Merging OKR concept on our PI Planning event

Nicolas Martin
ManoMano Tech team
Published in
6 min readSep 9, 2022

In spring 2021, we decided to build our own agile framework and commit to share it regularly with the agile community.

It is named MACS and stands for ManoMano Advanced Collaboration System.

You can learn more on those MACS introduction articles:

Today, I will share with you how and why, in our PI Planning events, we moved from PI Objectives with business value assessment to a complete OKR approach from company to team level.

To get the most of this article, I assume you know:

Our two first PIs with PI Objectives

A PI objective is an objective written by the team during the PI planning event. The team is able to write this objective because they build a plan and secure it by anticipating the risks and dependencies.

On ManoMano, we followed that way during our two first PI. Here are some examples of objectives the team committed on.

During the PI event, on day 2, we have a time called “Business value assessment” on which the train leaders (head of product, engineering, Agile delivery leader and main business owner) go to one team after the other and discuss those objectives to assign them a business value.

Goal is to trigger a discussion between engineering, product and business people to prioritize those objectives with a business value note from 1 to 10.

Pros and cons of PI Objectives and business value assessment

Pros

  • PI Objectives are a great way to have team focus on outcome and not output
  • Business value discussion is a great way to create a proximity discussion between developers, product managers and business owners to align them on why they are working on those topics, engage the whole group of train members on the same goals

Cons

  • Using the business value criteria to prioritize objectives gives a ‘by default’ more important voice to the business roles
  • What about the user value or tech enablers that do not have direct business value?
  • Is a goal rated 6 in Team A less important than a goal rated 8 in Team B?
  • It was not easy during the whole PI (one quarter in ManoMano) to check if the objectives were progressing on the right track
  • It wasn’t easy at the end of the PI to assess if the objective was reached yes or no, most of the time we were on the gray zone.

The OKR MACS experiment

In MACS, we try to avoid jumping on solutions. We identify ways to improve, we experiment and if it is satisfying, the experiment becomes part of the system.

How did we integrate OKR in PI Planning?

At ManoMano, we have three levels of OKR.

Company Strategic OKR

The ManoMano executive board defines company objectives to be reached each 6 months. We call them Company Strategic OKR

The key results are attached but not fully defined, KPI criteria can be mentioned but they do not have a definitive target defined.

Intention is to have those Company Strategic OKR updated and shared one month before each PI (sometimes it can be one week 😅)

Besides being a prioritizing tool in conflicting situations, Company Strategic OKRs are guidance on where to focus energy, they give clear directions to the teams on what are the main problems the company wants to solve and what are its main challenges.

Team PI OKR

During the PI Planning, Teams confirm on which Strategic Company OKR they can contribute to. And by defining the Team PI OKR, they define to what extent their efforts will help the company reach its objectives (given their scope, their capacity).

Our template for team to write down PI OKR

Train Aggregated OKR

At the end of PI Planning, as we have more than 50 teams in 7 trains, we need an abstraction layer to get an overview of all Team PI OKR.
So, each train leaders aggregates their Team PI OKR to generate Train aggregated Objectives.

Those train aggregated objectives can then enable the Company Strategic OKR to be updated with accurate Key Results.

Hence we have a top to bottom to top OKR update.

Three levels of OKR, top to bottom to top

Is it better ? Let’s get a step back to check if our cons from previous situations are answered.

Cons

  • Using the business value criteria to prioritize objectives gives a ‘by default’ more important voice to the business owner role
  • What about the user value or tech enablers that do not have direct business value?

OKR are not business centric or user centric. They are agnostic. It is up to the teams to define the best objectives they can set to participate at the global train and company strategic levels.

We do not have a Business Value assessment where business owner’s note objectives from 1 to 10. Now, on the middle of PI planning day 2, we gather train leaders and business owners to discuss each team OKR one by one.
The discussion is focused on why they choose to commit to this objective, what are the KR attached, are they ambitious enough, clear enough for all?

The OKR focuses on the incremental value that each team will deliver which gives a better sense of what the collective energy will reach (compared to a BV 1–10 score, that is relative and could vary a lot from a team to another).

  • It was not easy during the whole PI (one quarter in ManoMano) to check if the objectives were progressing on the right track

We specifically try to follow Key Results as SMART items. That way, we can measure their progress on each 2 weeks iteration and assess if we are on track or not.
(let’s be honest here, we still have some Key Results that are not smart, but most of them are great; big kudos to the teams here).

Example of a team OKR

And so, what is next?

We are globally satisfied with the move from PI Business Value objectives to OKR but we are now facing two main difficulties on which we are currently experimenting.

First, on which sync meeting, which participants update and assess KR progress? We already have multiple synch rituals in place, and we want to avoid having too many to try to make them as useful as possible.

Second, but that is linked to the first point, which tool do we use to monitor those OKR?

We had a Google sheet but reaching more than 50 teams now in 7 trains, we are not convinced it is scalable.
And we would love to have it all within our Jira that already regroups all delivery items, so that we could link on the same place ‘delivery items’ and ‘global OKR’.

Answer on next episode, :-)

We ❤️ learning and sharing

I took a lot of pleasure to write this article, feel free to post your feedback below and reach out to me on LinkedIn. Whether you had a similar or totally different experience, I’d love to hear about it.

Oh, and by the way: we are hiring in France and Spain.

--

--