This article is an abbreviated version of a book I wrote called OKRS in 30 days, which is available on Amazon. If you purchase it — i’ll give you a big virtual hug.

Moving to OKRs in 30 days: A step by step guide

I have been talking to a LOT of people over the last few months about OKRs — more than I thought I would actually. It is such a hot topic, and I like OKRs, so it’s fun to talk about, but I felt with each conversation I was mostly covering the same ground.

So I decided to write a complete guide on getting your team from where they are, to fully using OKRs in 30 days. I don’t just cover what are OKRs or how to set them, but how to actually manage the change element. How to have meetings with the team, what to cover, how to get low performing teams back on track and then how to persuade everyone in the company to use them too.

It’s totally tool-agnostic — you can use a piece of paper if that works for you.

It is not a quick read, but it is by far, the most comprehensive way of getting you to learn everything you need to about OKR’s without having to read a full book or go on some course.

Read this article if…

• You don’t have time to read Measure what Matters
• You want to move to OKRs, but are finding it hard to sell the idea internally or are a bit confused about the implementation process.
• You need a practical, step by step approach to get it done.
• You need help with the change management parts
• You want to know the tricks to do it right.

*Note. The screenshots in this article are from an OKR performance tool called OHNO which no longer exists. Sorry 😥, that means you can’t use it. But there are others you can use, lots actually — even a simple spreadsheet will usually surfice.

Imagine… what if you could get everyone rowing in the same direction?

Creating some OKRs is one thing — but how do you actually move to the process of using OKRs? How do you make sure everyone buys into the idea and embraces it enthusiastically? Well, I decided to write a super simple guide on how to make the transition. If you follow this, in 30 days, you’ll be done, and you’ll never look back.

I decided to write this because I was getting messages from people about how they actually roll out OKRs. At first, I decided to run some free workshops for our customers. These worked well but to make the process easier, (and so I could stop repeating the same content over and over) I thought a step by step guide might be able to help more people.

So this is a comprehensive guide to setting goals using OKRs and rolling out the framework in your team, and then more broadly, to other teams, if that’s what you want. This is everything I know about the process, and what I have learned from talking to different teams, and chatting about where things have gone wrong.

I do have some alternative views on OKRs, which I’ll go into in detail, but I think I have found which compromises and adjustments make sense for me, and which ones I suspect resonate with most people.

This guide is tool agnostic, you can track OKRs on a whiteboard if you like — you don’t need anything fancy. I also make sure to cover the more human elements. How to run meetings, how to help your team succeed, and generally just how to ensure the change management process is taken care of.

Part 1: What’s your number one priority?

Picking your one thing

So you’re looking to make the transition to OKRs in your team? Nice move. I like your style. The very first step on this journey is a big one, but I’m going to walk you through it. Grab a pen, make some notes, and follow along.

The first thing you want to do is set one big priority. This will become your very first objective (probably).

Why only one? When you first start with OKRs you don’t want to take on too much. Most teams make a mess of it by adding on too many objectives and are disappointed with the results.

Trust us here. The first cycle you do will be a mess, just accept that and get the cadence and routines down. The second cycle is where things start to shine.

So it’s best to just pick one thing and to do that you want to make sure it’s important.

Important vs urgent

We use a prioritization matrix (some of us at least) called the Eisenhower priority matrix. It’s designed to get you thinking about the differences between things that are urgent (like annoying phone calls or unproductive coffee meetings) to things that are important, like professional development or attending your daughters' football game.

If you’re going to pick just one goal for your team, make sure it’s both urgent & important. I’ll give you some context and an example to help.

It might be urgent for you to raise venture capital, or perhaps its really urgent you close a big sales deal

or get a proposal done for your largest customer, but is it more important than hiring for your team?

Hiring more people, while it might not be super urgent, will enable you to delegate some things around and lighten your load. By doing the recruitment items first, you’ll have more time, and can tackle more urgent things as they come up. So think deeply on this point. Whatever objective you have, it should be both important and urgent.

What comes first?

As we touched on above, if you get really stuck, try to think about the order in which things need to happen. Are you picking something that needs to be done in the next 3 months, or can it wait?

Think about the dependencies of your objectives. Do some need to come before others? A good example of this is recruiting salespeople. Sales might be very important for your business, so you might think talking to customers and winning deals is the way to go. But hiring salespeople first might make more sense.

What really counts?

When you are really unsure of your number 1 priority, try this technique. Ask yourself, what's one thing that will send you broke if you don’t do it soon. Sometimes thinking of the worst possible scenario and its implication will help you clarify what’s important.

A second technique is to ask yourself if you accomplished “Hire 10 salespeople” today, would you sleep better at night? Sometimes knowing a task is done is a huge relief to teams, it can relieve pressure, which frees up teams to move onto their next important thing.

How your mission can shape your priorities?

Values led businesses kick ass. You should be one. Your mission, your purpose and your values should be your north star. So why not let them lend a hand when it comes to shaping your number 1 priority.

Ask yourself questions like, if we need to arrive at our mission early, what’s the first sign we’re getting closer? An example of this might be (if our mission is to land on Mars by 2024), our first sign we are getting closer is we probably have;

Built 1 rocket.

Another way to work this out is to write down your first step, then write down all the subsequent steps to getting to your mission. When you see your first step and the subsequent steps, something may jump out at you as the immediate thing you need to focus on.

When everything is a priority

One alarm bell when using the Eisenhower priority matrix is when everything is in box 1. Usually, I have a rule, no more than 2 items in that box. Everything else has to go elsewhere. This is a good way of helping you cut some of the items from the list and narrow it down to the best task.

Key Actions for you:

• Your first objective should be both urgent & important.
• Of all your objectives, which one needs to be done first?
• If everything is urgent, pick 1. You have to.
• Your mission and purpose can help.
• Understand dependencies.
• Still stuck? Try the priority matrix.

