelve to fifteen factors of engagement,” like teamwork, management, work-life balance, compensation...there’s a couple of really big ones. We say, “You pick your top five. What are, in order, the five things that are most important to you in terms of motivation?” Then we ask them branching questions off of those to understand how well those motivators are getting met. So, the employee should be left with a picture of, “This is my engagement.”
In the future, what we plan to do, is allow them to discuss this openly with their manager. A traditional engagement approach is like, you never really talk about engagement at the individual level, you talk about it at the aggregate level. However what’s most important and interesting from my point of view, is at the individual level, how well are someone’s motivators getting met?
So, every time I’ve done this with a direct report, and it’s pretty easy to do - I’m always left with really fascinating insights, which is like, “Oh, I didn’t know that was so important to you. I didn’t know that before. I can sort of see why you’re frustrated, or why maybe things have been tapering off for you more recently.”
Are you suggesting for managers to take that one-on-one time to review what motivates that employee? Are managers supposed to remember what motivates each employee? How does a manager move forward when they have several teammates that all have something that’s motivating them differently?
The answer is basically yes, however I think it’s probably easier than how it sounds. What I’m suggesting is a couple times a year check-in and see how well the motivators are getting met. See what’s changed.
I did this with one of my direct reports a month late, and it was relatively easy. She’s like, “Hey, here’s some changes...this thing’s becoming a bigger, more important thing for me.” Each person will be different. Without a tool, without it operationalized, you would just be putting these in a Google drive some place, and maybe hyperlink to it in a one-on-one doc where you're keeping a running list of notes, but you would know that, OK for this one person, they’ve listed out seven things and clearly one thing in there that is the biggest issue. It’s 40% of what motivates her. It’s learning.
She said, “You know David, I’m busy, I have stuff to do, I’m just not learning.” Then you ask branching questions off that, “What would move [this factor to satisfied]? What would move this to [very satisfied]?”
“Well, I need to be part of some of these other projects you mentioned on other calls. I want to be closer to those, sort of strategic HR for a project, I want to step outside of my comfort zone.”
The goal is to solve it jointly. The goal is not for me to say, “Oh gosh, here’s some homework I have to do for some things I can’t even control for.” You’re trying to zoom in on what you can kind of control versus not control...what’s most at risk? In general, those are things that you can make progress on.
Now, if what someone told me was, “David, I want to be creative. I want to paint, and my job doesn’t allow me to paint.” And I would say, “Ya, it doesn’t. Gosh...OK. So we have a big impasse here. You’re trying to do something, fundamentally different than what I need you to be doing right now.” We’re having a different conversation, which is like, how do we get you to go do that thing that’s outside of the company that we clearly don’t have? That’s a different conversation than someone who is like,
“Work-life balance for me is a growing and important thing. I have a newborn at home. I can’t be up at night until midnight on instant messenger. I used to do that when I have time for this but now I can’t. I need to set some boundaries, and that’s my biggest thing.”
“Ya, that’s totally doable, we can fix that.”
And so, the first goal is let them speak. Let them sort of get it out in the open, their hopes, dreams, and fears, like what’s worrying them right now. This is going to be quality information, much better than you’d ever learn from like an exit interview. You never want this to come out in an exit interview. She will tell you exactly what’s important and you’ll be left with clearly actionable things. This is, in itself, a performance conversation.
What motivates you? If half of what motivates you isn’t getting met, I kind of have a disengaged employee here.
What motivates you? If half of what motivates you isn’t getting met, I kind of have a disengaged employee here.
And, they're not doing as well as they know they can do. They’re saying that without even saying it. You get to plan that out, and focus on that, and turn the dial. And so, the theory is if you can turn those things, the performance stuff should all come from it, versus thinking it in the reverse - “I need to have a performance review to tell her how to improve.” That’s old school thinking. That doesn’t really care about what’s important to her. And so, it’s not really an authentic, collaborative conversation. That’s why we think of these together.
Now, in a performance conversation it should be more of a reflection, sort of holding up the mirror instead of painting their portrait. Hold up the mirror and sort of say, “These are how things are panning out for you, but here’s a perception of your performance from your colleagues, and from myself, and from yourself, and hopefully you can see from any sort of blind spots or self awareness gaps that helps you sort of make it clearer is a starting point for engagement.
That makes a lot of sense. How long would you say these conversations are? How often are managers having them?
In this engagement conversation, there’s a little bit of pre-work to fill something out that they could do in 15-20 minutes at the coffee shop. When you have the conversation, you can build it into your one-on-one cadence. I would suggest quarterly, or twice a year to check in on this stuff. Any time you do it, I’d look for maybe a half hour conversation.
Your direct report will walk you through and say, “Here’s what’s important to me, and the first one is work-life balance.” The manager’s job is to just ask them questions. Make sure you’re not defending yourself. Inevitably, they’ll say, “I was put on this project, and it wasn’t really interesting,…” and you’re like, “Oh well hold on, let me tell you why I did that.” No, that’s not your role. You’re just asking questions to seek and understand...and get some clarity.
That takes about a half-an-hour. You’re taking notes in there. You’re seeking to understand how you can help the motivators get met. This is thirty-minutes a couple times a year.
How do you know that understanding the motivation, and having these one-on-one conversations, is really moving the needle for performance?
On the performance side, you’re assessing perceptions of performance, and also understanding, what are themes of high performance. The way that we have ours configured is we use text analytics, with a little bit of machine learning that’s inherent in our platform.
When you look at that even more closely you could say there are five to ten factors that you can say that are present in a positive sense that were associated with high performance. That could be, you know, assertiveness, or collaboration, or possibly a resilience, or learning...this person seeks to learn and they persevere through challenges. Each company might have slightly different things that are commonly associated with high performance but when you find them and they’re consistent, those are the things you’re looking for. When you use some machine learning with text analytics, that can help you cut through the bias of a traditional performance review.
