TL;DR: A startup can eventually pivot to a better market or a different product. It can get funded, can hire better people. However, without a healthy culture, everything will crumble. The healthiest culture is a culture built on accountability and operational excellence, not on fancy values. One way to build accountability and performance from the start, and define the right culture, is to use Objectives and Key Results (OKRs). OKRs are a management performance framework, first created by Andy Grove at Intel and used today by Google, Twitter, and thousands of high-growth startups. This article offers a practical way of implementing OKRs, based on my 20+ years of experience as a founder, and on my many mistakes.

[wd_hustle id=”6″ type=”embedded”/]

Contents

  • Everything is about Phase Two
  • Objectives and Key Results (OKRs)
  • Yearly goals to quarterly OKRs
  • Quarterly OKRs to weekly priorities
  • Weekly priorities to daily tasks
  • Routines
  • Tools
  • Lessons I’ve learned

Everything is about Phase Two!

Building a startup is much more difficult than building a traditional business. Any tech founder knows it. The tough part is that, in most cases, the startup not only faces uncertainty, but insufficient resources, entrenched competition, lack of experience, and many other challenges. Still, there is hope! Uncertainty can be fought through experimentation, resources obtained, go to market strategies devised, more experienced people hired. But, without operational excellence, the opportunity will be wasted.

I’m going to stop here, to share with you one of the best business lessons I have learned in the past ten years. It does not come from a serial entrepreneur who raised billions and billions of dollars, nor from a thought leader, or a known VC. It comes from a bunch of third-graders in the popular South Park TV series. Before you watch this short excerpt, let me give you a bit of context. The kids have to write an essay about business, so they gather at one of their homes, drink coffee, then end up staying late thinking about what they could write. While debating how they should approach the topic, a few gnomes come in the room. the gnomes open drawers and steal the kid’s underpants. Watch this for the rest of the story:

 

So what’s the lesson here?

Objectives and Key Results (OKRs)

What is your phase 2? How do you get from an idea to a successful, profitable business? Despite all the learnings, entrepreneurship and management are still more art than science, and subject to different mental models, cognitive and personal biases.

There are problems with all traditional management models—waterfall, agile, management by objectives, six sigma, lean startup—heck, you could even consider micromanagement an approach. Especially if it can help you break through the early stages. The downside is that the waterfall model can’t plan for uncertainty. Agile management is too short-sighted. Six-sigma will help you improve your startup process, but what if it’s the wrong process? Lean Startup is an excellent philosophy, but it encourages cognitive biases and a lack of accountability as a methodology.

All the traditional models have shortcomings when applied to a small startup organization, that’s why Objectives and Key Results could be a better fit for startups. Based on the management by objectives methodology, OKRs were first adapted by Andy Grove at Intel, then brought by John Doerr at Google, from where they spread throughout the startup world. Christina Wodtke, a former product manager and GM at LinkedIn, MySpace, and Zynga, further evangelized the methodology through her articles and books. (I recommend both Radical Focus and The Team that Managed Itself).

While there are as many detractors as there are supporters, I won’t get into that argument. I have implemented this and adjusted it over the past seven years in all my companies (Grapefruit, Eternime, Exploratorium), and with tens of startups at MIT, Singularity University, Techstars, and other accelerator programs. It just works. Here’s how.

An OKRs implementation guide

Now it’s time to get your hands dirty. This is exactly how we do it at Exploratorium (the startup organization that builds DE Toolbox and Metabeta). It is one of our essential processes as it helps us move together as a team from the vision to why it finally works. It helps us move from a long-term vision beyond ten years to the hard question of what we should do today, then it allows us to be accountable.

Our process has two essential parts:

  • The model we use to zoom in from a 10-year visionary horizon → to a medium-term (3-year) horizon → to yearly goals → to quarterly OKRs → to weekly priorities and, finally, → to daily tasks.
  • The routines that keep this process on track.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.

From vision to yearly goals

Our long term vision is simple and expressed on one page:

We help investors to find and grow the best startups. We do this by focusing on data-driven decisions, not only on gut feeling.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.Let’s descend from the high clouds of visionary thinking and get down to numbers. How does this look? A clear finish line supports our vision—our exit strategy. What’s yours? Forget the IPO (less than 0.2% of surviving startups go public) and be realistic. I have to admit that I did not have an answer the first time I was asked this question by an investor, back in 2005. Thank you, Konstantinos, for this very important lesson.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.