Part 2: Get the team excited

Convincing your team

This is one of the bigger steps, so hang in there. The rest will be smaller, I promise. But it’s probably the most important step. You want to make sure everyone is 100% onboard with the changes that are coming.

One of the biggest mistakes people make when rolling out OKRs is not taking into consideration just how much people hate change. You might already be using KPI’s for individuals, so make sure that when you talk about the move to OKRs, you are sensitive to the fact that change management, in any capacity, is hard to do well. But we’re going to help you at every step.

Arrange a meeting

The first thing to do is set up a meeting with your team. Get everyone in the room together. The very first thing, before you even dive into what OKRs are all about, is to just talk about general performance over the last year. If you’re not sure where to start, try the following questions to get the teams talking.

What does everyone think our number 1 priority is in the next 3 months?

You’ll often find that individuals will have vastly different views of what the priorities are, and the fact there is a discrepancy here is a great case for simplifying goals using the OKR system. If you want a great result to this question, have everyone write down what they think the main priority is on a piece of paper, and have them all turn it over together. Once they’ve done that, have them talk about why they think that’s the number one priority.

If you are very lucky, everyone will say the same thing. And if that’s the case, you’re in better shape than you think.

Ask them what has caused under-performance in the past?

This gets teams talking about their real problems. Often teams aren’t given an opportunity to sit together and talk about what's really slowing them down. You’ll find just by being listened to, they’ll be more open to solutions moving forward. And really listen too. Don’t just ask the question and move on. Ask why. Go deeper. And really have them reflect on the problems and think about why solutions were never implemented.

Ask them how often performance is monitored?

One thing I’ve seen frequently is goals are set, and the next time they are mentioned is 3 months down the line when teams are reviewing performance. By then, anything that’s gone bad is too late to fix. The process becomes mostly ceremonial.

If you’re in this situation, you’ll find people see goal-setting mostly as a waste of time. Goals are set, not met, no one cares, and you go through a bit of a song and a dance at the end of the quarter, but mostly nobody is held accountable, and you just keep repeating this cycle to infinity.

It’s no wonder most teams don’t really respond to new objectives being handed down from above.

Finally, ask everyone to think of 1 objective they think is the most important in the next 30 days.

Get everyone's input on what they think the most important objective should really be in the next 30 days. This is in contrast to what they thought it currently was. If they’re unsure how to define important, you can use the following frame of reference. It should be something that serves the customer, gets you closer to your mission, increases profit (or revenue), and is something staff are enthusiastic about working on.

Write these objectives on the whiteboard, and leave them there for now.

Explain who uses it

The truth is most teams I’ve worked with, you can spend a lot of time talking about the value of OKRs and how they work, but truthfully, what people tend to latch onto is the brands that use them today. They think, well… if those successful people use OKRs, and we use them, we’ll be successful too.

This is quite silly, but in practice, it carries a lot of weight. So show them some companies who are using OKRs today. here is a list if you need help.

Explain the gist of it

Google published all their management training online a few years ago, and there is a 1 hour video on OKRs. It’s a good video but pretty long-winded, but the key take-away really is they work, and they are simple once you get going. But a story I often tell to explain the concept of OKRs is a football analogy they use during a video. I have found this the best explanation of OKRs for people who are still trying to wrap their heads around them.

It goes like this.

If the objective is: Win the Superbowl

The key results might be:

• All players must be able to run a 4-minute mile.
• Sell 50,000 premium season tickets.
• Each field kicker must kick above an 85% average.
• No injuries during the playoffs.

Talk about outcomes vs outputs

One of the biggest blunders teams make is to focus on outputs (doing ….you know, things) vs outcomes (results you want to see).

If you create key results that are outputs, you can possibly end up tracking arbitrary numbers that don’t really get you any closer to your objective. A good example of this would be, (to use the football analogy again)

Make sure everyone is at training at 9 am on Tuesdays.

This might sound like something important, but let’s say everyone is at training at 9 am, but they still kick horribly and are terribly unfit and they lose all their games. They won’t win the Superbowl if they are out of breath and can’t kick straight. Or put another way.

Doing 10 Pushups on every second day = Output, vs Losing 100 kgs = Outcome.

You want to ensure that your teams understand the difference between outcomes and outputs. Make sure everyone gets this because it plans into whats next.

Key results mean numbers

Marissa Mayer from Google / Yahoo is famous for saying that it’s not a key result unless it has a number in it. If you only remember 1 thing about key results, this is it. Make sure you can actually measure the key results at the end of the cycle.

having a key result like “Make more money” is great, but what makes it better is saying “Make $1.2 USD selling Monitors by June 13th”. By ensuring key results have numbers in it, you’ll remove the wiggle room during the scoring process.

Outcomes mean more autonomy

Most teams I’ve worked with want more autonomy. There is also huge volumes of research that says this is one of the three big items that lead to motivation.

People want to decide how they do their job because often they really do know better than people in leadership positions. They are closer to the action and are more creative than people give them credit for.

This is important because when you set an outcome (like all players must be able to run a 4-minute mile) you are leaving the how up to the team.

You are not saying

Train 5 times a week.
Do 10 leg presses and 20 minutes on cardio
Or practice running between 5 pm and 6 pm on Fridays.

These might get you to run a 4-minute mile — but maybe not. If you are updating progress each week on outputs, the numbers might be growing in the right direction but you are really just kidding yourself.

What you’re really trying to communicate here is that through the OKR setting process, you’re making it really clear to teams that they will have the autonomy to decide how you’ll hit those numbers.

It allows them to pivot and change course if some of their outputs are not working mid-way through.

Let’s say the key result is run a 4-minute mile. They might try training every day for 2 hours, doing running drills. But a week into it, when they report on progress, they are still only running a 6-minute mile.