How do you track these competencies? Or, factors of performance? How do you know an employee is using them?
The platform we’re using uses text analytics and looks at a string, whether it’s a 360 or a peer review, self review, or manager review...it looks at all the content to have multiple inputs on one person rather than basing it off of one thing. It looks at the input and it tags each sentence when I say, and I’m talking about myself, “I think I do a pretty good job of translating vision and a strategy.” Then it would say, “positive tag for communication: David believes he communicates effectively.” It looks at those words as if it’s translating and it tags that as communication. Or, it could say, “I think I’m working well with others, but there might be times when I have difficulty responding in a timely way.” So then it says, “Responsiveness,” and that word is tagged as a negative.
It looks at the input and it tags each sentence when I say, and I’m talking about myself, “I think I do a pretty good job of translating vision and a strategy.” Then it would say, “positive tag for communication: David believes he communicates effectively.”
There are times it doesn’t tag it the right way, and so we have a team that looks at those for when it tagged something wrong to edit and tweak the language, but it is using your own text and then tagging it as either a positive or a negative on one of thirty-three strings.
Most things are positive, and then I can click on each of these and say, “Ok, this person was tagged ten times for collaboration. Great. They must be a highly collaborative colleague across peers, direct reports, and their more senior people. And there’s, uh-oh, there’s this other thing over here that was tagged seven times for responsiveness, so they must not be doing a great job in their responses or not responding with quality information.” You’re going to have to click that to see more about what it means. It’s going to do most of the first half of assessment for you, then you’re still required to understand a little bit more.
From that data, we can then do all of this analysis of themes of high performance and we can prove a factor of gender differences in bias and language, and try and control for that. This is in the same platform of an engagement understanding, so ultimately we understand those alongside each other.
Going through these analytics and taking time, does the employee get to see her own analytics? How much time would you say it takes from management, or would you say the HR department team to go through them?
We want to do this twice a year. In general, we have four quarters, so we’ll do an engagement cycle, performance cycle….engagement cycle, performance cycle. So, a quarterly cadence for checking in.
One of the most time consuming tasks of performance for the managers traditionally it’s reading all that stuff. If you’ve got a big team, if you’ve got a direct report, and each of them are doing five to ten inputs on the peers that they work with, this is a ton of information. This is like mountains of text and so managers have to go through all that stuff.
Traditionally they go through it in a somewhat biased way, “I’m kind of looking through the key strings that confirm my opinion on this person, so I’m going to copy this string word-for-word, put that in your review because your peers think you’re not doing a great job responding as I’ve seen from this one thing I’m going to quote.”
One of the helpful aspects of this platform that we’re using is that the managers are confronted to balance the sort of positive negative ratio of the feedback on specifically one of the themes that are coming up the most. That’s what you first see, and so you can click into them and understand them. A lot of the time consuming stuff that causes some risk in this cycle that we’ve already accounted for, by the way we’ve configured it, is to see the summary of the feedback from the text analytics from the person that you log in.
It sounds like you have a lot of data that drives your decision making or what to do next, and that’s pretty key in regards to performance metrics and measuring.
Ya, being analytical I think is incredibly important in understanding performance in a non-biased way, and traditional approaches haven’t really allowed for that. I’m excited by advancements in the machine learning and text analytics that help you understand sort of complicated mountains of texts. That being applied in both engagement and performance is an incredibly exciting advancement.
Ya, being analytical I think is incredibly important in understanding a non-biased way, and traditional approaches haven’t really allowed for that.
I agree, this is all very exciting. Is there anything that we didn’t cover that you’d like to add?
I mentioned exit interviews. In the real world, you understand your engagement so well on the individual level that you don’t find these things out in exit interviews. I think a related aspect here is that what I’m hopeful that we can do is better understand flight risk, and true causes of attrition by correlating that intent to stay question with what’s most important to the person, his/her motivator.
When you see there’s a population here that intends to leave and the biggest thing that probably is a worry is training (for example), it not only becomes part of that lever that you can pull in response to limiting the flight risk, but also kind of replaces exit interviews, frankly. It replaces that in how you understand what’s happening in our attrition.
What do you think other companies are going through regarding this specific challenge of performance? Do you think they share the same challenges? Do they have slightly different parts of performance that’s challenging to them?
I’m part of this HR community in tech, and there’s a genuine intellectual curiosity from a lot of colleagues in other really impressive tech companies, and they’re all curious and interested in cracking this and coming up with a better model. One thing that I think is a challenge that they might be faced with, either knowingly or not knowingly, is that the way these one-on-one conversations are evolving and being addressed, is through what I would call a surface level approach.
They ask, “How can we use this on the iphone? How can do performance reviews on mobile? How can we make it easier on the manager?” is a lot of the conversation that I see evolving, which is like making it simpler, making the user experience more friendly, kind of Web 2.0-ing it.
There isn’t a fundamental flipping of a model emerging in a lot of these conversations, it’s like, “Let’s try user experience changes.” So when we say, we’re going to do away with the rating, OK that’s interesting, that’s an experiential thing you're doing away with or changing, but how has that really fundamentally flipped the model?”
Or when we say, “We’re going to use it on a mobile phone instead of on the web now, and allow people to do like, real-time, ongoing notes on their phone about performance, and do the real-time thing.” Ya it’s interesting, but it’s still sort-of a flawed conversation based in a Frederick Taylor style performance review, which is what most performance reviews are.
Are you experiencing the same challenges in your organization? What are you doing to couple performance and engagement?