You Got This!

WTF are OKRs


Object Key Results (OKRs) are a goal setting tool used to communicate what you want to accomplish and what milestones you'll need to meet. OKRs are defined collaboratively and are designed to cascade through an organization - from senior leadership and down through every level. If the objective is to "exist in 100 years", each key result may be items such as "increase donations by 30%" or "reduce exhibit shipping costs by 5%". But each key result becomes an objective to the next level of goal-setting.

Things Michael learned the hard way:

  • The OKRs themselves don't matter. The real value is in openly communicating organizational goals.
  • OKRs don't fix dysfunctional orgs. OKRs are great for gaining alignment, not building strategy.
  • If you don't have a baseline, set one.
  • Pair quantity metrics with quality metrics. Pumping numbers is easy. Focus on outcome over output.
  • Hitting 100% of your OKRs means you're not reaching high enough. You should hit about 50%.


Good morning, afternoon, or evening, everyone. Welcome to OKRs -- you can fill in the rest! I'm Michael. I'm @mheap. You can find me on Twitter there about the talk. If you want to say bad things, put it -- DM me. I'm an unashamed OKR lover when they're done correctly, but, unfortunately, they're not done well a lot of the time. Today, I'm hoping to help guide you through what a good OKR looks like, and how they can help you in your work life, your personal life, things like that.

We've only got ten minutes, so I'm keeping it quite brief. We will start at the beginning, and spend two minutes on what is an OKR? Why are they helpful? How do you know if they're working? Then we are going to walk through how to build an OKR. This is where it goes wrong in a lot of companies. Three minutes on that. Next, hard-earned tips on OKRs. Things I learned on the hard way. And then some examples of OKRs.

Part one. What is an OKR? OKR is an acronym short for "Objective Key Results" -- OKR. OKR is an effective goal-setting and leadership tool for communicating what you want to accomplish and what milestones you need to meet in order to accomplish it. And the important parts here are these two that I've highlighted, what you want to accomplish and what milestones you will need to meet. Now, it's just like any other goal-setting. The acronym OKR is really just a fancy way of saying what do we want to do and how will we know when we got there?

That seems pretty straightforward, right? So why are your OKRs so popular? We know how to set goals, make them SMART -- all of that stuff. Why do people really keep going on about OKRs? The secret's in how they're defined. They're built collaboratively rather than everyone choosing their own goals independently. In fact, this is a great time to take a look at how those OKRs get built.

In most orgs, you have a goal, and your manager has a goal, and your team has a goal, and the colleague has a goal, the CEO has a goal, the CTO, this other team, that other team, get what, they are usually all different, and most of the time, they're probably directly in conflict. The engineering team say fantastic, we're going to move to the cloud. And the finance team are saying, no, no, we're trying to cut our cloud spend, we are going on prem, because it's cheaper.

This is where OKRs come into their own because they're designed to cascade through an organisation, so the company leader setting an objective which informs your team objective which informs your personal objectives. We are all working towards the same thing but the objectives get smaller as you head down the org. This is all very theoretical.

So let's just build an OKR together right now and to prove to you that OKRs work in all kinds of environments, I'm not going to build a tech-related goal. Instead, I'm going to pretend that we all work for a museum. And that our goal is to ensure that we still exist in 100 years. So, the person asks their goals to set goals for the next 12 months to make sure that we still exist in 100 years. They get the following response. We're going to increase donations. Attract new exhibits. Target new demographics, and decrease costs. And those all make sense to me. We are going to make more money through donations and revenue, decrease costs, and make sure that we still have exhibits that people want to come and see.

And this is okay, I guess, but they're not exactly measurable. How do you know when you've achieved any of these? So we go back to the drawing board. Instead of increased donations, how about increase donations by 30%? Seems ambitious. But we will definitely know if we hit it or not. How about attracting new exhibits? Let's go "Run one visiting exhibit per season", that's four times a year, keeping things fresh and having people visiting repeatedly. How about new demographics? Let's be specific. We need to target younger generations if they want to visit, bringing their children and grandchildren into the future. And then finally, decreased costs. This is always on the list.

But we still need to be more specific, and we need to align to the overall objective because it would be easier to say we are going to decrease costs. Let's reduce head count by 40%. If we did that, how would we achieve all of our other goals? OKRs are about working with other teams towards an objective that we all share, and with that in mind, let's think about how we can decrease costs whilst supporting our key results. How about we reduce shipping costs by 5%? If we're running lots more visiting exhibitions, the 5% saving on shipping will be worth even more to us in the future.

