OKRs: Getting Started in an Early-Stage Company

Introduction

Many articles have been written on the OKR methodology, most of them approaching the matter from a quite theoretical angle. This is great and very much necessary. We, operators, do not need to reinvent the wheel but instead can leverage this excellent framework to align everyone in our organization.

But when we decided to implement the OKR methodology with more attention at Upflow, tapping into this knowledge did not give us a clear understanding of how to translate it operationally in a series A company (growing fast from 20 to ~100 people).

This article is an attempt to share our experience adapting the methodology to our context. We hope it will be a reference point for other early-stage companies willing to get started with OKRs, and wondering how to navigate the first 12-15 months of adoption.

The theory, briefly

Early days

The creation of the OKR methodology is attributed to Andy Grove, former CEO of Intel, back in the 1970s. He taught John Doerr, now a venture capitalist, who wrote a famous book “Measure What Matters”. Since then, many successful companies leveraged OKRs to start and scale their organizations: Google, Microsoft, Amazon… It is now likely that any successful startup we look up to uses OKR.

Benefits of using OKRs

The OKR methodology invites every level of the organization to take a step back, and define goals and measurable results for the foreseeable future.

• OKRs allow all teams to focus and rally behind a small set of carefully chosen priorities.

• It provides a method for the entire organization to align its goals at every layer with its ultimate purpose.

• It creates commitment from the parties involved to choose and stick to agreed-upon priorities.

• It allows teams to track their progress toward a goal and know earlier when to change tactics.

• It empowers teams to set goals that stretch beyond “business as usual”, and make significant, meaningful changes.

What are OKRs

The OKR methodology ensures everyone works on the same important priorities throughout the organization. It invites us to build a cascade of objectives (directions) and key results (milestones) at the company, department, and contributor level.

An Objective describes what is to be achieved, where you want to go. It is qualitative, action-oriented, and inspirational. It can be long-lived and roll over from one period to another. Its benefits are to clarify the direction or outcome you are working towards, it prevents fuzzy thinking and ineffective execution.

A Key Result describes how to get to the objective. It is a milestone, measurable, and time-bound to the period you run OKRs on (often, quarters). It thus evolves every quarter as you progress. Its benefits are to measure your progress to attainment. There is no grey area whether you reach it on not.

Here are two examples

Why and how adopting OKRs at Upflow

Since Upflow started, the team always ran on a simple version of the OKR methodology, driven by our CEO. This definitely helped roll it out further, because (A) the CEO has been a promoter of the methodology and actively sponsored the initiative since the early days, and (B) we already had a small history with OKRs upon which we could build.

We decided to push our OKR implementation further a few months after raising our series A, when I joined as Chief of Staff. As we will discuss in more detail later in this article, this move is easier to perform when you have someone in charge. At that time, we were a team of 30 people, growing fast.

One of our operating principles is to work in small batches with the long term in mind. We strive to work using the 80/20 rule (or Pareto principle) - and have applied that principle to this OKR project too. We felt and still feel that adopting a step-by-step approach is the right way to go at our stage. Why?

Because we are not Intel or Google (yet!). No one at Upflow is that far from the vision/mission (ie. founders) of the company to need a cutting-edge planning and reporting process. Given we are less than 100, our team is mostly made of people with an entrepreneurial mindset, at ease without a super strict and by-the-books organization. And in any case, we do not have the bandwidth and man power to set up a team fully dedicated to animating the creation and tracking of our OKRs. So we started simple, because it’s the best we can do, but most importantly because it is more than enough at our stage. We just needed to unlock the most powerful features of OKRs with minimal investment.

In the following sections, we will present how we navigated our first 12-15 months with the OKR methodology.

Step 1 - We enabled people on the methodology

With common knowledge…

First things first, we started with building some internal documentation (on our knowledge base, Notion) about the OKRs, to ensure everyone had a similar understanding of the key components of the methodology, and how we want to translate it operationally at Upflow.

