OKR consulting that changes
what happens on Tuesday
Most OKR work fails quietly: the objectives get written, everyone nods, and the week carries on exactly as before. I work with leadership teams to make them decide something: fewer priorities, named owners, and a review that can actually stop work.
Three ways people arrive
First time
You've read the book, you can see why it should work, and you don't want to spend a year discovering the expensive mistakes for yourself. The goal here is a first cycle that survives contact with a real quarter.
It didn't stick
You rolled OKRs out, enthusiasm lasted about six weeks, and now they're a spreadsheet nobody opens. Something specific went wrong. It's usually diagnosable, and it's usually not the thing people blame.
You've outgrown it
What worked at forty people is creaking at a hundred and fifty. More teams, more dependencies, and objectives that no longer connect to what any single team can actually move.
Where rollouts actually go wrong
Rarely in the writing. Almost always in what the organisation does with them afterwards.
The roadmap, renamed
Objectives that are the projects you were already committed to, relabelled as goals. Nothing was chosen, so nothing was dropped, and the list is exactly as long as it was in December.
Key results that count activity
"Ship fourteen features" is a plan, not a result. The test is simple: if you could hit every key result and the business would be materially unchanged, you're measuring effort.
Stretch goals wired to pay
Tie an ambitious objective to someone's bonus and you've asked them to negotiate the target downward. They will, sensibly. Sales quotas work precisely because they're committed numbers that everyone treats as committed. The mistake is running stretch OKRs through the same machinery.
Owned by a committee
"The leadership team owns it" means nobody does. Every key result needs one name against it. Someone who'd feel personally awkward reporting no movement three weeks running.
There are three structural failure modes underneath these, and they're worth understanding before you relaunch anything: how OKR rollouts fail, and what a diagnostic looks for.
What I actually do
Clarify the strategy
You cannot write good objectives against a vague strategy. The fuzziness just moves down a level and multiplies. Often the first sessions aren't about OKRs at all, but about getting the leadership team to say the same thing about where the business is going.
Cut the list
The hardest and most valuable part. Deciding what you're explicitly not doing this quarter, out loud, with the people who'll have to defend that decision to their teams.
Design the architecture
How many objectives, at which levels, owned by whom, and how much of the organisation is in scope for the first cycle. Usually less than you'd expect.
Facilitate the sessions
I run the setting sessions myself, in the room, including the uncomfortable parts. An external facilitator can ask the question everyone is avoiding, which is most of the value on the day.
Fix the measures
Turning intentions into key results that would actually tell you something mid-quarter, and stripping out the ones that only report effort.
Install the cadence, then leave
The review rhythm, who runs it, what happens when something is off track, and coaching your internal champion until they're running it without me. If you still need me after two quarters, I've done this badly.
What it takes
A first cycle is a quarter. Not because the setup takes three months, but because a quarter is the shortest period that proves whether the cadence survives a real one.
Weeks 1–2
Diagnosis and strategy clarification. What's already in flight, what's actually being measured, and where the leadership team disagrees without knowing it.
Weeks 3–4
Architecture and the first setting session. You leave with a short list, named owners and measures that mean something.
The quarter
Fortnightly reviews, coaching the person who'll own this after me, and a genuine retrospective at quarter end on the process, not just the scores.
Who this works for, and who it doesn't
A good fit
Leadership teams who will be in the room themselves, own their own delivery, and are prepared to cut the list rather than relabel it. Typically 20–500 people, or a unit that size inside something bigger. It helps enormously if the person at the top is willing to say out loud what the company is not doing this quarter.
Not a good fit
Anyone shopping for an OKR tool rollout, or for OKRs to be "done to" their teams while leadership stays out of it. If the real goal is a reporting layer that makes the board pack look tidier, I'm the wrong person and I'll say so on the call.
Start with evidence, not a proposal
I don't quote for an OKR programme off the back of one conversation, because at that point neither of us knows what's actually wrong. Nearly everyone starts with the Execution Diagnostic: two weeks, fixed fee, and a straight answer, including "you don't need a consultant for this" when that's what the evidence says.
£7,500 + VAT fixed · no follow-on commitment
FAQ
Which OKR tool should we use?
A spreadsheet, until the process is working. Every tool I've seen bought early was bought to avoid a conversation about priorities. Once the cadence is real and the list is short, the tool is a ten-minute decision and almost any of them will do.
Do OKRs have to cascade to every team?
No, and cascading is where most rollouts die. Copying objectives down the org chart produces alignment on paper and nothing on Tuesday. I'd rather have company objectives that a handful of teams genuinely own than a full tree that nobody reads.
Should OKRs be tied to bonuses?
It depends what kind of number it is. A sales quota is a committed target, everyone treats it as one, and paying against it works fine. A stretch objective is a different instrument: the moment an ambitious number decides someone's pay, you have asked them to negotiate it down, and they will. The common mistake is taking compensation logic that works in sales and applying it to stretch OKRs across the rest of the business.
How often should we review them?
Fortnightly at minimum, and the review has to be able to change something. A meeting where everyone reports green and nothing gets stopped is a status meeting wearing an OKR badge. If nothing can be re-scoped in the room, move the meeting or cancel it.
Who should own a key result?
One named person, every time. Not a team, not a function, not the leadership group. Shared accountability reliably becomes no accountability, and the first honest test of a key result is whether anyone will put their name against it.
Do we need an internal OKR champion?
Yes, and it should not be me. Someone inside the business has to own the cadence once I've gone, or it decays within two quarters. Identifying and coaching that person is part of the work, not an afterthought.
We tried OKRs and people hated them. Is it worth another go?
Often, yes, but not immediately and not the same way. A failed rollout leaves scar tissue, and relaunching on top of it usually fails faster than the first attempt. The Execution Diagnostic exists partly to answer this question honestly, including when the answer is no.
Tell me where your OKRs are stuck.
Thirty minutes, no pitch. If OKRs aren't your actual problem, I'd rather tell you that early.