March 3, 2016
DX and Open Sourcing at Netflix
In this Speaker Series presentation, former cloud architect for Netflix and now Battery Ventures technology fellow Adrian Cockcroft discusse...
The talk today is "Using OKRs to Drive Results," and the subtext is "Secrets to Crushing Your Goals."
We're going to talk a lot about goals today, and I don't think you would be here to listen to this talk if it hadn't been for Google promoting the work of OKRs and how OKRs can be applied at scale.
A great book came out in the last year from Eric Schmidt, "How Google Works." In that book he actually said, "In 1999, John Doerr introduced OKRs to Google and changed the course of the company forever."
It's amazing to me that from 40 people, when Google was a small company, now it's 60,000, they still use this process quarter after quarter, basically 15, 16 years later.
I actually wouldn't be here if it wasn't for John Doerr as well. John is actually an investor in my company Better Works alongside with Bing Gordon at Kleiner Perkins, and I have the great pleasure and privilege of working with two major Googlers. One is Shona Brown.
Shona actually built out the entire business operations and people-operations function at Google. She was there for about 10 years. She's now retired. She hired Laszlo whom I mentioned runs people operations there.
It's an amazing kind of group of people I have worked with, and we'll talk a lot about OKRs, how they emerged, how they were established, where they are today and where things are going. But for me, my personal story is just feeling very privileged to have worked and to be working with these people.
We take the Google metaphor to an extreme at my company. By chance, this wasn't planned, we ended up moving in to the very first Google office on University Avenue in Palo Alto.
The photo above is when Google was 40 people, and they did their round of financing with Kleiner Perkins. The photo below was actually last week. We decided to do Hawaiian shirts and go out there and take a fun photo of that.
Now let's talk about a quick history. In the beginning we heard about MBOs, management by objective, and goal-setting in general. Obviously that's been around for a long time.
Peter Drucker instituted management by objective, which is a more formalized process of goal setting and some really good things came out of that. Obviously, getting people to focus on a results-oriented type of philosophy was great.
In terms of the gaps or opportunities, I would say generally they're not very frequently updated, probably annually. This was how people thought of management by objective. It was fairly siloed. It wasn't very cross-functional.
It was tied to performance reviews and compensation, and so as a result, once you tie things to performance management, they have to be private. You're not going to share your goals with people, because it's more of a personal kind of private thing. And very top-down as we'll see.
Starting in the '70s, when John Doerr was working with Andy at Intel, OKRs emerged as the management practice for how they wanted to run the organization. John was so impressed by what that meant in terms of galvanizing the organization that he actually took that to all of the subsequent companies that he's worked with. Whether that's Amazon, or Google, or many, many, many others, it's basically been the legacy of John and also the Silicon Valley.
To move to this much more modern, operating cadence around, how you do goal setting, I think this is really something that Silicon Valley has the opportunity to share with the world. The key benefits we can talk about here are shifting from an annual process to quarterly. Think about what that means.
If you were doing annual goal-setting today, that's a pretty infrequent kind of cadence around checking in on things, aligning your organization, being responsive to the market, being dynamic, being competitive.
Clearly, moving to a quarterly process has benefits, this concept of being transparent and open around your goals versus private and thinking of them just as performance evaluation weapons.
Another key thing we're going to talk about here is this concept of stretch thinking. Once you're freed from the constraints of using them as a performance management vehicle, you can now encourage your team to set stretch goals and really push themselves as hard as possible and to stretch as possible. We'll see that has the opportunity to even potentially open up new opportunities of innovation when you encourage that kind of behavior.
So let's just start with "What are OKRs?" By the way, show of hands, who actually does OKRs right now? OK, about a third of you. And who's thinking about OKRs? OK, a third of you and then a third of you came for the free drinks and dinner, I guess.
OKRs are simply a management methodology that helps companies focus their effort on what's important and to do that at scale across the organization.
I don't think you can be too small to think about OKRs. One or two people in a team, or a group, or a division, or a department, or a company, or up to literally 60, 80, 100 thousand people could do this at scale. If we think about OKRs, objectives and key results, that's what we're talking about, the objectives are "What do I want accomplished?"
I do think they can be very aspirational and significant on a personal basis. What's going to get you out of bed in the morning and also be significant for the company? They are basically aligned with the bigger picture of what we're all working towards.
When we think about key results, the key phrase here is actually "as measured by," my objective as measured by my key results. How am I going to accomplish that? How am I going to measure my progress on that thing?
From a key-results standpoint, make the objective as clear as possible, measurable, in limited time. It starts to be very specific about how you're going to do those things.
Why use OKRs? We talked about this a little bit, but the key here is disciplining your thinking and communicating to the rest of the team what you believe is important.
Get everybody coordinated around the key set of initiatives.