Because they are only tracking the outcome (4-minute mile) they can decide, you know what, I’m going to find out what the research says. They then come across an article in Science Daily that says creating 1 rest day after a 2-hour run increases endurance by 300x. So then, they start training every second day and sure enough, they hit the 4-minute mile in under 2 weeks.

This is exactly the kind of behavior you want to see.

Experiment with OKR setting with the team

We want to get teams into the habits of setting OKRs quickly and walking them through the process. So to start things off, create one key result per objective that’s on the whiteboard. And then have the teams contribute their own.

Because this is your first time around, let everyone come up with as many as they can. Let everyone know these won’t be the real objectives, so don’t worry too much about the ideas, what you’re trying to do is get the team into the process so that when it comes time to set realistic objectives, they know what to do.

Homework time

Now that everyone knows how to come up with OKRs, tell everyone that over the next week, you want them to reflect a lot on their priorities, what has worked in the past, and what your customers really need.

You’re going to come together in a week's time and you’re going to set the real ones. Everyone is going to come to that meeting with their full attention and be ready to commit to one objective, and 5 key results as a team.

One caveat. If you feel you already have those OKRs set on the whiteboard, and from the discussions today you feel you nailed it and everyone is super pumped — by all means, run with what you have. But what we go over next may help you validate and sanity check what came out of the session, so make sure to go over to the next step before you lock anything in.

Let everyone know you’re going to send out some information to help them prepare, and you’ll also send them the objective so they only need to come up with the key results.

Key actions for you.

• Arrange a meeting with your team
• Discuss general performance over the last year.
• What does everyone think the #1 priority is?
• Get the team talking about underperformance and why it’s occurred before.
• Clarify how performance is monitored currently.
• Talk through who uses OKRs today.
• Explain outcomes vs outputs
• Outline how outcomes are better for autonomy
• Run an OKR drill with the team. Define some objectives and key results quickly so everyone gets the hang of it.

Part 3: What’s really possible?

Sanity checking your OKRs

The first time you set OKRs you’ll make mistakes. Make peace with that. You’ll set the wrong goals, you’ll set the wrong key results, the wrong cadence, you won’t check-in enough with the team and generally speaking it will take a few weeks before anyone knows what you’re talking about. But relax, you need to just cut your teeth on them, and I promise you, the second cycle, performance will go through the roof.

But to give you the best chance of success, there are some tips that will help you set good, realistic goals, and that's what we’re going to cover now.

Stretch goals vs being authentic

By far the most famous company that uses OKRs is Google. It is well documented, and talked about at length in the book Measure What Matters, by John Doerr . One fact that is well known from their implementation is they aim for key result scores of between .6 and .7.

I know we haven’t covered scoring yet so if that means nothing to you, relax, it’s fine. We’ll get to it.

Google wants its teams to score about 60% — 70% of its key results. They feel if you are scoring 100%, your objectives were too low. If you score 20%, you’re falling behind.

The overarching truth is there are no hard and fast rules with OKRs. Some people think stretch goals are a great idea, some think (including me) that you should just spell out what you want your teams to do. In the end, it’s up to you. It’s your call.

How to pick great key results

We have already covered the difference between outputs and outcomes. So you know how that works. But to pick really good key results, you want to know a few things. I’ve written on this stuff before — but this is the short version.

What’s normal

Let’s say you’re setting key results for a sales team. And every quarter for the last 5 years, they have sold between $1.5m and 2.2m in revenue. It is safe to assume this is considered normal. This is your performance baseline.

So if you are setting new key results, setting it at $20m, while sounding inspiring and getting everyone really fired up, could make people vastly depressed if (when) they fall short.

As you go through more OKR cycles, I encourage teams to get more aggressive with their targets, but for your first time, try and set the bar slightly above what is considered normal. I think you should most certainly set stretch goals in every case. But don’t set goals arbitrarily high. You don’t want to treat your team like lab-rats in an experiment. I feel as time goes on, I’ve found the most effective key result strategy is this;

Be ambitious, but not ridiculous.

What is possible

The next thing to do is try and find out what is possible. There is lots of information out there about industry-standard benchmarks. I’ll give you an example of how we made some mistakes here.

When we first launched OHNO, one thing we really obsessed over was a key result for finding new customers. We wanted to convert about 30% of users who viewed our sign up form into trials. We were tracking between about 16% and 18% consistently and it seemed really hard to get this higher. It made us feel like we were failing, and the engineering team didn’t feel like the product was attractive enough, which lowered morale.

But then we did some digging and spoke to similar SaaS companies and uncovered that the industry average was more around the 8% mark. It turns out we were crushing it and our feelings about our performance were all in our heads.

It wasn’t our circumstances that were underperforming it was our outlook. Once we had the empirical industry data, it clarified what was considered standard in the industry.

So when you are setting your key results, try and sanity check them against industry benchmarks. If you have a team that doesn’t have a lot of experience and they want to set key results, sometimes showing them what other companies are doing will provide a frame of reference that can be useful. This is often called Anchoring, and can have downsides, so be careful.

If you’re a Saas company, this presentation talks about the results of a large scale survey, which points out the most critical findings. It will highlight average conversion rates, how long sales cycles are, and even what are the average pricing models and how they impact revenue over time. Take a look.

Apples and Oranges

When comparing competitive benchmarks, make sure you are comparing the same kind of thing. Let’s say you are KPMG and your competition is Deloitte. You might be encouraged to think if you know what their revenue is in New Zealand, you can set the same kind of revenue goal. But there may be huge differences between the two organizations.

Are they the same size? Have they been in the market as long? Are the teams at the same level? Make sure you take this into account when looking at industry benchmarks.

Once you have some industry data, gut-check it with others in the team and perhaps even those externally to the business — a coach is great for this. In the end, remember any kind of goal setting is nothing more than guessing. We fool ourselves into thinking humans are clever, and that we can predict the future and see around corners, but this is usually far from the truth.