Ten years is enough to reach most of the ambitious goals. How do ten years from now look for you and what are your goals? What does a closer success look like, in three years? Now is the time to get more specific. There is still a lot of uncertainty but the process is about breaking a tough task (find the pirate treasure) into smaller, more manageable stages (sail to the Caribbean and find the map, hire locals for help, and stack provisions).

Don’t get too specific—these milestones should be achievable. One important thing about this breakdown is that it helps everyone on the team understand our purpose. Where our ship is sailing, why, and what’s on the map.

From yearly goals to quarterly OKRs

We used to define OKRs at a yearly level. Mature businesses do plan in yearly cycles and objectives, but we still have a lot of things to learn, sod we realized that being very specific gives us a false sense of certainty. So we keep the yearly objectives still very broad. This changes when we get at a quarterly level. It is a very close horizon that allows us to be specific. Very specific.

We set one Objective (which has to be inspirational, not SMART). This is our North Star for the quarter and everyone on the team can rally behind it. The next question, “How do we know we’re there?” gets us our Key Results—three measurable outcomes of the quarter. The rule here is this: one Objective and three Key Results. No more no less (in the wise words of Master Yoda). There is a rationale at the end of this article if you have the patience to continue.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.

There are a lot of ways to formulate the Key Results (this website has some good examples for all business areas). Here are my suggestions:

  • Key results should be expressed as results, not tasks.
  • They should be measurable (i.e. countable or with a specific deadline).
  • Their progress should be measurable (i.e. it’s not an on/off thing).
  • They should be stretched. You can easily evaluate this by measuring your confidence level for attaining them. It should be 50% at the start of the quarter. Reaching 80% of the Key Results is considered a good outcome.
  • Key Results should cover the three most important areas of your business at the current stage (in our case product, sales, and marketing for the current quarter).
  • They should be very clear, concise, and simple.
  • Don’t cram two Key Results in one (eg. 1,000 new customers and 10 advertising campaigns). Choose the most important one. Stick to the rule of three.

Implementing OKRs is the job of the founders. It’s not easy to decide what’s important now and leave aside all the things you want to do. Still, the path to failure features a lack of focus.

These were the company quarterly OKRs.

[wd_hustle id=”6″ type=”embedded”/]

Individual quarterly OKRs

If your team is big enough (16+) you may choose to have departmental OKRs derived from the company OKRs and only then move to individual ones. Otherwise, for the sake of simplicity and against bureaucracy, go to the individual level.

The same general rules apply to your individual OKRs as for your company OKRs. However, there are some specific aspects of individual OKRs that I’ve learned the hard way. Hopefully, you won’t have to. Here they are:

  • They should be decided by the individual. That’s where accountability comes from. You can’t have it without delegating decisions. The only thing you can ask is “How does this support reaching the company’s OKRs?”. I can’t stress this enough. It is perhaps the most important rule of implementing OKRs.
  • The OKRs should be related to the team member’s everyday job. Extracurricular things don’t work (like studying something, getting certifications, non-essential projects). I often get the question “What should the person do first, OKR-related stuff, or daily-job-related stuff?”. If you’re asking this thing, OKRs are not set the right way. Don’t set your team up for failure. Focus on implementing OKRs right.
  • They should be decided in a common meeting, where everyone should be confident that by putting together their individual OKRs, the company OKRs are achievable. Let your team discuss and explain to you how the team as a whole will get there.

Quarterly OKRs to weekly priorities

Let’s assume we decided on the following. By the end of this quarter, we’ll set sail to the sky-blue seas of the Caribbean (that’s the Objective). As Key Results, we need a 15-ton sailing ship ready to leave, a team of 20, and have raised 15,000 doubloons (to be honest I have no clue of the exchange rates to date). My Objective as a captain would be to get the expedition funded. My Key Results would be Key Result 1: at least 10 ships reviews before securing one; Key Result 2: two senior pirates hired by mid-quarter; Key Result 3: 50 fundraising meetings to secure the money.

The question is, “what do I do this week”?

This week me (and every other pirate), will choose three priorities (or results) to achieve. Not tasks. Unlike key results, which are progressive, these priorities should be binary, meaning that, by the end of the week, they are DONE or NOT DONE—“There is no try”). These priorities should help people make one more step towards reaching their quarterly results.