Here is what’s included in this documentation:

1. Why using OKRs (see above “Benefits of using OKRs”)

2. What are OKRs (see above “What are OKRs”)

3. How to prepare your team/department OKRs

4. How to track progress in our OKR tool

Not only did we try to keep the document as light as possible, we also wanted to make it approachable for everyone, regardless of their favorite way to learn. To do so, we embedded Loom videos at the beginning of each category, where I orally explained the content. This way, people could choose between reading the document or watching a series of four 5-min videos, as a small online & async training.

… and a tool.

You understood it from the 4th category: we immediately got a tool to log OKRs and track progress during the quarter.

We have been using Lattice for a long time now to power our 1:1s, 360 reviews, and engagement surveys - and it also offers an OKR feature. We saw several benefits, among them: not adding another tool fully dedicated to OKRs while we were not yet mature on their implementation, and being able to visualize OKR progress next to everyone’s 1:1 agendas easily. So we gave it a go!

At our stage, we could have also started with a nice Notion dashboard or even a spreadsheet. A good tool (whatever it is) appeared to be a necessary but not sufficient condition for the adoption of OKRs internally. In any case, it is necessary to have a tool to support the process - otherwise, it may fail - but founders’ sponsorship, discipline, and many other factors are also key to making it work. So we’d recommend not over-engineering it too soon: a good spreadsheet or Notion note could do.

Step 2 - We rolled it out at the department level, starting with Business and cross-functional teams

Focusing on the department level…

As suggested in the 3rd section of our documentation (see above, step 1), we decided to start our OKR implementation focusing on the department/team level. Ultimately, and as per the “theory”, OKRs will be rolled out at the company level, the team level, and the individual level. To start, we picked only the level lying in the middle: the department/team.

Being a team of 30 people at this time, everyone could relatively easily connect to our vision and mission, and we thus felt we didn’t need a very detailed set of OKRs at the company level. We also thought that OKRs at the individual level were not the priority: as our teams were quite small, it was easy for our teammates to relate to (and even own) the priorities of their departments.

… and only the Business and cross-functional team.

More than that, we narrowed it down further to only our Business and cross-functional teams.

For context, our structure looked like this:

• Product & Engineering - reporting to our co-founder & CTPO

  • Product (scoping what we need to build)

  • Design (designing what we need to build)

  • Engineering (building what we need to build)

  • Data (ensuring we make data-driven choices)

  • Security (keeping our customers and us safe)

• Business - reporting to our co-founder & CEO

  • Business Development (generating opportunities from outbound)

  • Marketing (generating opportunities from inbound)

  • Alliances / Partnerships (generating opportunities from indirect channels)

  • Sales (closing new business)

  • Success (ensuring retention and generating expansion)

  • Business Operations (helping the above scale)

• Cross-functional - reporting to our co-founder & CEO

  • Finance

  • Talent & People

  • Founders Office

Supported by 10 recommendations…

We were and are convinced that rolling out OKRs should help our teams, not hold them back or slow them down. We shouldn’t implement a new process for the sake of it, but ensure the benefits are net positive for the organization.

We thus came up with those 10 recommendations (logged in the 3rd section of our documentation, see above, step 1) to help the team leads write their team’s OKRs.

1 — Your team’s OKRs is not a list of team members’ OKRs

Your team’s OKRs should be an independent set of priorities. You are free to work on personal OKRs with each of your team members — even though Upflow does not ask you to. But your team’s OKRs should be another set of OKRs encompassing your team’s efforts, not a list of all your team members’ OKRs.

2 — Before you start, look to the company strategy

Your team’s OKRs should be grounded in the company’s current strategy. So before you start drafting them, make sure you’re aligned with your manager or the founders on the company direction for the upcoming quarters. One of the values of OKRs is to provide alignment — it starts here.

3 — Keep it to maximum 3 Objectives

One value of the OKR methodology is to drive focus around a set of carefully chosen Objectives. A goal-setting exercise is a lot about prioritization. You should expect to experience this feeling of having to say no for now, to deprioritize — it’s normal, healthy, and actually what we are looking for here. We encourage you to write down what is not a priority for the upcoming quarters. It appears to be very helpful to draw clear lines between what should be and what shouldn’t be your focus.