The point of OKRs is to progress forward, with the most effort possible. As long as teams are getting better and growing, you’re always going to be in great shape.

Metrics that Matter (MTM)

Every company should have what is known as its MTM or Metric that matters. It’s important to have just one, and it should be your north star. It’s the one number you want everyone to obsess about.

When setting OKRs your MTM could be an objective or a key result. But what’s important is that for your first time setting OKRs you 100% want to make sure the MTM is included. For us, our MTM was weekly responded surveys. This is the biggest test of engagement in the platform and we track it (and do our best to manipulate it) each week.

Examples of common MTM’s are things like NPS, Monthly Recurring Revenue, Gross Profit, Monthly Unique Website views or even, total pizzas sold.

Leading vs lagging indicators

Your MTM should ideally be a leading indicator. The difference between lagging and leading can be a little confusing, so let me try to simplify it.

If you track something like revenue, that’s great. But you only get revenue by doing something, right? You get it by selling products or making users happy. So in this case revenue would be a lagging indicator and something like NPS score would be a leading indicator.

Your NPS score tells you (arguably) how your clients feel about your product or service. If that number is going up, it may (lead you) to think revenue will climb later (lagging behind).

You can think of lagging indicators as something that happens after you succeed, and leading indicators are things that happen before you succeed.

It’s also fine to track both if it makes sense for you. But you should only track both if there is a reason the numbers could fork, and go in different directions. For instance, if your NPS was really high, but revenue was declining. This would be a reason to track both.

Other great key results

Some of my favorite numbers to track are things like lifetime value, cost of acquisition, and customer churn rates. These are really important numbers you should be tracking. If you track these as part of your first objective — that’s up to you, but they can often tell a lot about how your business is performing.

Now you’re educated

Now that you have some good ideas from the industry and some good ideas from us, you should have enough information to set objectives and key results with the rest of the team.

So share what you discovered as discussion points with the rest of the team so that they have some ideas for when you catch up and set your objectives next.

Include in this email the objective you want to aim for. You could leave this decision up to the team when you get together, but I find picking the objective for first-timers is often easier coming from the leadership position.

If you don’t do this, you may find that when the team gets together to pick their OKRs, they may all have vastly different views of what is the main objective. I know we said to give teams as much freedom as possible, but sometimes giving everyone the same objectives makes the process much easier for the first cycle.

Key Actions

• Pick your main objective and send to the team
• Send any industry benchmarks around to anchor expectations if teams are unclear of what's normal or realistic.
• Don’t set key results too far above normal for your first cycle.
• Know what your MTM is and include it as one key result.
• When comparing your goals to the competition, make sure you’re in a similar circumstance (Apples vs Oranges)
• Define what success looks like. Do you want 100% key results or 60% — 70%?

Part 4: Setting your first OKR

Set your first real OKR

So when you get your team back together, it’s time to pick your very first OKR. You should be up to speed on how to pick the best OKR now, so I’ll assume you’ve managed to pick something. If you’re still struggling, just shoot me a message on twitter and I’ll try help where I can.

Like I mentioned before, the first cycle will probably be a bit of a mess, so it’s less important what you pick than how you manage the process and making sure everyone gets the hang of things.

You want everyone to finish up the first cycle really excited and pumped up — wanting to prove that they are going to absolutely crush it the next time around.

The right number

There is no right or wrong answer about how many key results to have. Having said that, 5–7 is pretty good for most teams. If you have too few, when it comes to scoring, the averages won’t be very meaningful.

For instance, if you only have 2 key results, you can get 0 for one result, and 100% for the other, and you’ll be at 50%. Which isn’t very reflective of performance. So try and come up with a few extras. Don’t force it if it doesn’t make sense. I know some teams that have 1 objective with only 1 key result, and for them, it’s totally fine.

How long should you go?

No matter what OKRs you set, you are going to need to decide how long the time frame is. How long your OKRs are for (the cycle length) is really up to you, I’ve written a guide about why we use 6 week cycles before — but know that there is no wrong answer.

The most common approach is to have quarterly objectives and yearly objectives. I think yearly objectives can be good to set the high-level vision of where you want to be in a year, but my experience has shown me that most teams simply experience too much disruption and variation over 12 months to make those kinds of commitments authentically, and still be relevant a year later. I know there are lots of people who find this comment jarring, but I find the world is moving at such a rate, the shorter the cycles you can adopt, the more agile and adaptive your team will be.

If you chart what tends to happen with teams who set 12-month OKRs is that their relevance decreases the longer the OKR period is. But know this — you do need a vision for the long term. You want to know broadly that in 12 months' time, you want to be in a certain kind of position. But whether OKRs are the right mechanism for this I’m less convinced.

We have found a magic sweet spot for us, 6 weeks. When we set OKRs for 6 weeks, we found its just the right amount of time to pick a target, aim for it, and then review, ready to start all over again. This time period may be too short for most teams, but other companies we have spoken with, Fullstory and Basecamp being good examples, also tend to stick to these 6-week planning cadences. But like I said, find what works for you and don’t worry about what others are doing.

For the purposes of setting your first OKR, I would recommend either 30 days or 3 months. Nothing more, nothing less. 30 days might sound crazy, but as I mentioned, you are going to make lots of mistakes, so keeping it short to cut your teeth on the process might be a good trial period for you to experiment with. Just be aware that if you set 30-day OKRs, you will need to make the key results much more realistic and attainable.

Rubric

A rubric is a set of rules. This is probably something that nearly everyone makes a mistake with on their first turn. Let’s say you set a key result of
“50 Daily Active users”. When you come to scoring, you will need to know what you really mean by that.

Is it 50 daily users, every day of the quarter, even on day one?

Or are you just hoping on the last day of the quarter, you reach 50 daily users?

