The Fastest Way to Speed Up Your Team Is to Do Less

When a team feels slow, the instinct is to add structure. Better planning. Clearer roadmaps. OKRs.

That instinct is almost always wrong.

I have been preaching Eisenhower for years. Those who have worked with me know this. Saying no consistently is one of the most powerful forms of performance optimization available to a CTO. Not the most comfortable, but one of the most powerful. And yet it's the first thing that gets abandoned under pressure.

OKRs are a symptom treatment

Teams introduce OKRs because something feels broken. Things take too long. Priorities are unclear. Everyone is busy but nothing meaningful ships. OKRs feel like a solution because they impose structure on the chaos. But they don't remove the chaos. They add a layer on top of it.

Quarterly goal-setting cycles, key result definitions, weekly check-ins, scoring sessions, alignment meetings about the alignment meetings. All of this is overhead, and none of it touches the actual problem. The team is still saying yes to too many things. People are still working outside their core competence. Decisions are still not being delegated to the right level. After the OKR rollout, the team has more to manage, not less, and the underlying causes remain exactly where they were.

The real causes of a slow team

In my experience, when a team is consistently too slow, it comes down to three things. Nobody is saying no, so every new request gets added to the pile and the team ends up spread across twenty initiatives moving at full speed on none of them. Delegation is happening at the wrong level, with CTOs staying involved in things that don't need them while letting go of things that do, creating bottlenecks in one place and quality problems in another. And people are working outside their core competence, which means a senior engineer spending half their time in coordination meetings instead of doing what they were actually hired to do.

Focus is not a mindset problem. It is a discipline problem. And it requires someone in the room who is willing to say no clearly and often.

The problem with saying no

Eisenhower is easy to understand and hard to apply. The matrix makes the problem visible. The organizational discipline to act on what it shows is a different challenge entirely, because Eisenhower is not a self-running system. Someone has to have the courage to act on what it reveals, at every level of the organization, not just at the top.

The most common failure pattern looks like this: a lead says no based on capacity and priority. A stakeholder escalates to another head. That head escalates to C-level. C-level says yes because the pressure is real and everything feels urgent. The lead now has a team doing more than it should, and the team has learned that no is negotiable. That destroys two things simultaneously: the prioritization culture and trust in leadership decisions.

Saying no as a CTO only works if the CEO and management hold the same no. If a CTO rejects an initiative and the CEO lets it in a week later, every stakeholder will fight for exceptions from that point on. And management is often where the biggest focus problem lives, because that's where the most external pressure lands. Investors, customers, market, competition. Everything has an advocate. The natural reflex is to let everything in. Real focus at management level is harder than at team level, and it's also where the consequences of losing it are most severe.

Delete more than you think you should

Saying no stops new things from coming in. Deleting removes what is already there. Both are necessary, and deletion is the more radical act because it leaves no back door.

A backlog with four hundred items is not a sign of good planning. It is a record of too many unfiltered yes decisions from the past. Every task sitting in that list costs cognitive energy even when nobody looks at it. The brain registers unfinished things as open loops, and open loops consume capacity that should go toward actual work. This applies at every level: backlog tasks that haven't been touched in quarters, code that is never called but still needs to be maintained, emails marked for later that will never be read, documentation nobody opens, tools still being paid for that nobody uses.

When you delete aggressively, the team doesn't just have less overhead. It thinks more clearly about what remains. Deletion is Eisenhower applied retroactively. Most of what has been accumulating is neither important nor urgent. It is just waiting. Let it go.

What actually works

The combination that consistently speeds up teams is simpler than any framework. Say no by default, and make sure the entire leadership chain holds that no when it gets pushed. Use Eisenhower not as a personal productivity tool but as a team-level filter. Delegate decisions, not just tasks: the seven levels of delegation give you a vocabulary for getting this right, and most leaders never explicitly develop this skill. Protect core competence ruthlessly, which sometimes means bringing in a Founders Associate early so the CTO focuses on technology, the CEO focuses on vision and stakeholders, and everything else goes to someone whose core competence is exactly that. And automate what cannot be delegated or dropped.

Eisenhower, delegation, core competence, deletion, automation. These are not productivity hacks. They are a coherent philosophy: the highest performance comes not from doing more but from being ruthlessly clear about what deserves attention at all.

OKRs ask you to define your goals more clearly. That is useful. But it skips the harder question: are you willing to say no to everything that isn't on that list, and will your management back that no when it gets pushed? Without that answer, the OKR is just more overhead on top of an unfocused team.

The fastest way to speed up your team is to do less. Fewer initiatives, fewer meetings, fewer open loops, fewer people working outside their strengths. What remains gets done faster, better, and with less coordination cost than anything a framework could deliver.