Objectives and Key Results — A practical guide to implementing this performance management framework.

Some guidelines:

  • The tendency here is to express them as tasks. “Contact all sales leads”, “Publish 10 social media posts”. This leads to the false impression that the priority has been (almost) achieved by the end of the week. I insist on changing them to outcomes. “All sales leads contacted”, “10 social media posts published” etc. to encourage accountability.
  • All have to be connected to individual OKRs, otherwise, they might not be a priority.
  • Of course, everyone decides for themselves, but everyone in the team should easily see these priorities.

From weekly prioritization to daily tasks in implementing OKRs

You probably expect a full system, top-down (vision-to-task) process for implementing OKRs, but I will disappoint you on this one. I found out that every human being has an essentially different process and way of being productive. The time of day differs, tools differ, working styles differ and there are a lot of other variables involved. My goal as a CEO is to set the direction, expectations, and to help them reach their expectations.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.

  • For to-do lists, we use whatever we feel comfortable with, Jira, Things, Trello, Todoist, Google Calendar, text files, pen & paper, etc.
  • We only enforce tools for projects that are collaborative. Like product development (where we use Jira) or sales (where we use Salesmate), but, as a small team, we stop there. Everyone is free to use their personal system. In my case, I use a combination of Things (to collect tasks from everywhere in the universe), pen and paper (to reflect), and Google Calendar (to timebox and schedule my tasks for the day).
  • We have a daily stand-up meeting at the end of the day, but that’s covered in the next chapter.

Routines

As a process freak, what I found is that the best processes die without the proper routines. It’s almost too easy to postpone the important tasks and focus on the urgent. When doing this, you wake up in the last week of the quarter and realize you’re only 10% on your way to your OKRs. Because—you know—life happened, clients needed stuff, your team needed stuff, and a million other excuses. So we put in place a system of routines (meetings and personal tasks) to make sure that we keep the main thing the main thing.

Daily routines

We have a 15-min daily review with our whole team, where we report on the status of our weekly priorities. This forces us to be accountable on the priorities, not the other tasks that fill our to-do list throughout the week. Otherwise, it would be impossible to make progress towards what is important, not towards what our email inbox, Slack conversations, or whatever else others put on our list as supposedly important. We usually try to avoid small talk, yet with remote situations, like the one caused by the pandemic, we do a bit of chit-chat after reporting. This is possible and quite pleasant because the meeting is scheduled at the end of the workday.

Weekly routines

We have CEO one-on-one meetings on Mondays and Tuesdays, where everyone gets 30-mins with me to discuss their priorities and OKRs. We talk about what’s needed to make progress on the OKRs; what would be the steps beyond the current week; what are the bottlenecks, needs, and any general feedback? If there is time left, we discuss any pending project issues and needed help. Or we get back earlier to our to-dos.

This takes the place of any formal annual or quarterly employee reviews, as it is a chance for mutual feedback. It also avoids the fakeness, bureaucracy, and inutility of a last century management artifact.

For larger teams, one-on-ones are not feasible with CEOs but should be with the team/department lead.

The other important weekly routine is our end-of-week team review. We do it last thing on Friday, it lasts one hour, and it’s an overview/report of the current week. It has two parts. The first 30 minutes are for reporting, the last 30 minutes are for discussing any pending issues or information and for deciding the next week’s priorities. We use the first 30 minutes to report in the same document on our priorities/areas of responsibility (OKRs, marketing & sales, product, financials, other). Everyone is in the meeting, writing down stuff or reading what others write and adding details or clarifications. This way we avoid hasty reporting, done in the 5 minutes before the meeting.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.

We discuss why the priorities were not done, the lessons learned, and the things we need to change. This is when and where we talk about how to improve the process so that we keep focused and effective. When it’s done, unless we’re in lockdown because of a pandemic, we try to get out for a drink or pizza or we try to do something fun together, to celebrate the end of another week and the steps we took in our pirate treasure hunt.

Quarterly routines

We end the quarter with a quarterly review 1-hour meeting, similar in format with the weekly review. Then, we get in-depth feedback on why some things worked and some did not work. We already have tactical details from each weekly team update, so we are more analytical and reflective in these meetings.

A few days later, the founders decide on the company’s OKRs and we have another 1-hour meeting quarter planning with the whole team. This is where we brainstorm and decide on how to move from company OKRs to individual OKRs.

