Timestamps

1:08 - What an OKR is

5:02 - Why startups aren't too small to work with OKRs

13:47 - What a bad OKR looks like

21:18 - Who should own the OKRs

30:20 - How often you should discuss OKRs

About Martin Voynikov

Martin Voynikov is the director of product at Gtmhub, an OKR software for missions that matter. He previously helped create a mobile AR game called Visualz Inc, which guided kids through museums in a fun, enriching way. More recently, he was the product manager at eezy.ai, the mobile personal assistant, and Babbel, the language learning app. He is also a guest lecturer at the University of St Gallen, teaching product management analytics and heuristics.

Before coming up with some OKRs, you should make sure you have a good idea of your vision ("what will we become in 10 years?") and mission ("what purpose do we serve?").

How to create a good OKR

Let's say your startup wants to be the best in terms of product: a good objective would be something like "becoming the category leader".

The key results would be:

1. Have an NPS of +50

2. Have 100 5 star reviews on the app stores

3. Have the product team launch a new feature each month

This last key result is debatable: some might say that's just output, not outcomes. But if you tie it to another OKR, like "0 bugs introduced", or "all new features having a customer effort score of 5", you're basically counter-balancing it.

How many OKRs do you need?

Answering this question requires making two important distinctions:

1. How many key results should you have in your objectives?

2. How many objectives should you have?

Data has shown that the most successful objectives are the ones who have up to 4 key results. Numbers have also shown that an individual should usually own no more than 2 objectives. Teams, depending on their size, can accomodate between 2 and 3.

What timeframes should be used for OKRs?

The vast majority of companies are using quarterly OKRs, that tie to an annual OKR, which in turn might tie to a 3 year plan. Circumstances for companies change pretty fast, so monthly OKRs are probably the optimal option — but the quarterly basis might also work for you.

If you feel like even a quarterly OKR is too time-critical for you, keep it nonetheless and transfer it to the next quarter if need be.

How often should you check OKRs?

In order for OKRs to really work for your company, you should create some ceremonies around them: your cadences of review should be regular. Gtmhub sees clients using mostly weekly reviews, but bi-weekly can work as well.

OKR meetings should be quick and should focus on the next steps and on your learnings. No need to let people see you update the OKR card. Each person should be allotted 2-3 minutes to speak.

Memorable Quotes

"A good OKR has an aspirational objective, measurable key results and outcomes. A bad OKR is a mediocre objective with non-actionable key results."

"OKRs have already become a competitive advantage in many industries."

Resources Mentioned

- Oceans 11 twitter thread

- Gtmhub's OKR marketplace

If you would like to listen to more conversations on startup organizational methodologies, check out our episode with EOS implementor Christian Burger.

190期：月读·日本的过去会是我们的未来吗？- Japan’s past, China’s future?