What do you do if you get 50 daily users on one day, but then it drops to 2 for the rest of the quarter?

You want to know if the scoring will be totals, uniques, cumulative numbers, whatever. It’s key to write this down and be very clear about it when you set the goal.

Individual OKRs?

Some people also have individual OKRs. These can be really great for tracking professional development goals like “Start doing my MBA” or “Give 3 public talks at University”.

Individual OKRs are really powerful for developing staff capability, and I would encourage everyone to have at least one individual OKR — but not for your first cycle. Keep this for when you’ve done at least one cycle.

Everything right now needs to be about a singular focus.

If you chase two rabbits, you will catch neither one — Russian Proverb.

What to do with your first OKR?

You want the OKR to be very visible. The ideal state is that people should be able to recite the key results from memory. It may be good to find a way to prompt users about one of your objectives so it keeps your goals front of mind for most staff. You could do this through pulse surveys, some kind of check-in tool, or even just a regular email or during a standup.

For most teams, printing it out and putting it on a wall is great. There are some obvious downsides to having a paper-only version of the OKR though, and when you have remote teams you want to ensure you have a digital copy.

There are actually lots of OKR tools out there but the tool is actually not that important.

Most teams I know who use OKRs start by tracking them in a simple spreadsheet, and this works fine and scales pretty well.

Where things get a bit more complicated is when more and more teams start using them, and you want to compare scores. As soon as you move to two teams, you’re going to want something that keeps track of the OKRs in a digital fashion.

Share it with the executive responsible

I would assume that by now you’ve kept the powers-that-be across your plan to roll out OKRs, but don’t underestimate this step. I find this is critical.

By making a verbal commitment to the senior leadership in the organisation, what you’re doing is making this less of a game, and more of a real commitment. Drive home the point to the team that you’ve shared the agreed-upon OKR with the company, and that they expect you to deliver it now.

Tell the executive (or whoever is worth telling) that this is what your team will be working on over the next 30 days (or 3 months) and that you’ll give them a progress update roughly once a week on how things are going.

Key Actions for you:

• Make sure everyone knows the OKR by heart
• Keep it to 1 objective and 5–7 key results
• Don’t use personal OKRs for your first cycle
• Pick a time cycle that works for you, but 30 days — 90 days is best to start.
• Store the OKRs in a spreadsheet or some OKR tool
• Tell your boss what the team is committing to
• Plan to update your boss roughly once a week on key results progress.

Part 5: Checking in on progress

On the very first day of your OKR cycle, quickly get the team together (or do this during your standup) and communicate a few key things. I can’t stress how important checking in is. The purpose of checking in is 3 things.

Ensure what the key results mean is clear.

Nudge teams back on course if they are struggling &

Create accountability to results, so that teams know the scores really do matter.

Things are different from here

Make sure everyone understands, from now on, things are different. You are no longer just working on the usual things, the usual behaviours, the usual tasks. Moving forward, your one single obsession is to focus on your OKR.

Next — give everyone the authority to say no to things that don’t get them closer to their objectives. If you talk about making OKRs everyone's focus, but staff can’t get free from the day-to-day tasks that are hindering them from focusing on their OKRs, they will very quickly abandon all enthusiasm for the process. I would say in 90% of cases where I have seen OKRs fail, this is the smoking gun.

I once had a team leader from one of my support teams come to me when we first set OKRs and ask me about problems he was facing, especially about conflicting priorities. “What do you think I should work on?”, he would be asked

“Whatever gets you closer to your OKR” I said.

OKRs are a license to say no. No to things that don’t lead to progress on what you’ve collectively decided will be important.

There is no point in rolling out OKRs if it’s really a way for teams to just add new priorities onto their existing ones. It must replace all current priorities.

Updates

As circumstance would have it, I have always implemented OKRs in teams that are both a combination of remote and on-site people. In my experiences, you don’t need to spend heaps of time discussing weekly updates (10 mins at most), but it’s important to have someone simply update the key result numbers each week ahead of any discussions.

When numbers are trending in the right direction — great. Celebrations all round. On a practical level, you really only need to focus on low numbers.

What I have found is that the first few weeks people will still be unsure how scores are calculated a bit, so the updates tend to be a lot about re-clarifying how scoring works and helping provide guidance on strategies to move those metrics that are important.

Some OKR tools make this easy because every time someone updates an objective or key result, it can send a message to our slack channel showing who updated what, what the update is, and the % increase or decrease this week.

People from the team can then jump in and chat about it right inside Slack, which creates more dialogue and often, encouragement.

When people are ‘conceptually’ stuck

In my experience, when someone has not had a lot of autonomy before, they struggle to come up with activities (or tasks) that will get them closer to their OKRs.

Some people will need some guidance, so feel free to provide suggestions about actions that might help move the numbers if you think it’s appropriate.

Let me give you an example.

Let’s say your key result is;

1,000 monthly web views.

Someone may not have heaps of experience in coming up with actual activities that help them achieve that. So they freak out a little, and the worst possible scenario is that they stay quiet, or just pick anything to do and try and look busy.

If you see someone in this spot you might want to suggest some ideas they could try. Don’t be too prescriptive as you don’t want them relying on your ideas, and you don’t want to give them excuses if they don’t hit their key result. If that happens they may instinctively pass the blame to you, vocally or not, because it was your idea in the first place. They need to own the how. So give people enough room.

But if they are stuck, don’t be a douchebag and let them fail. Give them the right kind of nudge to get them on track again.

When people are ‘operationally’ stuck

The truth is if you’re not actively seeking out barriers to progress, you’re fucked. I have seen so many teams set goals, then get to the end of the process, and only then, do they talk with such clarity and articulation, about every problem that stopped them achieving their results. Only then it’s too late.

If this is happening to you, I have a solution.