Keep in mind that it shall start here. Do not jump into working on your Key Results just yet. Take a step back on the “why” of the company and your team, and explicit your Objectives for the quarters. Only then, you may come up with meaningful Key Results.

So please, keep your team’s OKR to a maximum of 3 Objectives.

4 — Provide strategic context behind each Objective

As you start working on your team Objectives and drawing a very explicit line from those to the company strategy, take a moment to document the strategic context behind each of your team’s Objectives.

Ask yourself why you plan to work on this Objective. Do it three or five times. With your findings, provide strategic context behind each by formulating the why in writing with 2-3 sentences.

This Objective’s strategic context will be added in its description on Lattice.

5 — Start with action verbs, and be understood

Another value of the OKR methodology is to create commitment from the parties involved. To do so, start all your O & KR with an action verb.

Also, OKRs are public to everyone at Upflow. Anyone can go on Lattice, navigate the cascade of objectives and key results across all teams, and get a sense of what are the current priorities and progress. So write your O & KR in a way that is understandable to us all. Take someone very far from your specialty, and ask yourself, “Is he going to get it?”, if not, reformulate, write simpler.

6 — KR are not for business-as-usual

As you understand, the OKR methodology has many values. One of them is to empower teams to set goals that stretch beyond “business as usual”, and make significant, meaningful changes.

An easy trap is to write KRs meant to simply continue (not particularly improve) a current situation. OKRs should describe a specific focus of your team intended to make the company progress beyond the day-to-day. So write your KRs as changes to our current situation, not as a description of how they will make it persist.

It doesn’t mean that your KRs can’t cover any day-to-day item (ie. close X deals) but take this opportunity to craft them to drive a meaningful change, for instance on a subset you can improve (ie. close X% of our deals in this new geography, or on this new typology of clients).

7 — KRs are black-or-white and measurable

A good KR is a KR that is easy to identify as complete or incomplete. Be specific. There should be no gray area. At the end of the quarter, your team’s progress toward the KR should be unequivocal. Watch out, not making your KR black-or-white is a very common mistake.

8 — KRs have a single owner and a due date

OKRs create commitment. To do so, assign a single owner and a due date to each of your team’s KR. It can be you, as team lead, or someone else owning the delivery and completion of the KR.

Also, be smart about the way you set due dates. Don’t put them all on the last day of the quarter. Some KRs need to be completed earlier to either enable others or yield results. Use OKR to build anticipation in your team’s rhythmics.

9 — Identify your KR types

Depending on the stage of development of the objective you are working on, key results can be of different types. They can focus on other things than the outcome, as long as they are real commitments. Identify if your key results are a discovery, input, output, or outcome type. It will be helpful to align expectations.

People often think that OKRs are only meant to drive amazing results (outcome-driven). But it is important to be able to accept that an initiative may not be mature enough to think in outcomes just yet, and that it can actually be best to use a discovery or input KR to kickstart a project.

The KR type will be added in its description on Lattice.

10 — Craft tasks behind each KR

Once your Key Result is defined, align around a pragmatic plan for how to get there. Work on your battle plan in writing, make it explicit. It can absolutely (and likely will) be modified as you start executing and learning new information.

Well-crafted tasks help define what resources and steps are going to be necessary to complete your KR. It’s not critical that you get everything right up front, it is just critical that you have a believable plan.

This step can be done together with your team. It is actually a good way to bring them on board.

These tentative battle plans will be added in its Key Result description on Lattice.

… and a template.

More than that, we very early on built a template that team leads could follow when preparing their team’s OKRs for the quarter. The goal was both to save them time preparing and to save time for the Chief of Staff and the CEO when reviewing the use of the methodology and the proposed priorities.

Here is what it looked like in the early days. We invite team leads to complete the sections in orange. We propose later in the article a version of this template updated with our latest learnings.

Retro-planned to get started on day 1

Pulling out OKRs across several teams requires planning and sound execution. That’s why we paid attention to communicating early with the stakeholders on what was expected from them.

