Will OKRs Work at Your Startup?

Roadmap planning, goal setting, target projections etc. are all things that can make grown people cry themselves to sleep. Admit it, you’ve been there too! One tool often used to take some of the bite out of those exercises is OKRs (Objectives and Key Results). Most recently made popular by Google after being picked up from Intel, many swear by them. According to others OKRs belong where all three letter acronyms should go – which is not a nice place. In reality it depends what you need OKRs to do for you and how you use them. For a primer on OKRs see Rick Klau’s Google Ventures talk.

OKRs have been hugely helpful for me. They have helped me think about different planning horizons, balancing stretch goals with real progress, and to consider personal vs team vs company goals. When they function well they can help individuals, team members and different teams in a company maintain a shared framework of priorities that are clear enough to help decision making and flexible enough to let people make the decisions they need to make. There are times when they’re not great and like any other tool, outcomes depend on how they’re used. No tool will work well with a bad process.

Stage where OKRs are worth considering

• Growing company that is developing into discrete teams (Eng, Marketing, Sales, Account Management, Customer Support)
• Growing size of individual teams (>3 people)
• Some level of constant changes internally and externally

OKRs (or other planning) framework helps

• Focus on the right things (“good idea, but not a priority re. the objectives”)
• Make rapid changes to strategy and focus when needed (“this is now the priority instead of that”)
• Individual decision making (easier to “run with things”)
• Individuals and teams maximize progress through stretch goals
• Spot where teams need to cooperate, who to inform (ex. Marketing doing content strategy, AM rewriting Desk.com content -> coordinate)
• Make plans visible by being in one location where they can be referenced (site, deck, poster, tattoos – whatever works)

Risks of beginning to use OKRs

Things change too fast for the length of the planning period. Strategy, priority, product changes fundamentally within three months.

Too much uncertainty for multiple planning horizons. If all you know is what you’re doing right now and have no idea what will be next until you’re done, then don’t waste time planning at multiple levels.

Death by planning: hours/days/weeks spent writing, coordinating and tuning plans. Mitigate by forcing simple format, ex. max 3-5 objective and corresponding key results, described in max two sentences per item. And by using a format that is useful in standups, weekly team meetings etc. Hunt down duplication of content with a vengeance.

Death by performance management: OKRs are related to performance management and operational metrics which are other key tools to manage teams. But OKRs are directional, not absolute! They are meant to inform direction and help people hit their max by aiming for stretch goals without fear of failure. Using as direct performance measurement tools encourage sandbagging. Difficult balance, but should be kept separate to some degree.

Death by being useless: not a risk really, if it’s not useful it should die. It will take a test period (first three months probably) with some level of implementation pain to get going. Must be put in place, managed and championed by CEO since it has to become culture.

OKR structure

OKRs should be a very lightweight structure that makes everyone do better work. There are very few elements to getting started.

Objective

Stretch goal, obtainable under perfect conditions, but almost (if not completely) out of reach. Should feel slightly uncomfortably challenging without being stupid. 3-5 of these. The fewer and bigger the better.

Key Results

Specific, measurable results that can be graded. 3-5 key results per objective.

Process

At Google we set OKRs at a company, team and individual level. Each person set their own OKRs based on team OKRs. Teams set their OKRs based on company OKRs. It wasn’t a strict top-down or bottom-up approach. More like a few rounds of iterating at all levels to get to the final set.

Company and team level OKRs only plus separate individual productivity metrics may work for some teams depending what the work is. An engineer and a sales person have different circumstances when it comes to stretch goals vs. performance. They have different ranges of ability to impact the goals and the outcomes of their work.

Grading was done on a 0.0-1.0 scale. 0.6 – 0.7 was the target area. If you hit 1.0 you probably hadn’t set yourself enough of a stretch goal.
It is part honor system, based on the assumption that everyone is already performing well, and further motivated by challenging goals. Hence the partial separation of performance evaluation from OKRs. OKR grades could impact, but were not the same as your performance rating.

Bonus: Weekly snippets to help productivity and internal communication

Most teams used some form of weekly updates (snippets, Idonethis.com or whatever works). Helps internal “peripheral awareness” of what others are working on, as well as personal productivity.

Again, has to be extremely lightweight (~10 minutes every Friday or similar) and should directly help people evaluate their progress and make course corrections in addition to keeping team or company updated. Lets you follow people and projects you want to pay particular attention to, to see how their work is progressing. Without doing round-the-room updates with everyone present.

Overlaps with any current stand-ups and check-in meetings, so not something that should just be added on top.