Each week, ask every team member to highlight one problem (barrier) that is stopping them from meeting their objectives. Everyone should be able to think of 1.

Prioritise the blockers

To save time, you can prioritise the issues on a matrix, which might help you work out which ones are the most critical to address. You won’t be able to deal with all problems slowing you down, but even solving just one is better than most teams do.

Have the teams bring this list of problems (prioritised, ideally) to a meeting (maybe your weekly meetings or however you are checking in with them currently) so you can talk about it.

The reason most people don’t proactively offer these problems up is that there is no trigger to do so. Most people know what the issues are. They can recite the problems, but very rarely do people communicate them.

Diagnosing the problems

Take the top 5 problems you have identified with the team, and diagnose what’s really going on. Getting to the root cause of these problems is key for long term productivity and velocity. But the best reason to go deep on these issues has nothing to do with goals, and everything to do with a concept known as psychological safety.

When teams are afraid to raise problems, they tend to stay quiet. This is bad for everyone. You want to foster a culture where there are more candid discussions in the meetings where key decisions are made, instead of in the hallways.

To go faster, you need to ensure everyone feels great raising problems, talking about why they occurred, and coming up with actions that solve those issues for good. No egos, no agendas, no vendettas. Just be real, and get it done. If you can pull this off, your team will trust one another like never before and move at a pace that will put your competition to shame.

It doesn’t matter how you discuss these issues, but to keep things fast and simple, I suggest using a process known as the 5 Whys.

The way it works is fairly easy to wrap your head around. Bring up a problem then ask why did that happen? The answer is usually another problem. Ask again, why did that happen. Do this 5 times, and you’ll uncover 5 additional problems.

Most teams struggle to get to all 5 but go as far as you can.

There are some downsides to using the 5 Whys, and it’s not a perfect approach, but it will uncover things that are currently going unaddressed, so when all is said and done it’s still a great way of getting to the heart of whats going on.

Assign resolutions immediately

Once you’ve diagnosed the problem adequately, create some basic actions to solve this problem for good. This is one of the reasons why I say you need to prioritise. You can’t fix every problem, and when you go through the 5 whys process, you won’t be able to execute every action.

But if you can do something it’s better than doing nothing. Take whatever actions have surfaced, and put them inside whatever action/task management tool you use.

If you’re working in Sprints, its good to never add items into the sprint while it’s in progress, so what I like to do is push the top 3 (if practical) actions into the following sprint. This way, it feels like it’s been booked. People on the team will feel if you do this regularly, that their immediate operating environment is improving. And this is great for morale.

Dealing with high performing staff

Another standard thing to hear when talking about OKRs is that they should never be used for performance management. What I mean by that is you don’t want to reward or punish people for their scores.

In my experience, while you may not want to associate key result scores with who is a good performer, I have found that the best performers will typically score well, and the lower performers will often struggle. (Who would have thought…)

If you have a high performing team that is scoring low or struggling to move the numbers in the right direction, I have discovered they tend to become more frustrated than other staff members. I think this is because they are used to succeeding. So make sure to spend time with your high performers and help them get back on track.

Whether you use OKRs to pay bonuses or reward staff is up to you, my advice is not to do this, but this opinion has more to do with the science of motivation than it does to do with the best way to roll out OKRs.

Sharing scores is key

Make sure to send a weekly update to everyone on the team (if they are remote) to ensure that they know that every Friday, or whenever you get together, the scores will be shared.

You want an element of peer influence involved here. You want teams to see how others are progressing, so they can see how they fit into the broader landscape of high-performing teams. It will generally get people pointing in the right direction.

Also, copy in your boss who you said you would keep up to date on the scores. It will also make the process more serious.

Key Actions for you:

• Tell everyone things are different now. Draw a line in the sand about focus.
• Give them the power to say no to things that don’t move their OKR scores
• Have someone update the scores once a week.
• Identify barriers & blockers weekly
• Prioritise blockers
• Discuss the top 5 blockers and go deep on them
• Come up with actions for each blocker
• Keep your boss and your team across the scores
• Watch high performers closely.
• Tell your boss what the team is committing to
• Plan to update your boss roughly once a week on key results progress.

Part 6: Scores

Like I mentioned before, your first cycle is likely to be a bit dysfunctional, so don’t worry about the impact of the scores just yet. Instead, focus on the process and routine of scoring.

Getting a final count

You will need to count up all the key result scores and tally them up in your spreadsheet, or on your bit of paper, or inside whatever OKR tool you use.

In a spreadsheet, your scorecard will look something like this.

Given we haven’t spoken about scores before, I’ll give you a very basic overview of how OKR scores are calculated. In essence, you score them out of a total of 1. But if you want, you can think of it like a % to complete.

If you need to bake 10 cakes, and you only bake 3, then you would be a score of .3 (or 30%).

True or False scores

In your rubrics, (where you explained how the scores would be calculated), you should have spelled out the specifics of the scoring. But one thing that comes up a bit is about binary scores.

Some key results you will want to score as a false/positive outcome. Let’s use the following 2 examples.

KR#3 — Make the Baseball team.
KR#4 — Build our first rocket.

For these kind of key results, you can’t really half do them. You can’t build half a rocket and say you did 50%. And you can’t half make the baseball team.

In these cases, you are going to be scoring them either a 0, or a 1. When you have only true/false key results, you end up with very dramatic averages. You can have teams score a total of 0, which can be quite depressing, even though they might have made great progress.

Flexible scores

Flexible key result scores are a bit different. This is where you might have a revenue target of $1m USD in the quarter. If you sell $650k, you might score a key result of .65.

But, you also might score it 0.

It depends on if you want to count revenue as a true or false key result. I’ve known teams who do both. And I’ve known teams who give binary scores to only high performing teams.

There is really no method to the logic of which key results you will score flexibly, and which ones you will be more binary about, but what is critical is that you are clear about how they will be scored upfront. You don’t want these conversations happening at the end. People will feel vastly betrayed that the rules of the game were not clear when they started playing.