We want our quarters to start on day 1. For instance, Q2 shall start on April 1, not April 20 because we take the first 3 weeks of the quarter wondering what shall be our priorities. So strong anticipation is built into our planning. We start thinking about OKRs one month before the previous quarter ends - especially for department-level OKRs: it’s the duty of managers to ensure their team can hit the ground running on day 1 of the quarter. For instance, we start planning for Q2 on Mar 1 to be ready on Apr 1.

Our retro-planning looked like this in the early days (it’s evolved, as we will see in step 5). For simplicity, let’s assume we want to define our OKRs for Q2.

• ⏱️  Mar 1 — The Chief of Staff (or CEO) kickstarts the OKR preparation for Q2, sharing the resources and template available again.

• ✍️  Week #1 to #3 of March — Each team lead prepares a set of Q2 OKRs for their team, and anticipates dependencies with their peers. It is a bottom-up preparation process happening within the departments. The Chief of Staff remains available whether they have any questions.

• 👀  End of week #3 of March — All team leads send their OKRs proposal to the Chief of Staff for a first review and anticipate and prepare for the CEO’s questions.

• 🤝  Week #4 of March — Each team lead meets 1:1 with the CEO (in the presence of the Chief of Staff) to present their priorities, seek feedback and thoughts, and get formal approval.

• 📊  Before April 4 — Finalize tracking of Q1 OKRs. And add Q2 OKRs to our tool (Lattice). As owner of the planning process, and both to save time for team leads and to ensure OKRs are set up properly in our tool, the Chief of Staff takes care of entering every team’s objectives and key results on Lattice. It takes me an hour to set it up correctly, while it could take several hours for the team leads to do it and possibly miss something out. We are small enough to do things that don’t scale but promote a fast execution.

• 📣  First Friday of Q2 — Each team lead presents progress towards their team OKRs last quarter, and their priorities for the upcoming quarter, to the whole team during a 90-min all-hands meeting.

• 🚀  Q2 officially starts.

Step 3 - We added a light company level

After two quarters running with only the team-level OKRs, we felt it was time to enrich our practice and add a light company level in order to (A) reinforce alignment across the departments, and (B) to enable every team member to link their priorities to the company-level and thus reinforce a sense of purpose.

We kept it nice and short, for the reasons listed above: we were still small enough (~50 people of the entrepreneurial type) to be happy with something high-level.

Beta of Company Objectives

Our “beta version” of company objectives was the following. We decided on purpose to keep it quite generic, yet meaningful. It helped people relate to a high-level goal at the organization’s level.

1. Prove acquisition repeatability on 2+ channels

   We would simply tie all TOFU (Top Of The Funnel: Business Development, Marketing, Alliances) and MOFU (Middle Of The Funnel: Sales) functions’ efforts to this objective.

2. Win customers’ loyalty

   The Success team’s OKRs would be nested under this one.

3. Gather and build an inspired team

   The Talent and People team would contribute to this objective.

4. Control our corporate destiny

   And the Finance team to this last one.

Looking backward, this set is far from perfect. If they are problem-oriented, actionable, and relatively focused and cohesive, they are not very opinionated (they don’t translate a strong opinion of what to do and not to do). But that’s ok. I’d even say it was expected. This is working in small batches, following the 80/20 rule. You don’t get it right on the first shot - but you solve a big portion of the pain point (here, the team wanted to relate to a higher-level priority) with minimal investment, and can easily change as you mature. So overall, we were happy with this set of company objectives.

Beta of Company Key Results

Same logic here. In a group of 50 people, working on demonstrating the strong repeatability of their revenue generation engine, one metric mattered: ARR.

So our one and only company KR was our ARR target, as defined in our budget. It was simple to break it down into quarterly targets, and then follow progress on a monthly basis, and even in real-time (weekly).

Again, far from perfect - but it did the job really well of aligning everyone behind one metric that encompasses everyone’s contribution.

Step 4 - We onboarded the Product & Engineering teams

After another two-quarters of iterations on our implementation of the OKR methodology at the department (Business and cross-functional only) and company level, our Product & Engineering team expressed interest in joining the party. Note that it could 100% have been the other way around: starting OKRs with the Product & Engineering team, and scaling it to the Business and cross-functional functions. The key point is to start somewhere and ensure it eventually deploys to the entire organization.