These four items are our key results. If we hit them, we can be confident that we will exist in 100 years. But the beautiful thing about OKRs is that these key results are also objectives for the next level down the org, so the museum director says we need to exist in 100 years. And one of the direct reports says, okay, I'm going to take a look at increasing donations. And that becomes their objective. So to increase donations by 30%, we can add a super supporter tier as a higher donation threshold, show people how much it costs to run a museum. We can run an email campaign soliciting donations. And we can educate people about Gift Aid so that we can get an extra 20% from their donations at no extra cost to them. And it just keeps going.

Let's take the five interactive screens. This key result to increase donations by showing people running costs can be broken down again into actionable items that individuals can pick up, someone might be tasked with research, reporting on the top three screen manufacturers. Someone's got to build what the screens display. We need to decide where the screens go. And, of course, we need to get health and safety sign-off.

OKRs help you understand how the work you're doing benefit the org's wider role. Your individual research report on the top three screen manufacturers allows us to install interactive screens showing museum running costs which helps us increase donations by 30%, and all of these are done in service of our main objective. To exist in 100 years.

Once you know what your org goal is, OKRs are super useful for every organisation. With OKRs, if the project is working towards the overall objective and we should be helping each other, or it's a distraction, we should say no and focus on the things that move us towards that objective.

Now, I'm going to take a second and go off on a tangent and tell you about the British rowing team who unexpectedly won the gold medal in the Sydney Olympics in the year 2000. It is worth it, I promise! They achieved this amazing goal by asking one question of themselves and each other: "Will it make the boat go faster?" Should I do this 90-minute work-out? Will it make the boat go faster? Should I go to the pub tonight? Will it make the boat go faster? Should I have eggs for breakfast? Will it make the boat go faster? With OKRs, we can ask: will this help us achieve our objective. Things I learned about the hard way. I've made a ton of mistakes so hopefully you don't have to.

Biggest tip I have is that the OKRs themselves don't actually matter. You will hit some of them, you will miss some of them, but the world will move on. The best summary I have came from a Hacker news comment. The real value in OKRs lies in the process leading up to the process defining them not in the objectives and the results themselves, because those OKRs make you discuss what your goals are as a company. They make everyone get on the same page and maim your goals explicit because you have to write them down.

That being said, OKRs don't fix dysfunctional orgs. The process of discussing goals and explicitly committing doesn't work if your exec' team changes their mind every other week. OKRs are great for gaining alignment, not building strategy.

If you don't have a baseline, set one.

That is your first key result. Focusing purely on quantity metrics, and my favourite example for that if you want a thousand new users, I can do that. Just give me budget for pizza, and I will visit a university campus. What you need is a quality metric. 1,000 new users with a 20% conversion to paying customers. You need to focus on the outcome over the output.

Let's go back to the museum example, because we had a key result of Gift Aid education fliers. This is an output as a key result. We can do a thing, but we don't know if it is having the outcome that we want. We can run a very successful on time, on budget campaign, zero printing errors, and still get zero new gift aid donations. Instead, we should focus on the outcome of our key result, we aim to increase the number of Gift Aid donations by 20% and the flier campaign may or may not be the right way to do this, but now we are focused on what we are actually trying to achieve.

Limit the number of OKRs you're working with, two to three per team with two to three key results per key objective. Any more than that, and you will be pulled in too many directions. Things you say no to are as important as the things we say yes to. Hitting 100% of your OKRs is a mistake. If you're hitting every goal you're not reaching high enough. Common guidance is to hit about 50%. This doesn't work if goals are tied to compensation such as bonuses. It should not be. If goals and bonuses are tied, then there are organisational issues to work out if you really want your company to thrive.

Finally, it will take you multiple attempts to get this rights. OKRs are easy; organisations are hard. It will take a few iterations before you feel you're hitting your stride. All right, we're on the home stretch. Let's take a look at some real-world examples.

The first is from Allbirds who make shoes. Their aim is to have the lowest carbon footprint, no pun intended, in the industry. To do that, they're looking at the supply chain, shipping infrastructure, 100% zero waste, paying carbon offsets, and 20 per cent of their material is compostable and 25% biodegradable. These are focused and measurable for Allbirds. OKRs work for individuals in addition to companies. Imagine you're training for a 10k race. You might be aiming to run it in under 50 minutes. So you go for a run three times a week for at least 30 minutes. You increase distance of the run by 1k every week, and you decrease the time per kilometre by five seconds every week. You will quickly know if you're on track to run that 10k in 50 minutes by March. Then the last OKR is one for you.

You can use OKRs to set OKRs. If this talk has resonated with you, why not give them a try in your organisation? Start small, with just your team. Organise an OKR knowledge-building session with your team, two to four key results each, and get 100% commitment back from the team to try OKRs for the next six months. All right, I'm out of time. Thank you for your attention. I've been mheap, and you've been awesome!