I was speaking with a CEO recently, and he said that he was worried they didn't do OKRs. They didn't actually do any kind of open, collaborative goal setting. It was a very traditional, annual process, and he said that he was worried that up to 25% of the organization was working on the wrong things. How would they know?
They're not malicious. They're doing their best, but because of this lack of clarity, they're wasting time. They're duplicating effort. They're working on the wrong things. OKRs are established indicators for measuring progress. They focus our effort.
A great quote from Rick Klau who's over at Google Ventures. Rick has a great "How Google does OKRs" video on YouTube that I would encourage you to watch, and basically his quote is, "OKRs tell you and your team what you're doing, and most importantly, what you're not doing."
Not only are we going to talk about all the things around tracking our progress such as setting goals, writing them down, coordinating with our teams, but it's really important, at the beginning of that process. to say, "What's important? What are we really trying to achieve? By expressing that, what are we also choosing not to do?"
That's part of leadership, basically having the courage to know what's important and know what needs to be differed, delegated, or just ignored altogether.
The following couple slides are borrowed from a deck that Andy Grove and John Doerr worked on at Intel, but I've updated them to keep it a little bit more modern. John uses this football metaphor. I ended up calling the name of the football team here "The Sand Hill Unicorns" because, I guess, unicorns are popular animals nowadays.
You can see here that if you're John Doerr and you're the general manager for the unicorn's team, your number one goal is to make money for the owners. So you might say my objective is to make money for the owners. We need to do two key things. I know if I do these two key results it will fulfill the objective of the organization, and those are win the Superbowl and fill the stands to 88%.
You can then start to see how I need to get Larry on my team involved in supporting this objective and how can I get Jack driving this as well. Our objective is going to be win the Superbowl, and the key results are a 200-yard-per-game passing attack, third-ranked defense in the NFL and getting the punt return to the particular target that we need.
Similarly, Jack has his own objective to fill the stands to 88%. Maybe Jack has to go back to OKR school because his key result is "hire three colorful players." I'm not sure exactly how we're going to measure that. So maybe we need to do some coaching there for Jack. But a couple other key results make sense.
Interestingly, and all of this can cascade down in this particular example, we actually see that the key results for my objective might actually become the objective for the person that works for me and subsequently add an additional layer of key results.
I think that whole cascading process demonstrated here can happen sometimes, but it's not a requirement, and all we're doing is talking about, at each level of the organization, what's important and how do we make sure that we're fully coordinated on every element of the plan.
Then if we talk about some practices, clearly if we're going to be focusing on what's important and what's also not important, then we probably can't have too many objectives. Typically we see that having three, four, five objectives tends to make sense.
If you're familiar with the Alphabet news from Google, I actually think that one of the reasons why they're moving to the six-pronged business model, of having six key businesses now, is so that they can actually have three to five objectives at each of those six.
Imagine if you're Larry today, and you've got drones, you've got longevity, you've got health care and you're doing work in self-driving cars. With all these different programs and the ad business, how do you come up with a set of objectives that represents all of those and keeps that at three to five across one entity? That's really a difficult thing to do. By breaking them out into those six divisions now, I think that actually allows them to do their OKRs in a much more scaleable way.
Interestingly, at Google, they do encourage that a very large percentage of their objectives are sourced from a bottoms-up standpoint. The example I gave in the kind of football analogy was a very top-down, cascading kind of model.
The boss then sets the key results which then become objectives and so on. But there's a lot of research, and we'll talk about it, that around when people actually set their own goals, and they set their own targets, and they're creating this bottoms-up aspect to how they should express that, it actually, interestingly, increases the sense of ownership.
When they create their goals on their own, people that end up setting their own targets against their own goals set a higher target than they otherwise typically would have been assigned. And their belief of attainment is actually higher in that higher target than when it's assigned to them.
Basically the point is is that clearly, as leaders, we have to set the directional trajectory of the business. But in terms of how the details are going to culminate, there's a lot of benefit to allowing the organization to bubble that part up, if that makes sense.
Does that mean at Google, with the examples that I've talked about here, does that mean that they don't do performance management? Well, yes they do still continue to do performance management. The point here is that goal management, and how we do goals, and what we're working on, and how are we aligned and coordinated is separate from how are you going to be evaluated for your performance.
That doesn't mean that they're completely unrelated, that they don't cross paths ever, but it does mean that your goals might inform some of your performance conversation. It's not the entire basis of your performance conversation.
John's actually a great example. If how you set goals is going to directly inform your performance evaluation, then you're obviously encouraged to set very conservative goals, which is going against the spirit of what we're talking about if we're really trying to foster stretch thinking and innovation.