With about a year of “local experience” running OKRs in the Business and cross-functional teams, we shared with the Product & Engineering leaders our best practices, resources, and templates. We held a meeting where the Chief of Staff replied to their questions and shared a few additional insights. Nothing ground-breaking, basically going through them the first 3 steps of this article.

We kept the same planning calendar and foresaw an opportunity to improve cross-collaboration, especially with the Business team. I, as Chief of Staff, thus encouraged everyone on all sides to talk to each other, and anticipate the dependencies between their projects and priorities. For instance, the Marketing team may want to implement a CMS to autonomously iterate on the pitch on our website - but this would impact the Engineering team. Better safe than sorry, the Marketing and Engineering leaders shall take a moment to align on the expectations to ensure this CMS project can be delivered smoothly (or decide to postpone it).

On top of orally encouraging leads to talk to each other while preparing their departments’ OKR, we added to the template placeholders so they can point each of their department Objectives toward a company Objective. We also added this additional checkbox at the end :

We then sort of assumed everything would organically go right.

Step 5 - Reinforcing cross-collaboration

In the first quarter, we implemented OKRs in the Product & Engineering team worked quite well. Their leadership found it useful to go beyond the day-to-day and steer changes in the team behaviors (ie. implement a new process). They saw the methodology as a great tool to (A) align the Business and Product & Engineering teams, and (B) to provide a clear understanding of how the Product initiatives/roadmap are linked to the company priorities.

One important thing that could have been done better is the cross-collaboration when the team leads were preparing their departments’ OKRs. They could have engaged in better communication.

To address this issue (in small batches), we proposed adding a new step to our planning calendar. We also leveraged this opportunity to add a second new step and reinforce the communication on the company direction.

Here is what our new planning calendar “in W” looks like, again in the example of preparing for Q2.

• ⏱️  Mar 1 — The founders kickstart the OKR preparation for Q2 in a meeting, gathering all the team leads. They reiterate the company direction short-term (for the next ~2 quarters), and answer questions. The Chief of Staff follows up by sharing the resources and template available.

• ✍️  Week #1 + #2 of March — Each team lead prepares a set of Q2 OKRs for their team, and anticipates dependencies with their peers. It is a bottom-up preparation process happening within the departments. The Chief of Staff remains available whether they have any questions.

• 👥  End of week #2 of March — All team leads gather in a 90-min meeting, in the presence of the founders’ office. In silence, they take time to read and review the proposals of each of the other team leads, and ask questions by commenting on the Notion notes. At the same time, they reply to the comments that are made on their team’s note. Toward the end of the session, if there is some time left, we engage in a live discussion to align everyone - or take it out of the meeting. (This step is inspired by Partoo)

• ✍️  Week #3 of March — Team leads integrate others’ feedback on their teams’ proposal, and meet to address the last alignment issues.

• 👀  End of week #3 of March — All team leads send their OKRs proposal to the Chief of Staff for a review, and anticipate and prepare for the CEO’s questions.

• 🤝  Week #4 of March — Each team lead meets 1:1 with the CEO (in the presence of the Chief of Staff) to present their priorities, seek feedback and thoughts, and get formal approval.

• 📊  Before April 4 — Finalize tracking of Q1 OKRs. And add Q2 OKRs to our tool (Lattice). As owner of the planning process, and both to save time for team leads and to ensure OKRs are set up properly in our tool, the Chief of Staff takes care of entering every team’s objectives and key results on Lattice. It takes me an hour and is set up right, while it could take several hours for the team leads to do it and miss some things. We are small enough to do things that don’t scale but promote a fast execution.

• 📣  First Friday of Q2 — Each team lead presents progress towards their team OKRs last quarter, and their priorities for the upcoming quarter, to the whole team during a 90-min all-hands meeting.

• 🚀  Q2 officially starts.

We will keep rolling it out and improving it in small batches

Our OKR journey is just starting. There is a lot to improve in our setup, and we know we will need to change it all when we outgrow our current practice.

