OKRs are not only for big companies such as Google or Twitter. They can also help align the work in a startup. This alignment is very valuable since it's relatively easy in a startup to get distracted with many irrelevant but time-consuming tasks.

For instance, if your main objective is to raise money, how important is to dedicate time to the following things?
  • Designing a nice logo
  • Legally establishing your company
  • Creating a company Facebook page
  • Creating a company Google+ page
  • Creating a web page

The answer to this question becomes clear after setting up the startup OKRs. So, let's create an example of OKRs in a startup that wants to raise money. This imaginary startup is developing a mobile game called MyMobileGame and wants to raise $1M in 6 months. 

Instead of showing the OKRs as a pyramid we use the following alternative representation, which is visually easier to scale to a large number of OKRs (think of it as a folder structure representation):

OKR Example startup company business
Example of OKRs for a startup raising $1M for a mobile game

We see three levels of objectives with their corresponding key results. For the sake of brevity some objectives have been left empty. 

In order to achieve the startup's main objective to raise $1M in 6 months, the startup founders believe that they need to obtain three key results.

OKR startup objectives results

Key result 1 is to get to the point of discussing the details of the term sheet (governing the legals of the investment) with at least 6 investors. Once you reach this point when discussing with an investor, the probability of successfully closing the deal is high, but does not reach 100%. The founders are probably thinking of securing $250,000 from 4 investors but want to deal with 6 in order to have a stronger negotiating position and to have a plan B in case 2 investors decide they are not interested.

Key result 2 is to reach 50,000 installations of their mobile game. The founders in the example are probably thinking about impressing the investors with this number, which is called a vanity metric

Key result 3 is about finding a product-market fit. This means that the founders have developed a game that people like playing. The founders added this key result because they think it is necessary in order to attract the attention of the investors.

Discuss the Term Sheet with 6 Investors

The CEO of MyMobileGame has assigned to herself the objective of negotiating the term sheetOKR startup example investor objective with 6 investors. Notice that this was a key result of the company objective, but now that it's assigned down the startup hierarchy it's become an objective by itself!

OKR startup example marketing objectives

The CEO thinks that in order to fulfill this objective she needs to obtain three key results:
  • to go and pitch in 20 different pitching events, like conferences and meetups
  • to reach out to all her contacts and get 30 intros to investors
  • to fill in an application and send it to 20 different Business Angel Networks

Get 50k Installations

One of the startup founders specializes in marketing (he's the CMO). He's assigned to himself the objective of getting 50,000 installations of MyMobileGame. Just like before, this objective is regarded as one of the key results to the main startup objective. 

OKR startup example product objectives

The CMO thinks the best way to get 50k installations is by:
  • contacting bloggers in the mobile game industry and offering them to write a post about their mobile game
  • launching 3 ad campaigns; in Facebook, Twitter and Minimob. 

Validate Value Proposition (Product-Market Fit)

There is a third startup founder with a strong tech background. She has taken upon herself to iterate and improve the mobile game until she can obtain direct evidence that there is a product-market fit, following the lean startup methodologyOKR startup example objectives.

OKR startup example objectives results

She knows that the way to reach this goal is to:
  • Reduce the cost with which the mobile game acquires new users below $1.3 per install (CPI stands for Cost per Install)
  • Increase the percentage of average daily active users beyond 30%. A user is considered to be active in a given day when she opens the mobile game app. 
  • Reduce the churn rate below 20%, which tells how fast the mobile game app loses installations

Notice something interesting here, which is that the Acquisition key result (KR1 - average CPI < $1.3) has a direct dependency with the Marketing key results (be featured in 20 blogs and launch 3 media campaigns). In fact, the average CPI is ultimately affected by:
  • How interesting the game is (value proposition)
  • How efficient the marketing campaigns are (Marketing)
Dependencies between key results belonging to different objectives naturally arise in OKRs. Their management require coordination between the corresponding people or teams (in our case, between the marketing-oriented and the product-oriented founders). 

Get More than 30% Daily Active Users

This is an example of how we can have a third layer of objectives. The objective that requires getting more than 30% daily active users can be measured and fulfilled with more key results. 

OKR startup example objectives results

In our case, the product-oriented startup founder knows that the way to increase the number of daily active users is by:
  • Implementing mobile notifications that invite users to play the game, and 
  • Giving a chance to claim a daily reward (that provides some kind of advantage in the game) to those users who play the game every day

When do we Have Enough OKR Levels?

The recursion of interpreting a key result as an objective with its own key results could in principle go on and on forever. However, as we add a new level of OKRs two relevant things happen:
  1. The number of objectives grows
  2. The level of detail of objectives increases

There is a limit to the number of objectives a given team can handle, and objectives with an increasing level of detail stop being useful at a certain point and become just tasks. This means we have to cut the pyramid at some level. 

In practical situations, it is useful to cut the pyramid at a point in which each person has no more than 3-4 objectives assigned. Going beyond this limit means too much context switching, and that people won't have their list of objectives on top of their heads. 

Adding Tasks to the Last Level

Each key result at the lowest level can be accomplished through a task list. For instance, in order to implement the daily rewards (KR2 to get more than 30% daily active users) we may want to break it down into a list of tasks that looks like this:
  • Define table of daily rewards
  • Implement Android version
  • Implement iOS version
This list of tasks is managed outside of the scope of OKRs, typically with some other methodology like Scrum Agile MethodologyOKR startup example objectives results. Each one of this tasks could be regarded as one or more user stories to be implemented in one or several sprints. (Watch this for an introduction to Scrum).

Getting Back to the Original Question

So, after writing down the OKRs for the startup, how important is to dedicate time to the list of things mentioned at the beginning of the post? This is how the founders may go about prioritizing this list of tasks according to the goals they've just set:

  • Designing a nice logoMedium priority. The logo is needed to pitch investors but it will probably not influence their decision much.
  • Legally establishing your companyHigh priority since investors will not invest unless a company is legally established. 
  • Creating a company Facebook pageHigh priority since a Facebook ad campaign is required to get 50k installations.
  • Creating a company Google+ pageLow priority since it is not related to any objective or key result.
  • Creating a web pageLow priority since it is not related to any objective or key result. 
One of the good things about OKRs is that you don't need to create a list of tasks like the one above and then see if it makes sense or not to execute each task. Instead, you just break down your top objective into several OKRs and then the tasks required for obtaining each key result at the lowest level will just come out naturally. This way you won't even have to lose time evaluating if a Google+ page makes sense or not.