I remember John told me the story that he brought in sand literally, sand bags, to the sales leader at Google, because they thought that they were being too conservative in what was possible. So, yeah, no sandbagging.
The question might come up, "OK, well, if we're doing the goals, and we're doing that quarterly, maybe we're checking in monthly, we're aligned. We're coordinated. How do we determine if we've done a good job, and how do we score or assess, not for the performance management kind of reasons, but even just to inform our own? How did we do, and what can we learn?
I would say grading OKRs is very common. Not every company that leverages OKRs is already getting sophisticated around scoring and grading them, but many end up actually taking on an approach to this self-reflection kind of concept.
Here you can see that you could score them green, yellow, red, smiley face, neutral face, sad face, one to ten, and zero to one. At Google they're an engineering culture, so I think they were using two decimal points for their scores.
That might be a little too many gradations on assessing them, but here you can see that we came up with a blended score on this particular objective of 0.7, and this is where you have to determine what's your appetite for stretch. From a leadership standpoint, how do you think about how we really want to encourage risk taking or stretch thinking?
At Google they encourage 60 to 70 percent overall performance; it's considered a good score. That means that you didn't sandbag, but you did perform at a very high level.
You set stretch goals, and then you achieved 60 to 70 percent of them, and the thinking there is shoot for the stars and land on the moon.
It is maybe the way to think about it. I think you have to decide what's best for your culture. Let's say you work in a large company. You've never done collaborative goal setting before. You've just been used to doing anual goal setting. You might want to not encourage a lot of stretch thinking on day one.
It might be, "Hey, let's just publish our goals openly and start checking in on them frequently. It's a huge win." But I think you might find, after several quarters of doing that, "Hey, you know what? Maybe we can start to set more stretch goals, because we're starting to develop this concept of innovation and risk-taking, and that seems to be working for us."
A couple things I want to cover. Now that we've set a foundation for OKRs, some of the work that we're doing at Better Works is, "OK, OKRs, they've been around since the '70s. They're really starting to catch steam now, but what do we do to leverage all of the best practices that exist within Goal Science and Goal-Science Thinking and all of the best practices in this field?"
At Better Works we actually talk about the five pillars of Goal Science. I'm going to talk with you about those things.
The first pillar is this concept of connectedness. We've touched on that a little bit. It's not just kind of top-down goals, cascading, but it's actually how you do top-down goal development, bottoms-up, sideways, and do that in an open, transparent and aligned way. That's important when it comes to goal development.
Secondly, how do we create a supported culture and attitude around our goals?
To me, once we publish our goals openly, it actually creates a social contract to perform at a higher level.
I'll share some data with you about attainment that changes once you actually start to think of your goals in a social construct. Progress-based. There's a ton of data around recording your progress during the period that actually influences attainment. That's something that OKRs haven't really, I think, innovated on.
They were kind of like, "Let's make a thing open and frequent." But about recording progress during the period, that's something that I think is an even newer concept that we're trying to explore, and I'll share some data on that.
Adaptability just means as the business needs change, and as the priorities change, you should be able to adjust things. Some people ask me, "Oh, if I have goals or OKRs set for the quarter, and the business totally changes, should I keep running with those original goals that we've set? And the answer is absolutely not. Change gears.
Archive that goal, or move it out of the way, and come up with something new. It should always reflect what's important for you right now. In fact the reason that I got into goals, the inspiration for me was, at my last company we grew to about a hundred people, and we didn't have a formal process for this. So I started thinking, "How do I make sure everybody knows what's important and that everybody can share with everybody else what they think is important?"
I should be able to, not that I would do this, but I should be able to wake you up in the middle of the night and say, "What are you working on at work to create maximum leverage and value for your position in the company?" And you should have the three to five things.
I'm working on this. I'm working on that. This is how I'm creating value. This is how I'm creating leverage. This is how I'm driving results. That's all we're talking about. If the business changes, you should be able to change that list.
We've already talked about this, but there's opportunities to drive stretch thinking by taking this approach. And then there's also that scoring process that we talked about that actually drives even further aspirational thinking, because if you don't score and reflect on how you did, it's hard to enter a new period of goal-setting with those learnings and that enhanced perspective.
I'll just spend a few minutes going into each of these really quickly, and I think there's some data that I can share as we go through this around what we're learning as we go out and help companies do this at a massive scale, typically in the thousands or tens of thousands of employees.
We talked about that as just top down, bottoms up, and sideways. By the way, there's a really interesting HBR article, Harvard Business Review article that talks about the impact on operational excellence. For the most part it's the lack of tracking cross-functional dependencies inside companies.
What they say is that they found that these companies do the cascading. We already talked about that and they're really good at cascading. They've been doing cascading for a hundred years, right? But what they're not so great at is, "Hey, did Bob say he was going to do that? Or did Lisa say she was going to do that?" Did that get on their goals?