Totals

Once you have put all your key result scores in, you take the average, and that becomes your overall score of that objective. Given you only have 1 objective, this is pretty simple, but over time, if you start creating more objectives, this will become more elaborate.

Transparent scores

Another popular philosophy with OKRs is that all scores are transparent. There are no secret goals in OKR land. This is one thing I do think is important to adopt. It removes the stigma that underperformance is a personal reflection, more than an objective reality. It’s important to make sure the scores are visible for other teams.

We can see how teams compared against each other. Which is sorted in the order of performance.

And we can see the specifics of every key result if you want to give in and talk specifics.

Talking through

It’s important to get everyone together to review the scores. This is probably one of the most important things you can do as part of the process. You can make a really big deal about it too. This is Nicole Brolan, Product Director from Seek, at their regular OKR day where they run through all the numbers.

You want to cover a range of things during this catch up with everyone, but this activity (It’s a ceremony really) is all about celebrating success, discussing barriers, and working out how to improve for next time.

Here are some things you want to cover during this session.

Winning

Take stock of people or teams who absolutely crushed the key results. Also in the back of your mind, ensure that next time around, you push any high performing teams that little bit further.

Low scores

With teams that fell short of their scores (I would usually say anything below about 60% is worth discussing), It’s good to have the team talk through where they think things were difficult.

Find out what they would do differently, or how they could have adjusted their activities to get closer knowing what they know now. This isn’t about putting their failure on display, but more giving them an opportunity to reflect and learn, so that next time around performance should see a rise.

This is also a great opportunity to increase psychological safety in teams. If everyone realizes that disclosing defeat doesn’t mean they are punished, they will be more likely to open up about mistakes, which is great for everyone.

It’s worth noting that most people think you learn from mistakes — I have found this to be mostly false. The evidence says people rarely learn from mistakes they’ve made, but instead, people learn from pain. When things hurt, you learn really fast. You want to make sure that presenting on poor results isn’t about humiliation, but it should be slightly uncomfortable too. You don’t want total apathy and indifference for underperformance. Missing your key results isn’t cause for celebration, but it’s also nothing to be afraid of.

So don’t be mean, but also, don’t hide from failure. It helps nobody.

Vanity scores

By now you’ll also be able to work out which metrics you were tracking, and perhaps these are even ones you succeeded on, which turned out to be vanity metrics. They looked good, but they probably didn’t actually get you closer to your goals.

Go deep on these. Ask what would have been better key results to track. You want to reduce as many vanity metrics as you can in the process, so it’s good for teams to not make the same mistakes a second time around.

Closing out

Now you have what you need, it’s time to close out the objectives. Inside your OKR tool or spreadsheet, it’s time to archive the particular objective. In the below example, we just click on the objective and select archive. It retains all the information but allows us to report on it later. Most tools will allow you to do this.

Setting up your second cycle

During the session with your team, you want to find out which objectives and key results will be needed for the next cycle. Some teams take another week and turn this into a big thing, but in my experience, if you can get both the scoring and the setting is done on the same day, you are in world-class shape.

Now that you’ve done your first cycle, you need to decided on the following things.

How many objectives do you need?

For the next time around, you probably want to create at least 2 objectives for a team. Don’t go crazy and make 5 or something, but adding one more objective is probably feasible now that you have the hang of it.

How many key results are comfortable?

You should also get a sense now of how many key results are practical for each team, and each objective. I know I said 5–7 is good, but depending on the type of work and the team you have, the real number should feel right. Go with your gut, and ask the team what they think is possible. Don’t have so many you can’t remember them by heart.

Do you need to split up people into different teams for this cycle?

It may actually become apparent that in the last cycle there were too many dependencies on other teams to hit key results. You want to ideally break teams into the smallest groups you can (The 2 pizza rule), and set them up in a way where they don’t need to rely so heavily on others to hit their goals.

In practice, this is nearly always impossible, but the more you can reduce dependencies, the better. So if you think re-jigging the team structures will get you better outcomes, this is the time to try it.

I recognise this is impractical in most teams too. You can’t just go forming and dismantling team structures without a lot of input from the company as a whole. So I understand the limitations here. But try as best you can, to make the teams small, cross-functional, and autonomous.

What learnings do you have that can be incorporated into the next cycle?

Do a deep retrospective of the entire time period. Find out what everyone thinks could be improved. Try and adjust your check-ins or structure to make things easier for everyone.

Do any key results or objectives need to be carried into the next cycle?

Sometimes unfinished OKRs should be put into the next cycle. This happens quite a bit really, so I would assume at least some of your key results will be put into cycle 2.

Are there any ‘squads’ that are required?

When I refer to squads, I’m talking of temporary teams that attack an objective together, outside of their normal team structure. It might be a small band of people that meet once a week to get a specific project done. A good example of this might be forming a team to build a new website. It might only take a week to do, but it could be worth tracking.

Can you shrink the cycle?

Finally, ask yourself if you are using 3-month timeframes if you can shrink down the cycle a bit. Like I mentioned before, I think there is real value to trying to get your time periods down to about 6 weeks.

Key Actions for you:

• Publish all scores transparently with the company.
• Setup Squads if needed for temporary projects.
• Punt relevant incomplete key results to your second cycle.
• Capture and action any learnings for next time.
• Evaluate team structures to make sure they are optimised for performance.
• Add at least 1 additional objective for next cycle.
• Discuss vanity scores.
• Talk through lowest scores and why they occurred.
• Celebrate success.

Part 7: Going company-wide

If you like the OKR process so far, you might be thinking about rolling it out company-wide? Everyone thinks this is fun, but don’t underestimate the challenges. Here I lay out some common things you’ll come across and how to overcome them.

