Stuck teams almost never lack effort — they lack a named bottleneck, and the misdiagnosis ("we need better people / more tools / more hours") is what keeps them stuck. Six questions locate where work actually stalls: where decisions wait, where handoffs blur, what gets discussed twice, what only the owner can touch, where done goes fuzzy, and where AI runs without an owner. In founder-led businesses the honest answer to several of them is usually the founder.
The Misdiagnosis
When a small business feels stuck, the explanations on offer are always the same. We need better people. We need better tools. We just need to get through this busy stretch. Each of these has the ring of a diagnosis, and each conveniently locates the problem somewhere that doesn't require changing how the business is actually structured.
But watch a stuck team closely and you almost never find a shortage of effort. You find effort pouring into a system that loses it — work that blurs between people, decisions that wait in someone's head, the same ground covered in meeting after meeting. The team is rowing hard. The boat is taking on water through joints nobody has inspected.
A bottleneck, properly understood, is not the busiest person or the most annoying process. It is the specific point where work accumulates faster than it flows through — and everything downstream of it starves no matter how hard anyone works. Every operation has one. Most operations have never named theirs.
In founder-led businesses there is an additional, awkward fact: the bottleneck is very often the founder. Not because founders are incompetent — because the business grew around them. Every path was originally routed through the person who could do everything, and nobody re-routed the paths when the volume outgrew the person.
The questions below are the ones that surface it. They work best answered in writing, honestly, with specific recent examples rather than general impressions. Vague answers are themselves data.
The Six Questions
1. Where do decisions wait?
Take the last ten meaningful decisions in your business — pricing a job, hiring, approving an exception, greenlighting a purchase. For each: how long did the decision take once the information was available? Not the analysis. The waiting.
Decisions that wait are the purest form of bottleneck, because everything attached to them waits too. And decisions usually wait in one of two places: in a queue in front of one person, or in ambiguity, where nobody is sure they're allowed to make the call. Both are structural. Neither is fixed by working harder.
2. Where does work change hands — and what happens to it there?
Follow one recent piece of work — a customer request, a project, an estimate — from entry to completion, and mark every handoff. At each one: did the receiving person know it was theirs? Did they know what done looks like? Did anything have to be re-explained, re-formatted, or chased?
Handoffs are where operations quietly bleed. Work that moves cleanly between people needs a shared definition of what's being handed over; work that doesn't gets absorbed — it sits in the gap between two people, each assuming it's the other's, until a customer asks about it.
3. What gets discussed more than once?
Pull up your last month of meetings and find the repeat topics — the initiative that gets 'aligned on' every week, the policy question that keeps resurfacing, the project whose status is always 'in progress.'
A topic that gets discussed twice without changing is not being discussed. It's being deferred with ceremony. Recurring discussion is the sound of a missing decision right: no one in the room is sure they own the call, so the group re-litigates instead. The cost isn't the meeting time — it's everything that stayed frozen between the meetings.
4. What can only you touch?
List everything in the business that stops moving when you're unavailable for three days. Quotes? Approvals? Client communication? Payroll? The answer to a specific technical question that lives only in your head?
This is the founder-bottleneck question, and the honest version of the list is usually longer than the flattering version. Some of it is genuinely yours to hold. Most of it is routing that made sense at half your current size and was never revisited. The tell is emotional: the items you'd defend hardest as 'faster if I just do it' are usually the ones most worth moving.
5. Where does 'done' go blurry?
Find the work in your operation that is chronically ninety percent finished — the proposal that's 'basically ready,' the process that's 'mostly rolled out,' the hire that's 'nearly onboarded.'
Work stalls at ninety percent when nobody defined the last ten — when done means 'when it feels done' rather than a state someone can verify. Blurry completion is expensive twice: the work delivers no value while it hovers, and it occupies attention the whole time. An operation with crisp definitions of done feels lighter at the same workload.
6. Where is AI running without an owner?
The newest version of an old problem. List where AI is producing work in your business — drafts, summaries, replies, research — and for each, name the person who stands behind the output as if they'd made it themselves.
Every blank on that list is a soft spot. AI-produced work without a named owner reads finished while committing no one, and it quietly converts your operation's drift into confident-sounding prose. The bottleneck questions above get harder to answer in an operation where the paper trail is fluent and accountable to nobody.
Reading Your Answers
Patterns matter more than any single answer.
If decisions wait mostly on you, and the only-you list is long, the constraint is routing: the business still runs on paths drawn when you were half this size. The fix is not delegation as a virtue — it's explicitly moving specific decision rights, with the authority to be wrong, one category at a time.
If handoffs blur and done is fuzzy, the constraint is definition: work isn't specified crisply enough to move between people without losing shape. The fix is unglamorous — owners on recurring work, entry and exit criteria on the handoffs that bleed the most.
If topics repeat and meetings multiply, the constraint is decision rights: your operating rhythm has plenty of ceremony and no mechanism for closure. The fix is deciding who decides, then letting the meetings shrink to fit.
And if the AI column is full of blanks, resist the urge to ban or evangelize the tools. The tools are revealing a structure problem that predates them. Fix the ownership, and the tools become genuinely useful; skip it, and they become a fog machine.
Naming It Is the Hard Part
None of the fixes above are exotic. Decision rights, clean handoffs, crisp definitions of done, named owners — every operator has heard all of it. The hard part is never the fix. It's the naming: seeing your own operation clearly enough to say 'this specific point is where our work goes to wait' — especially when the honest answer implicates the person doing the naming.
That's the case for an outside read, and it's the reason our work starts with one — a bounded, fixed-price diagnostic rather than a proposal. Not because the questions are secret; they're printed above. Because the founder answering them alone is grading their own homework, on the precise subject where self-grading fails.
But start with the questions. An hour of honest written answers will tell you more about why your business feels heavy than a quarter of pushing harder. Effort was never the missing ingredient. A named bottleneck, and the nerve to fix that one first — that's the whole game.
FROM NOTE TO EXECUTION
If this issue is active in your organization, we can pressure-test it together.
The article is there to sharpen the model. The engagement is there to change the operating reality.