I can't remember, and now I need that thing to happen in order for my thing to happen, and the nature of these businesses now is so dynamic and so competitive, and agility is so important that if you don't start tracking all of these things, that creates risk around execution and business results.
Just a quick stat, we look at who's looking at goals in the system, and it turns out that people actually spend more time looking at other people's goals versus their own goals. There's a voyeur aspect to it which I think is pretty cool. People looked at their manager's goals more than their own, but even their peers.
They spend a lot of time doing that. Imagine how great it would be if you just joined a company and you could go and see what are all the people working on, what's important to them. Why do they think that's important? How can I align around that? What's the CEO working on? How can align around the key divisional things? That's what this kind of stuff promises.
We've talked about this, but if you work in a social kind of contract environment, how does that impact attainment? Some really great research: 78% increase in achievement when your goal is shared with a peer. Why do you think the Runkeepers and the Fitbits and all that stuff publish all that data to Facebook? It's because it actually drives your behavior, and it keeps you motivated.
Or you look on Fitbit and you see, "Wow, I'm only 900 steps away from my target. I better go out there and walk around the block a few times, because you know, if I don't I'm going to get taunted or cheered by one of my peers. So creating that kind of environment is really important.
By the way, in our product we don't taunt because that would be weird in a workplace setting. We do cheer and nudge, and that seems to work. That's very effective. This was some research that showed how goal attainment differs with how I express my goals. The grey bar, the really low bar, is I didn't even write them down. I just thought of a goal, and then I basically didn't do anything. Or I wrote it down, but I stuck it in my pocket. I didn't tell anybody.
Then the next thing is I wrote it down and I shared it on the white, you know, blackboard, white wall, for everybody to see. And then, finally, I wrote it down, I shared it, I'm getting feedback and I'm talking about it with somebody.
Just making these things social impacts performance. Recording your progress, similarly, has a huge impact on performance. A great example of that is Fitbit users step 43% more because they track their progress. Do you guys use any fitness trackers? OK, about half the group.
Imagine if Fitbit just sent the users an email at the end of the year with their step count. That's basically how corporate America works today.
We'll just send you an email at the end of the year. Or you actually have to go through and read all of your emails from the year, cumulate your own set of steps, and then put that into your performance review.
We've talked about this, but interestingly, we actually see data where about 20% to 25% of the goals for the quarter actually get adjusted and edited throughout the quarter. Not the progress being recorded, that's a separate thing, but the actual goal, or the details of the goal, or the targets of the goal are adjusted. We think that's actually a really good sign.
We talked about the aspirational, but here you can see a scoring example. Again, you can do it one through ten, colors, smiley faces. By the way, I'm going to show you an example of the goals that we use at my house. I use this with my family. We do goal scoring with my family, and we do the happy face, neutral face, sad face scoring methodology.
This is an energy company example. Their average progress of the goal, their actual progress attainment, was 82%. The way that they scored that was 97%. When I look at that, that tells me they had set a stretch goal and they performed really well to that stretch goal even though they came in at 82%. The team felt that that was a great score. Does that make sense for everybody?
Now we're going to go into a few goal examples by department, then I'm going to share some personal goal stories and then we're going to go to Q&A. We actually have a white paper for each of these departmental examples that we're going to make available to you after the talk so you don't have to write too much stuff down.
It goes into each of the functional areas of a company: sales, marketing, engineering, product, support, design. I just wanted to paint some ideas for you as you kind of think about this.
In the sales example, and a typical thing you might hear in sales is like, "Oh, well, they have goals. They get paid on commissions. Why do they need to have like their quota as their goal?And why else do they need to have goals?"
Clearly we see the goal is to hit the bookings target and contribute the bookings; 250k, that's a key result, and we're going to measure that. But we also need to do other things like deliver certain amounts of pipeline growth, or open up that healthcare vertical, or deliver three case studies for this new product that we just implemented and create some stories around that.
Similarly, a sales representative might have to not just drive the outputs that we're trying to deliver, but deliver even more inputs. Like, "Hey, you know what? Let's get 30 Fortune-1000 meetings this quarter and so forth. To me it's a balance of, "What are all the key things in this department that we need to deliver?"
Obviously you can see some of these would roll up to the manager. Maybe some of these are more at the individual-contributor level. What's really powerful, by the way, is if you really do figure out the recipe for how to scale your machinery and operate at scale, you can start to institutionalize this for any new hire that you bring on. And it becomes a very quick way to drive productivity and ramp up productivity.
OK, marketing examples. Here we could say, for PR for example, "Drive awareness through PR." So what does that mean? As measured by 20 press pieces, hosting two media dinners, secure the speaking spot for TED, that seems like a pretty stretch aggressive, ambitious goal. You'd be crushing your goals if you delivered that. Reach out to X number of publications to cover x, y, and z.
