Focus on priorities: the discipline of less but better

Focus on priorities is one of the four cultural pillars of OKRs (alongside alignment, autonomy, and transparency). It means carrying a small number of Objectives per cycle, to avoid the dispersion that makes any strategy unworkable.

Andy Grove, the inventor of OKRs at Intel, put it simply: "if we focus on everything, we focus on nothing."

The rule of small numbers

Beyond those numbers, two symptoms show up: Objectives lose their force (no team member can recite them from memory) and trade-offs become impossible (everything is priority, so nothing is).

The hidden cost of not prioritizing

Not prioritizing doesn't mean "do everything". It means do everything poorly. Four measurable costs of a non-prioritizing organization:

• Cognitive fragmentation cost: a contributor juggling 7 topics loses up to 40% productivity to context switching, without noticing.
• Strategic dilution cost: no topic moves enough to produce a visible effect, the organization spins in place.
• Permanent arbitration cost: without clear priorities, every resource conflict re-negotiates individually, draining managers and teams.
• Engagement cost: aiming at everything, the team stops believing in anything. Motivation collapses within 1-2 cycles.

The most toxic anti-pattern: "everything is a priority, so nothing is". An organization that can't say what isn't a priority isn't practicing OKRs. It's filling out an ambition spreadsheet.

The direct link between focus and the number of Objectives

The cap on 2-3 Objectives per team and 3-5 per company isn't methodological vanity. It's the operational consequence of the focus principle. If a team exceeds that cap, it's mathematically because they didn't prioritize. The number is just the indicator.

Planning trick: instead of asking "what are our Objectives?", ask "if we could only carry 2, which two would you pick?". The constraint forces prioritization. And often, that's when the real strategy emerges, stripped of accessory good intentions.

Why focus is so hard to hold

Three pressures constantly push toward dilution:

• Inclusion reflex. Every team wants its own priorities reflected in the company OKRs. Leadership gives in.
• Fear of forgetting. "If this important topic isn't an OKR, what carries it?". Answer: ongoing operations and KPIs. Not everything has to be an OKR.
• Excessive ambition. Wanting to tackle ten topics in parallel instead of finishing three.

How to force focus in practice

• Set an explicit cap (3 company Objectives, 2-3 per team) before OKR planning.
• List explicit "non-priorities" for the cycle. More powerful than listing priorities.
• Separate OKRs and ongoing operations. A CS team that has to "answer tickets" doesn't need an OKR on that. It's a KPI to watch.
• Run the memorability test: every team member should be able to recite the team's Objectives without checking the doc.
• Refuse "and" Objectives. An Objective that contains an "and" is almost always two Objectives in disguise.

Common focus mistakes

• 5 to 7 Objectives per team. Classic symptom of a team that can't say no.
• "Catch-all" Objectives. "Improve product, marketing, and sales performance" is actually three poorly-named Objectives.
• Same Objectives every quarter. If you've never finished an Objective in 3 cycles, it was either poorly framed or never really a priority.
• Adding Objectives mid-cycle. Avoid unless a major event hits. What's not in this cycle's OKRs waits for the next.

Serendly insight: "what can we stop doing?" beats "what should we do?"

The real focus question isn't prioritization, it's abandonment. Most teams don't suffer from a lack of good ideas, they suffer from too many simultaneous commitments.

At start of cycle, ask each team: "what can we stop doing this quarter to really land our 2-3 Objectives?". The answers are often more revealing than answers to "what should we do?".

Install a real focus discipline

Focus is one of the hardest competences to sustain cycle after cycle. Let's talk about how we help teams hold the line.

Request a demo