How to Write OKRs That Don't Suck by Adrian Howard - Mind the Product

Consultant, coach and trainer Adrian Howard has, over the years, seen many organisations attempt to implement Objectives and Key Results (OKRs), generally in response to a sense that “everything is wrong” and with a hopeful attitude of “these seem simple and clear”. In this MTP Engage Manchester talk, he explains how to write OKRS that, in his words, “don’t suck”.

OKRs, Idealised

Objectives and Key Results sound deceptively simple on paper. In theory, you set an organisational Objective (a vision or mission, perhaps?), and then work out what measurable Key Results would represent success. And then there is supposed to be an elegant cascade, where each business unit and team derives their smaller-scale OKRs from the structure above them.

Sadly, a lot of organisations fall into the trap of cargo-cult OKRs, where they expect the rituals to solve their problems while losing sight of the underlying thinking and practices. They look for tools to solve their problems as fast as possible. They expend a vast amount of energy trying to create their first ORKs, and then… they hope. They operate as if just having OKRs will somehow, magically, make them more efficient, aligned and successful.

OKRs are not Magic

OKRs are a powerful and useful tool in your organisational design arsenal, but they’re not going to magically transform your business on their own. They are not a substitute for a strategy or vision, and they won’t conjure organisational alignment out of nothing.

As with all tools, they are powerful if they’re understood and used correctly. The good news is that many of the ways that organisations struggle with their OKR implementations are predictable and, even better, fixable.

OKR Antipatterns (and Possible Fixes)

Here are some common red flags that Adrian has seen as organisations struggle to start using OKRs, together with things you can try to get yourself back on track.

All-Encompassing OKRs, and Generic Objectives

Early deployments of OKRs often lead to organisations trying to OKR All The Things – every previous metric suddenly becomes an ORK, without thinking about why it should be an OKR. Alternatively, organisations often start with Generic Objectives (e.g. increase revenue by x amount in y time frame) – which give your teams very little clarity or strategic direction.

To start solving this, try going up a level in your thinking – Ask yourselves “Why do we want to track this metric? Why is this objective important?” The answers to those questions will hopefully lead to more informative OKRs.

At their best, OKRs are a form of strategy deployment. They’re a way to condense key strategic ideas into actionable, measurable goals that can be understood at every level of the organisation.

OKRs give you a way to distil vision and mission down into lower-level goals, which you can then use to actively steer your activities. They need to be at a suitable level, and framed in a way that helps you guide action.

Quantitative Objectives or Output-based Key Results

If you find yourself writing quantitative objectives, you’ll hopefully realise that these are rarely strategic – they’re just things you’d like to happen, with no guidance as to the “Why”. In a similar pattern, you might encounter output-based Key Results – KRs that specify milestones, features or releases. Unless there are circumstances where the output is the point of the work (these are rare), objective-based KRs give you no way to steer or make strategic decisions about your work.

To move beyond these patterns, remember that objectives should tell stories. They are qualitative articulations of your strategy, and the key results are just a way to better understand and articulate those stories. A good way to work towards that is to regularly ask “Why”. For example:

• Why do we want to shift these numbers? How do we think we’re going to do that? Why is that important to our strategy?
• Why do we need these features shipped? How do we know if they’re solving our strategic goals? What are the outcomes we’re hoping to see as a result?

Those might be slow and challenging conversations, but they will help you start to understand the things we should actually be working towards and measuring.

Copy/Paste OKRs all the way Down

When OKRs are mechanically cascaded down the organisation (e.g. a high-level KR becomes a low-level O), these top-down OKRs remove any kind of strategic understanding from the process, and rob teams of their ability to understand their decisions and steer their work. And if OKRs are pushed down to the level of individual contributors, this is generally an antipattern that can start to feel like process for process’ sake.

The work of teams and individuals generally operates on a different cadence to the organisations that surround them. Attempting to automatically cascade OKRs down the organisation will likely disempower your teams and contributors, and force them into work that isn’t actually fulfilling the organisational strategy as efficiently as it could.

At the team-level, it’s generally advisable to let leaders set strategic Objectives, and then invite the people doing the work to set the Key Results. This sets up a conversation where everyone can seek to understand the strategy more clearly, teams can work out how best to understand that measure success. At the personal level, using OKRs for personal development is a great way to understand your own strategic growth, but that will operate a different scale and cadence from the team around you.

OKRs That are Rapidly Cycled or Largely Ignored

If your OKRs are changing dramatically every quarter (and you’re not a startup that’s learning and pivoting), it suggests that you haven’t understood your strategy, or that the OKR system is being used to grade people, rather than guide them. On the other hand, if nobody is referring to or using OKRs on a daily or weekly basis, it’s a sign that there’s a misalignment in either their scope, level, or strategic focus.

If your objectives are useful to the teams and the end customers, they should be used on an almost daily basis. If that’s not happening, something is wrong with the OKRs as they are currently formulated. Likewise, while your OKRs need to be useful to teams and customers, they also need to be useful to the organisation as a whole. Strategic direction should generally change slowly, and your OKRs should evolve in step with that.

You should regularly retrospect on your OKRs, and attempt to understand how they have helped or hindered teams. Were they at the right level of granularity? Did they help steer teams to make impactful decisions? Did they reflect outcomes that benefited customers? And to make sure you can have meaningful conversations, find ways to bring the OKRs into your work on a regular basis. Talk about them. Put them on the wall. Use them for planning and retros. If they’re not helping in those scenarios, then that’s useful insight for the next retro.

Targets and Tools Skewing Your OKRs

Beware of jumping to a tools-first approach, and beware of Goodhart’s Law – “when a measure becomes a target, it ceases to be a good target”. Getting caught up in selecting an OKR tool in your first week will almost certainly mean you’ve picked the wrong tool for how you operate after a few months. Likewise, getting caught up on improving your key results can hide the fact that it’s entirely possible to actively improve a key result in a way that doesn’t actually support the overarching objective.

Thankfully, the ways to avoid these issues are actually very simple. To start with, don’t worry about tools – just use whiteboards and spreadsheets. They are flexible, cheap, and everyone understands how these kinds of tools work. They’ll let you adjust and evolve your processes, learn how your organisation needs to use ORKs, and then – if you really want to – you can make a more educated selection of a formal tool.

And to make sure you’re actually working towards meaningful objectives, rather than just making a bunch of numbers go up, try pairing your key results. Find a KR that you’re hoping to optimise for, and pair it with something related that will help you check if you’re doing something wrong. A great example of this might be New Customers v.s. Cost of Customer Acquisition – if both of those start going up together, you might be doing something wrong.

Key Takeaways

OKRs are a tool for strategy deployment. They must be relatable and understandable, and give the organisation the information and tools it needs to steer towards the desired outcomes. If this isn’t the case, then you should make adjustments.

Regularly retrospect on your OKRs – ask questions about how well they supported and guided decisions, or whether the key results actually lead to the objectives you wanted. Asking these questions will help you better understand the strategy, and better guide and align your organisation.