On the product marketing side, deliver an epic launch. We want 10,000 first impressions. We want to finalize the messaging. We want to create a certain number of case studies. You might say, "OK, I get all the quantitative stuff. It's super measurable. We're doing a good job at that."
But some of this stuff is a little bit more qualitative. How are we going to measure that? It could be as simple as, say, "Hey, you know what? Let's exit the quarter being on track in this area." It doesn't have to be more explicit than that if that's the best we can actually say.
On the engineering side, ship some of these features. Maybe focus on quality or focus on performance. There's always something that we can find that somebody can own in the organization that is going to drive value. Maybe squash a certain number of bugs or create more process.
An interesting story at LinkedIn, for example, they have results-oriented goals. They also have leverage-oriented goals, and so your results are like, "I'm going to squash 50 bugs. I'm going to do this. I'm going to do that." But the leverage is, "What are you doing to make your job easier for yourself and for others in the future? And what are you doing today?"
I'll squash the bugs. That's my results. But my focus is on making the process better so we don't have these bugs. I'm going to create a solution for that that's going to help scale my position going forward. Product goals, you can see the examples here: increase engagement. Maybe implement the new onboarding system.
On the technical writer it's, "How do we consolidate our redundant documentation, or implement a new style guide, or share and communicate that with the team, or standardize that across the team? Some good ones there.
Similarly, on the design side, we see major interaction, so we need some prototypes. We want to drive a certain amount of engagement or usage or MAU, weekly active usage and then user research. Maybe we want to start to go and do customer visits.
Actually, one of our key product goals across our organization is we do 50 customer visits for the product team every quarter, and we want the product team to hear from the customer in person and share the notes on the Wiki as to exactly how that process is going. When they actually check in their progress on that goal they comment a link to the Wiki, and everybody in the company can see that.
Then the last example is support, and so here we see making customers' experience with support more enjoyable. That could be around hitting some performance targets like service level agreements and things like that, or personal satisfaction scores, and then maybe there's a community goal. We're going to start to activate our community.
We want to publish content. We want to increase participation, and we want to start to identify and develop these evangelists and key contributors. It doesn't mean that we are going full metrics at metrics' sake. It's about saying what's important and expressing this in a measurable degree to the degree that it's possible.
With the last few minutes here I'll just cover some personal examples, and then we'll go to question and answer.
This is a real goal. I just screenshotted it this morning. This is my sales leader. The sales leader just joined the company about three weeks ago, and we have a practice at our company, we're 50 people now; for everybody that joins the company, we want them as productive as possible as quickly as possible. And so it means that the hiring manager has to do a welcome letter where we actually go into pretty significant detail on who the teams are, what people are working on, who you'll be working with, etc.
Everybody sets up a 30-day, fast-start goal. It cuts off about halfway through, but here you can see, "Meet with each of my team members. Meet with 15 prospects. Meet with three customers. Make five account introductions. Review all of the key materials like contracts and sales materials." There's a few other ones there.
What better way is there to have a new person ramp up more quickly than to actually know what's clearly expected, and to give them those guardrails and a clear path on being successful? Again, it forces a great conversation of, "What do I think 'productive' means? How do we get productive? And let's capture that.
A second example, this is my marketing leader, Minka. This is a goal from last quarter, and I went in and screenshotted this for two reasons. One is that here you can see that the goal was to deliver 750 marketing qualified leads, and you can see some description of what that means.
What's interesting there is I think it's really cool if other people can see why you have this goal and the context around this. It's not just, "The target was 750." Give me some more details. Help me understand this. You can see that she put that there.
Second thing that's interesting here is that you can see a little cloud halfway down the page, in the center, and that cloud means that it's powered by Salesforce. So she's not checking in every lead that she generated. She just connected it to Salesforce, and as she makes progress on her leads, it automatically checks that in for her, which is kind of cool. You know, automating that thing.
And then the third thing that's interesting here is, you can see in the stats on the right, under the little tree, there's a stats thing that says "180 views." I moused over that, and I took a screenshot for you guys on this point, which is, and it's a really long list, "180 people" or "180 unique views." But when you look at the list of people that were viewing this goal, it wasn't really the marketing department. It was engineering, the design team and product team.
Other people cared about how were we doing on this goal because they realize it's an important goal for the company. So I think that's a pretty cool thing, and you can see by the way she received 23 cheers to the right of that, which is pretty cool. It feels good.
Then this is another live goal. This is Jonathan, is my head of engineering. He actually was one of the co-founders of Siri, went over to Apple, came over recently to join us to lead engineering for us. And if you just look at his goals, you can see that he's working on quality. He's working on engineering capacity, security, and the release schedule.