Demonstrated performance increases

If you want to show the value of OKRs to the broader organisation, you will need to be able to show it has more value over what you currently do. This is why highlighting the process to the business leaders during your first cycle is key. You want to take them on the journey with you, and at the end of the second cycle, they should start to see that OKRs really are a simple way of getting everyone rowing in the same direction.

What’s the difference to KPIs?

In my experience, there are three kinds of organisations out there. Those who just want to use OKRs because that’s what everyone does now, and they like it, for whatever reason.

The second group is mostly larger corporates who will tend to use fairly old-school KPI’s, or have a kind of ingrained performance management process which would be too complicated to change at scale.

There is a third type, my favourite group really, and this is where there is no goal setting at all. With this group, moving to OKRs is great. It’s simple to wrap your head around and easy to setup.

For the corporate KPI adopters, KPI’s will be used for roles and goal setting, and depending on how ingrained this is, if you want to move to OKRs, you can expect to spend a huge deal of time just explaining the differences between KPI’s and OKRs and even then — most will still not get it.

Here is the best way to explain OKRs to someone who is only familiar with KPI’s.

Tell them “Comparing KPI’s and OKRs is a little like comparing Oranges and Fruit. One encapsulates the other.”

While this is not exactly true, it seems to communicate the broad strokes enough for people to get it. Tell them the KPI bits are the key results, and the objectives are more of a broader goal. If you use this analogy, I find it makes this process less painful and you avoid long meetings where people talk about OKRs in circles, trying to agree on definitions and terms.

Another thing to mention is OKRs put the focus back on key results. Most KPI’s I have seen are long lists of a mixture of behaviours, tasks, projects, and activities, usually a few pages long, which nobody can remember and only get talked about during performance reviews. The key in KPI’s is missing.

Something is only key if it’s important. But if everything makes the list, then nothing is really key anymore.

The scores of long KPI lists never really reflect actual performance. If you get someone who is passionate about staying with their KPI’s, don’t force the issue, but if you want to test their resolve, see if they can even recite them by heart. My gut says they’ll hardly remember 3 of them.

How do we pay people bonuses?

This would have to be question #1 when it comes to OKRs. Most performance management systems exist so they know who gets the cash prizes at the end of the quarter, and who to fire.

If this comes up, you can try 2 things. If you’re brave, you can show them some of the current research about how paying for performance actually leads to underperforming teams. Yes, you heard me right. If you pay people a bonus for performing, they will usually underperform. In my experience, going to war on this topic is very hard unless you have a very open-minded leadership team.

The second thing that paying bonuses based on performance leads to is an individual view of the world. OKRs ideally should be about teams, and if a team succeeds, you need to ask yourself, does everyone on the team get a bonus? This leads to socialism vs capitalism debates, which are not overly helpful.

How you pay bonuses is a complicated topic, but if you are used to paying bonuses for those who hit their KPI’s, my answer is usually the one Google recites.

“OKRs are not to be used for measuring performance and paying bonuses.”

So if you want to pay bonuses, you’ll need to find another way.

There are actually lots of ways to pay bonuses that don’t relate to OKR scores. You can use a separate assessment altogether, one that incorporates values, how much they are exhibited, individual contributions to the team, overall company profitability — anything really. But just know that when you start to pick apart this topic, it will rain on your OKR parade.

So if you want to roll out OKRs more broadly, and you are in an environment that pays for performance, my advice is to just split up those two concepts and talk to your HR team about it.

just say that you think OKRs represent a better way to drive teams forward, and for now, you don’t want to tangle that with those who get paid bonuses.

Keep in mind, this is probably the real sticking point for why most teams don’t fully commit to rolling out OKRs more broadly. When people don’t understand how OKRs fit into the performance benchmarking / bonus story — the company-wide implementations will fail. You really want your HR team onside with this goal-setting framework. They will help you navigate it better than anyone and can coach you on how to roll it out more broadly.

What makes it catch on?

The best reason OKRs get adopted more widely is that teams who use them typically have a greater sense of alignment, and that will lead to better results. When you see teams around you crushing their goals, it’s hard to ignore. If you can demonstrate to other departments and teams that things are going well — they will want some of that too.

Why you shouldn’t use OKRs.

But you must be open to the idea that OKRs aren’t for you. Personally, I don’t think they are for everyone. I think some organisations have very particular appetites and variations on OKRs and honestly, it’s not overly important in the grand scheme of things.

I think too much is written about the value of OKRs and not nearly enough on the value of just setting any goals. When I have met with new customers, I am really surprised by just how many don’t have stated and clear goals. I think if you’re a team who has no goals at all — simply setting something, using any framework, will be of great benefit.

But if you felt OKRs didn’t work the first cycle, as I’ve mentioned before, this is pretty normal. I think you want to try at least 3 cycles before you call it quits. I have found by cycle 3, things are usually really great. So stick with it for at least 3 cycles.

But in the end, I think if you feel OKRs are not right for you — don’t feel bad about it. Just find what works for your team, and have the courage and fortitude to know when something is not working, and pivot to something else.

What matters is not the framework, the tools, or the process, but that your team is engaged, reaching their goals, and getting better every day. If you can come up with something else that does that — I’m your number 1 fan.

Key Actions for you:

• Stick with it for 3 cycles before you call it a day.
• High-performance results will sell it the best.
• Don’t mix bonuses and OKRs.
• Make the concept simple for those that love KPIs.

If you finished this article, give yourself a big congratulations! It was a long read, but kudos. When I first wrote this, I was surprised by how many people sent me messages saying it was helpful for them, and it was those comments that inspired me to move it off the OHNO website and paste it here on Medium so others might discover it. Hopefully it was helpful — it is everything I have learned about OKRs during my research and interactions with other organisations.

This guide first appeared on http://ohno.ai. If you have any questions about OKRs or comments about this article, let me know.