Here are a few limits/opportunities that we foresee. They are neither exclusive nor ranked by importance.

• More thoroughly frame our OKRs at the company level. Our current setup works, but we could build a stronger and more opinionated set of company OKRs, which could provide the team with a healthy top-down direction, supporting everyone in deciding what to prioritise day-to-day.

• Support individuals who want to build their own individual OKRs. At the moment, we do not invite our team members to build individual OKRs. But it could make sense as we grow. We, on occasion, already see some team leads supporting some of their direct reports to do so but did not run any enablement session for everyone yet.

• Reassess our tooling. Lattice’s OKR feature has been working well for us so far. But we will likely need to benchmark the market, understand what tool companies 3 to 5 years ahead of us use, and assess if it could make sense to migrate now. Upflow implemented Salesforce with a team of 3 Business people, it’s in our DNA to anticipate tooling and adopt the best-in-class option that can scale effectively.

• Simplify the tracking of our KRs. For now, KRs can be tracked on a $ scale, a # scale, or in percentage. Assessing a precise percentage of completion for some KRs is not always easy. For instance, if one of my KR is to write a blog post on our OKR implementation, what’s the unequivocal milestone that tells me I’m at 20%, 50%, 75%, or 90% completion? It may actually be easier to select between 4 options, such as “Not started”, “Early progress”, “Good progress”, “Completed”. If I had to stop writing this article now, I’d likely pick “Good progress”.

• Improve the quarterly communication around our OKRs, especially the way we deliver them on a quarterly basis to the whole team. For now, we organize a Business Quarterly Review on the first Friday of the quarter and a Product Quarterly Review on the second Friday of the quarter. We could think about making it a single meeting where all teams gather, and publicly retrospect and present their priorities for the new quarter to the team.

Key learnings

To conclude, I want to leave you with a few learnings we have learned on our journey:

1. Get the founders’ active sponsorship. Most (to not say all) company-wide initiatives won’t work if you do not get the explicit buy-in of the ultimate decision-makers: the founders. You will need their top-down support to make it work.

2. Never forget why you implement OKRs. Don’t do OKRs for the sake of doing OKRs. It is supposed to help the organization, not hold it back or slow it down. Keep it all as simple as possible at any point in time. At our stage, we think that OKRs are mostly here to help people (A) know how their work contributes to the big picture, and (B) get a clear context in hand to be able to say “yes” or “no” to projects coming on the fly during the quarter. Your implementation shall ensure this is addressed, and it can be done in a very scrappy way.

3. Get someone who owns the OKR process. It may be a Chief of Staff, Strategic Project Manager, or Operator. It is very important to have a clear point of contact, and someone in charge who will ensure this planning exercise is successful. If you don’t publicly give this mandate to someone, chances are that the ball will often fall between players and that this OKR process will be very taxing for your team. It is a full-time job, at least for a few weeks every quarter.

4. Anticipate (a lot) your retro-planning. We found it best practice to prepare early so our quarters can start on day 1. Start sooner rather than later, and keep animating this planning exercise (again, it’s best to have an owner).

5. Embed formal cross-collaboration moments during the preparation phase. Cross-collaboration is complex, we recommend explicitly making some room for it, and supporting it with prepared rituals. Don’t assume it will happen organically.

6. Challenge the tracking of the KRs during the preparation phase. We had a few examples when one month passed in the quarter, and we realized that the teams’ did not have handy a proper way to assess progress toward their KR. They could be missing a dashboard or clarification on a definition. Ensure at the beginning of the quarter that every KRs can be easily tracked.

7. Communicate on your OKRs internally. You are putting some work on OKRs to drive alignment, so don’t let them die in a tool. Refer to them often, for instance in 1:1. Present your progress and roadblocks, for instance in team meetings. You are implementing them to organize, align, and give pace to your execution.

I hope that this article will support you in your OKR journey. It is not meant to be a formal playbook, but rather a point of reference and a source of inspiration for other early-stage companies.

I’d love to hear how you navigated your first months with the OKR methodology - feel free to reach out.