Also, we measure story points as one indicator of productivity. You can see, and by the way, we actually connect a lot of that stuff with JIRA to automate the collection of some of those data points.
Last example, this is my family or two of my boys. In this picture they're 12 and 10. Colin uses, we all use, Better Works at home. We don't actually have a home edition, so I bought a special access to this, and you can see the goals.
Colin's goals are, "Fight less with brother, only three times per week. Get straight A's. Hit a homer," and "Run on the weekends."
What was cool is that he did hit a home run last season, and as he was running from third to the home plate, he looked at me and he goes, "I'm checking this in."
The younger brother, the 10-year-old at the time, one of his goals was "Find a new friend." And he got to 100% on his goal. So I thought that was pretty cool.
All right, last slide. Why are we doing all this? We talked about the focus and the attention and what not, but a lot of the spirit that we have at Better Works is, "How do we help people get one percent better every day?"
Small improvements, small adjustments every day, consistently over the long term, have massive implications on yield, results and productivity.
Here's an example that I think is really profound. If you get one percent better every day, just because of the laws of compounding, you're actually 37 times better than when you started. Thirty-seven times better after one year. Compare that to the person that maybe doesn't care about getting better. In fact, they could maybe get one percent worse every day. They're actually .03 compared to the person that they were at the beginning of the year.
If you compare the two people, the one person that's getting one percent better versus the person that's getting one percent worse, it's actually at 1260 times difference between the two people. So it means that you could have, if you have one person that's focused on getting one percent better, effectively 1200 of the "other" people.
In start-up world, these companies are so precious. The results are so meaningful. Every customer matters. Every input, every action drives massive consequence. I think you can see the importance of being really focused and driving the execution with the team. That's the end of the presentation, thank you.
Is anybody working for a company here that has more than 50 people? OK, so these are all small companies. I guess what I would encourage is to shift away from the thinking that we do goals for the purposes of performance management. In which case, maybe, I would bet that most people here don't have a very formal performance-management process.
It's a very fluid or non-existent process, and so if you think of goals as performance management, you probably don't do a lot of goal-setting. Or you might say, "We might have a couple company goals, but we're not going to push that to the individual to really drive personal accountability, alignment and coordination on what we think is important.
I just wouldn't use performance management as the lens around how to think about why we do goal setting. The reason we do goal setting is to focus on do we all agree on the work that needs to get done. Does everybody have three to five things that they're really on the hook for? That's really it.
Obviously, step two is, "Can we align and coordinate these things so that your goals ultimately support mine?" Or you're at least supporting some key initiatives in the company, just so that you're not off doing your own thing, working on things that actually don't matter to the success of the company.
Then I would treat evaluating performance as a separate thing. For example, at our company, the way we do performance feedback is every hiring manager will say three cheers and three nudges for the person, and that's typically independent of your goal execution. It could be related to it, like, "I feel like you could be setting more stretch goals."
But the point is that you're still capturing some feedback through your three cheers, three nudges, three pros, three opportunities or however you want to frame that. I think that that part of it is far less important than, "Are we coordinated on what work needs to get done?"
I think you'll naturally find that the people that are performing, you're going to know who those people are. And the people that aren't performing, you're probably going to have a good sense for who those people are as well.
Everybody has goals, and everybody has three to five goals. Everybody's driving what's important. Some people may have compensation tied to some of their goals, and so let's look at that sales example. We came up with five goals for the sales person: sell the quota, open the healthcare market, increase our cross-sale percentage on this new product and brief 25 executives.
Only one of those goals is actually going to be tied to compensation. I think that can happen anywhere. That could happen in whatever system you want to use for your HR system, a spreadsheet system, however you want to kind of track that process.
Interestingly, you can also make a case for somebody getting compensation but actually not carrying any goal related to the compensation. So here's an example of that. For my leadership team, we pay bonuses for exceeding the annual bookings target of the company, and they don't carry commission on that.
It's just a bonus if we overachieve the bookings plan, because I think that everybody's involved in that. Product's involved in that. Engineering's involved in that. The whole company's involved in that. Only the sales leader in the company actually carries the booking's number as a goal in Better Works. So I think that the nuance here is that everybody should be able to express that.
Some people are separately, and maybe in parallel, incentivized financially for certain kinds of things, and in some of the cases, those are directly related to some of their goals. In other cases it might be unrelated to their goals altogether. I think a mistake would be, "Oh, you know what? If you do 67% of all of your goals as a blended average in Better Works, that pays out this bonus." That will then go back to reinforcing the sandbagging concept that I talked about earlier. So that would be the worst case example.
Absolutely, and I would also say, at the same time, there's a ton of research that supports a lot of these management by objective or, "You do these things on a quarterly basis, and then you get paid these things."