It might sound like I’m a meeting fanboy, but this means only 15 hours per team member every quarter to make sure that we prioritize things, we are on the same page, and we are accountable. Even meeting-haters in our team think this is worth it.

Tools for OKRs

I couldn’t care less about the tools. They are only a means and I focus on them as little as possible. Some advice:

  • Use a tool that your team is already using. Don’t introduce new tools and accounts to your team for this.
  • The only rule is: everyone should already use it, and everyone should use it.
  • Use specialized OKR/team productivity tools like NestorUp (one of the best tools for managing your team’s cohesion and performance), Gtmhub, Weekdone. You can use generic project management tools like Trello, Asana, and Jira. You can even use spreadsheets, Google Sheets, or documents, like Google Docs, Confluence, and others. Or you can use PostIt notes, a wall, anything that works for you as a team.

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.

  • You can also use Metabeta for weekly/quarterly reviews, to create a template for these updates, and set a schedule for creating them. So each Friday you’ll get a new weekly team update document to report in, every end of the quarter a new quarterly update and so on. Once completed, the updates are emailed to all team members, in case any of them missed the meeting. Create a startup account, play around with this new feature, and let me know what you think.
  • In case it hasn’t been obvious up until now, I’m a process freak. I use pen & paper (I love pens, ink, and sketching) to put down my personal OKRs each quarter and cover non-work related areas as well. I have even been accused of being too obsessed with this—but it’s my strategy of making sure I end up closer to where I want to be, rather than where life takes me. Let me ask you this: would you rather be a paper boat pushed around by any breeze, or build a small motor for the paper boat that will take you closer to your pirate island?

OKRs (Objectives and Key Results) — A practical guide to implementing this performance management framework.Lessons I’ve learned

There are some lessons I have learned over the years and I’d hate for you to learn them the hard way. That being said, I’m going to write them down. You’re going to ignore most of them and figure out later what and how works for you and your company.

  1. The process is more important than the right OKRs
    Few startups have a clear idea of where they are going when they start their journey. The discipline of having such a process in your company will build accountability and operational excellence. You can have the best OKRs and you can be the number one fan of implementing OKRS, but without the process of getting there, they are going to stay a dream. Having a process will ensure that week by week, quarter by quarter, you’re going to be able to set better OKRs.
  2. Don’t set your employees’ OKRs or weekly priorities
    The temptation to do that is high, especially with junior members in your team. Don’t. This is a process of teaching them strategic thinking, having them make their decisions, and supporting them in learning and owning their OKRs and priorities.
  3. Don’t use them for performance reviews
    If you do this, the whole intrinsic motivation will be gone. Carrots and sticks don’t work well in startups. Implementing OKRs is not about that.
  4. They are for everyone
    It’s not a system for your team. If this process does not start from the top, if you’re not following it, you cannot expect anyone to do it. Remember, this is about creating a results-only-work-environment (ROWE), a healthy and accountable culture for your startup.
  5. Three, no more, no less
    I know, you think I’m only repeating Master Yoda’s mantra over and over, but two OKRs will make your organization unbalanced. Too much focus in some areas, too little focus in other areas. Four or more OKRs will lose your focus. Not everything is that important. Simply eliminate the least important and stay with three.
  6. Implementing OKRs is how you make sure everyone is on the same page
    Lack of alignment leads to a lack of motivation, internal conflicts, personal frustration, and, of course, a mess in execution. Keeping the North Star in front of everyone at all times helps to share the vision of your startup journey in detail and it helps everyone understand what their role is in this journey.
  7. Yes, you can change OKRs in the middle of a quarter
    If they become obsolete because of a major shift or event, there’s no point in keeping them. Just make sure the new ones are implemented in the same way for the rest of the quarter.

[wd_hustle id=”6″ type=”embedded”/]

Last, but not least

Wow, that was a long write for me, and a long read for you. If you got this far, I am confident that it means you’re truly interested in this topic. So I am more than happy to share the templates for OKRs, weekly reviews, and quarterly reviews that we currently use in our team. See the details at the end of the article.

Now is the time to post questions about implementing OKRs in the comments or share this article with your team. Do it now, it’s easy. Just use the buttons below. I do hope this works, and I am looking forward to hearing from you when it does, to share your thoughts and lessons!

Good luck!