OKRs in strategy execution: when the method works and when it doesn't

What do OKRs do for strategy execution?

OKRs translate a strategy into goals you can review each quarter. An objective describes what you want to achieve; the key results make it measurable how you'll know.

The real gain is rarely the goal system itself, it's the rhythm that comes with it: every three months, the same group sits down and takes stock. Many companies that benefit from OKRs would have benefited just as much from any other fixed rhythm.

The second gain is visibility across departmental lines. When sales and production see their goals side by side, contradictions surface that used to go unnoticed in separate meetings.

What has to be in place first?

One thing, and it isn't negotiable: **the direction has to be decided.**

OKRs are a translation tool. They take a decision and break it down. If there's no decision, a goal system still emerges, it just describes day-to-day business in a new format. That only becomes apparent after two or three quarters, and then the method gets written off as a failure, even though what was missing was the decision beforehand.

A simple test: can management say in two sentences what the company is betting on for the next three years and what it's giving up to do it? If not, what's needed is a strategy process, not a new method.

How do you derive OKRs from your strategy?

In three steps, and the first one gets skipped most often.

• Take the fields of action from the strategy, not from the departments. If every department derives its own OKRs from its day-to-day work, you get a collection, not a system.
• No more than three objectives per quarter for the company. A leadership team can't seriously track more than that. Everything else is day-to-day business and doesn't need a goal.
• Key results measure impact, not activity. "Concept completed" isn't a key result, it's a task. "New customers from the new segment at 15%" is one.

The third point is the hardest, and most rollouts fail on it, not on the method. A goal system built from activities is a to-do list with extra maintenance.

When does the quarterly rhythm not fit?

More often than the OKR literature suggests. Three cases from mid-sized companies:

• Long project cycles. If you build plants and your projects run eighteen months, you can't define a meaningful quarterly milestone without inventing one.
• Highly seasonal business. When two quarters decide the year, an even quarterly cadence is an artificial frame.
• Small leadership teams under about a hundred employees. The same group already meets every week. The gain from formalizing is small; the maintenance effort stays.

In these cases, annual initiative planning with fixed review dates works better. It delivers the same thing, visibility and rhythm, without forcing a cadence that doesn't match the business.

What are the most common mistakes when introducing OKRs?

• Too many goals. Three objectives with three key results each means nine numbers somebody has to maintain. Five objectives means fifteen, and at that point nobody maintains them.
• OKRs for everyone. Rolling them out across the entire company at once overwhelms the organization. The leadership team alone, for two quarters, is the more durable start.
• Tying them to bonuses. As soon as money depends on the level of goal achievement, goals get written cautiously. That kills the benefit.
• Tool before method. Picking the software is the most enjoyable part of the rollout and the least important. The first two quarters work fine in a spreadsheet.

What are the alternatives?

The most common one is the simplest: a list of initiatives with owners, deadlines and a fixed review rhythm. For most mid-sized companies, it does the same job as OKRs, without its own vocabulary and without a rollout project.

If you want to steer more heavily by metrics, the Balanced Scorecard offers an older but robust framework. It's more cumbersome, and in return less prone to the quarterly pressure that leads to cautious goals with OKRs.

The choice of method rarely decides the outcome. The fixed date does.