There's actually some interesting research that says those things don't actually drive the right kinds of behaviors anyway because they have to be structured in such a manner where you're getting guaranteed payouts. Because otherwise they're too punitive or they over-emphasize the wrong behaviors, and so then it's like, "I'll do that at any expense, even quality."
There's a ton of examples in that area where we'll reward people for fixing cars faster, but then the quality of the repairs plummets. And then there's all these other quality and safety issues. So at our company we don't use MVOs.
We do have some goals that are commissionable, because I think that's an expectation for the front-office side of the house. Because compensation can be such a very difficult topic, I actually think that we just simplify things to say we pay people a base salary.
Equity is the upside for why people would join a start-up, and let's just make sure we're really aligned on the goals, what's important and how we're going to measure success.
It's almost like something that I think happens a little bit organically, and it's almost like you have to trust the process to work itself out. Some of it's going to be top down, some of it's bottoms up, and it's like if you've edited a Wiki page. How is that all going to come together? People start editing and adjusting, and then things emerge and the painting emerges.
I guess what I would say is it just takes practice. The first time you do it, I wouldn't actually worry about trying to get every detail. Like, "Oh, the book, page 98, said do it this way." All we're talking about is just, "Does everybody know what's important, and can they express that in a coordinated fashion?"
What you might find is that this department is really way too ambitious, and this department is way too conservative. Let's just go with that. We'll come back in a quarter. We'll learn from it and get better every quarter. Google's had 65 attempts at doing this back to back, so they're going to be really good at it. Versus if this is the first time you've done it.
I think some people are very goal oriented, and this is very natural. It's like, obviously, "Why wouldn't you do this?" Goals, metrics, accountability and performance, and we're getting results. And we're going to be very coordinated this way.
Some people, I would say, my experience has been, are more task oriented, and they don't have a goal muscle, but they're always busy, and they're staying really busy, but maybe the bigger picture of why we're working towards this, what's the compass or the map where we're going. That's harder for them to kind of express in that way.
So I don't know if that answers your question, but I guess it's practice makes perfect, and you'll find deadlines do a pretty good job of setting a framework for how we're going to do this. If you said, "Hey, you know what? I want all managers in the company to have all their goals in by Friday, and we're going to sit down and review them, and talk about them, and have some back and forth," I might say to one of my leaders, "Hey, you know what? I noticed that you didn't have security in there as a goal. You know you still have to have this gut check of 'wow, does this make sense?'"
Does it make common sense? Let's make sure we have a security goal in there, because I want to make sure that every quarter we're making progress on security. And then you might say, "Actually, by this Friday all manager goals are due, and then by next Wednesday everybody in the rest of the company needs to have published their goals."
That means that every manager needs to have met with everybody on their team to talk about, "Well, hey, why did you pick that?" Or maybe once they saw what the manager's goals were, they could actually start to build out their goals. So it has an emergent aspect to it, but you could be much more top-down in the beginning if that's how your bias is.
I think it might be hard to be entirely bottoms-up.
At Google, remember, 40% top-down,
So you definitely have to set
a general trajectory and let the people fill in the details.
If you're going to just say,
"Everybody just figure out what you think is best," that could also be very dangerous.
OKR Process for Small Companies
I think it's a hard question for me to answer, because I think everybody just needs to have a clear expectation. So, for example, I wouldn't want to work, personally, in a place where it's not clear what everybody's working on, that they can't express the three to five things that they're focused on, and that we don't have a common and shared understanding around how your three to five things, and my three to five things, result in company success.
The nice thing is, if you have the luxury of being small, you could probably set out your goals and OKRs tonight.
You go around the table and, "Bob, what are you working on? Lisa, what are you working on? Sarah, what are you working on? OK guys, that's what we're working on, and you know what? We're in month two of the quarter. We've got another month and a half. We've got six weeks. Let's see how much progress we can make.
"You know, Bob, I think you're stretching yourself a little bit on that goal. But if you want to keep the target where it is, I'm all for that, and that's going to be a crushing goal if you can deliver that. Lisa, I feel like what you're saying is maybe a little bit on the conservative side. I'd like to see you stretch that."
This is called leadership. It's having courageous conversations around reaching our potential, stretching ourselves and setting up conversations where everybody can feel successful.
Here's maybe an interesting thing I've learned at Google. At Google it hasn't been this exact approach 65 times in a row.
They actually took polls of what we needed for the company "right now," and "How do we want to adjust this philosophy?" Everything that I've talked about is not a magic wand. It's just the simple framework for expressing what's important and us all being coordinated. And you might find, "Hey, everything is going super well right now operationally, and what I really, as a leader, want to encourage is stretch thinking and thinking bigger.
Like, "What's our TED talk for every department? What would just crushing it look like?" And I want people to start thinking about that. We're so buttoned up operationally that what I really want them to do is start thinking out three, six months ahead, nine months ahead, 12 months ahead.
It could also be that, "We're missing our financial results. Product is always late. Forget landing on the stars or the moon, we just need to get out of orbit, and I want to make sure everybody has a very clear, crystal clear expectation on what they're going to get done this quarter so that we're at least operationally buttoned up."
I think it's part of leadership to modulate, "Are we more operationally oriented around goal setting? Are we more innovation goal setting?" And it would be a failure for me to say, "Oh, you know, just set stretch goals, and everything's going to be fine," if you don't have the foundation of even meeting basic fundamentals. That's not necessarily going to deliver success.
By the way, one thing I would say on financial goals at Google, those are not stretch goals. They're expected to deliver at 100% of plan.
Framing that is if you're pre-revenue, what do you do? It makes me wish I had pulled out my goals from Q4 of 2013 when I started the company, because that would have been a cool thing to show.
We were three or four people, and we were just starting to explore this area. I remember explicitly, one of our goals was in Q4 of 2013. We wanted to talk to 80 companies to verify our hypothesis that goal setting and doing goal management as a service would be a good business idea. That quarter we talked to 86 companies, and that was based on two per day, and then once you say, "OK, we're going to talk to 80 companies."
Then you probably have to figure out, "How many do I have to contact in order to get a reasonable number of people that are actually going to talk to me?" You probably have to reach out to whatever that is, three or four hundred companies, and then, if you're going to reach out to 400 companies, then you probably have to figure out, "How am I even going to find those companies? I'm probably going to have to look for lead sources on LinkedIn or whatever that might be."
So you can back into all of the work that needs to get done. I'm going to research 400 companies. I'm going to reach out to 400 titles and targets on LinkedIn. I'm going to convert maybe 25% of those and have 100 conversations, of which 80 actually might actually happen.
And then we're going to do an assessment on whether we're going to share the data on what we learned about those interviews or those discoveries. So I would say I don't think a company's ever too young to explore a way of measuring success. It might be as simple as validating your hypothesis. But I would do it in a data-driven way that is actually a very measurable and organized way that everybody agrees with upfront that would ultimately convince them that we're on to something.
I would say the goals are not comprehensive. They're not everything that you're doing. It's what you are trying to bring extra attention to. I'd ensure that my people-ops goals, my leader there, payroll is not one of her goals, but moving our office and all of the logistics around moving 50 people is a key goal for her this quarter.
There's a whole set of things that we're doing in the culture area which she's a key stakeholder to drive. Then our hiring. We obviously have to add a lot more people this quarter. So hiring, culture, and some major office logistics are the key things that are getting that extra attention.
Yeah, and that's how work feels. You're spinning the plates. You're going from conference call to conference call, meeting to meeting. And I think of goals like that, a bigger map, that compass where you're headed towards.
You could absolutely have taken a lot of meetings, done a lot of conference calls, and feel exhausted at the end of the day. But did you actually move the business forward in terms of what we're really working towards? Only you can be the judge of that based on your goal progress.
The most important kind of motivator, I think, is actually this concept of making the goals open and social. Back to that research of attainment around public and progress, I think that kind of that concept has the maximum impact on achievement and attainment.
We do have some privacy settings. You could add privacy to goals if you want to do that, and there's certain reasons why you might want to do that. We can look at the average attainment. Imagine this across 100,000 goals. We can look at attainment, on average, of public goals versus private and see fundamental differences between the two when all other things are equal.
I would say at Badgeville the incentive was, I don't think, necessarily extrinsic. You were getting a virtual incentive for delivering desired behavior. Here I would actually say that you are still building your reputation inside the company, and one reason why I think you're compelled to perform is you become somebody that does what they say they're going to do.
They support other people's initiatives by aligning correctly, and they're great collaborators. They support their peers. So I think a lot of the stuff that we're talking about is just being a good citizen inside the company, and if you give people a framework for how to express that, like, "Are they execution-oriented? Do they deliver on time? Do they support others," etc. I think it's very natural for people to want to play in that kind of that world.
We are just starting to do some interesting analysis in this area, and we're now starting to build out a data team mostly coming from the health and fitness world. It's how we're thinking about it.
The only research that we found is however active your goal expression is, it actually does have a relationship with your ultimate attainment, if it's very passive versus very active. Even the goal name, the title of your goal, ultimately impacts attainment.
So that's a fair question. It's just interesting to see. Some of the things that we think about are like, "Well, maybe if somebody expresses their goal a certain way we should encourage them to express it a different way," and we're going to start testing that as well.
Is everybody ready to do some goals? Thank you for your time.