Remove constraints, don’t add initiatives.

There's a reflex in almost every growing company, and once you notice it you'll see it everywhere. A goal starts to slip, and the response is to add something. A new initiative, a new hire, a new tool, a new weekly meeting to keep an eye on it. Revenue's behind, so we launch a partnerships programme. Churn's up, so we start a customer success project. Delivery's slow, so we bring in a new project management system.

Each of those feels like action, and each one gets a kick-off, an owner and a slide. And quite often, six months later, the goal is still slipping, and now the team is also running three more initiatives than it was before.

I think this reflex is one of the most expensive habits in business, and the third belief behind Clearahead is the antidote: remove constraints, don't add initiatives.

What Goldratt saw

In 1984 an Israeli physicist called Eliyahu Goldratt published a business novel called The Goal. It's about a factory manager trying to save his plant, and it's still one of the most useful management books ever written, partly because it's a story rather than a textbook. The idea at its heart, which became known as the Theory of Constraints, is very simple.

Any system, a factory, a sales team, a whole company, is held back by a small number of constraints at any one time. Speed up anything that isn't the constraint and the system as a whole doesn't go any faster; the work just piles up in front of the bottleneck. Improve the constraint, and the whole system moves. Then a new constraint appears somewhere else, and you go again.

Goldratt was talking about machines on a factory floor, but the principle holds for any company I've ever seen. Growth isn't held back by a general lack of effort. It's held back by a few specific things, and until those things are cleared, every bit of extra effort elsewhere is largely wasted.

Why initiatives feel right and rarely work

An initiative is a bet that the problem is a lack of activity. More outreach, more features, more process. Sometimes that's true. Mostly it isn't.

When revenue's behind, the real constraint might be that the sales team can't get prospects past the first call because nobody can explain the product in a sentence. A partnerships programme doesn't touch that; it just feeds more prospects into the same leaky first call. When churn's up, the constraint might be that onboarding takes three weeks and customers give up in week two. A customer success project with monthly check-ins doesn't touch that either.

Initiatives also have a nasty habit of never ending. Nobody ever formally stops one; they just slowly fade while still taking up a slot in somebody's week and a line on the dashboard. Add a few a year and before long a team of thirty is running fifteen of them, each one competing for the same handful of people. That's how you get a company where everyone is flat out and nothing is moving, which I wrote about in Alignment beats activity.

Most to-do lists are symptoms

Here's the shift in thinking that I find makes the biggest difference. If you sit a leadership team down and ask them to list everything that's wrong, you'll get a long list. The website's out of date. Leads are low quality. The product's missing a feature. We need another developer. Hand-offs between sales and delivery are messy. Nobody updates the CRM.

Almost all of those are symptoms. Real, painful, worth knowing about, but symptoms. Underneath them there are usually only one or two causes. Low-quality leads, a stale website and a messy hand-off might all trace back to one thing: the company has never agreed exactly who it's for. Fix that, and three items on the list get easier at once. Work on each symptom separately, and you'll be busy for a year and not much better off.

So the first job isn't to make a plan. It's to ask one plain question, and keep asking it until you get past the symptoms: what's holding us back? Not "what should we do?", which invites a list of initiatives, but what's actually in the way of the goal, stated as something the team can do something about.

How to find your constraints

A few things I've found help a team get from symptoms to causes:

• Start from the goal, not the problems. A constraint only means something in relation to what you're trying to achieve. Name the goal first, with a rough when, then ask what's stopping you getting there.
• Keep asking why. When someone says "we need more leads", ask why the leads you've got aren't enough. Then ask why again. Usually by the third answer you've found something that a new initiative wouldn't have touched.
• Write down why you believe it. Every constraint should come with its evidence and how sure you are: a hunch, fairly sure, sure. A constraint the team believes but can't back up is still worth working on, but you should know you're working on a hunch.
• Say how you'll know it's cleared. If you can't describe what would be true when the constraint is gone, it's probably still a symptom or a vague worry.
• Put them in order. Not a score, just an order. The one at the top is what most needs clearing now. I'd suggest three to five active at a time; more than that and you're back to running initiatives.

Then, and only then, decide what to do. Every action, or move as we call it, should have one owner and a due date, and should say which constraint it eases. If a move doesn't ease a constraint, it can still happen, but everyone can see it's other work rather than work on the goal.

Clear it, then find the next

The best part of working this way is what happens when a constraint clears. You record what cleared it, which becomes part of the company's memory, and the next constraint moves up. There's no big reset, no new initiative launch, no planning day. The team simply asks the question again: now that's gone, what's holding us back?

That's the loop at the heart of the operating model for the AI era, and it's what Clearahead is built to run. When you set a goal, AJ asks what's holding you back and pushes past the symptoms to the causes; you can try that conversation for yourself on the Meet AJ page. You come away with a short list of constraints in your own order, and moves under the top one. As people update their moves, AJ keeps the evidence, notices when a constraint looks cleared and remembers how you cleared it, so the next time something similar comes up, it can tell you what worked. There's a full guide to constraints in our docs.

Stop something this month

If this lands, here's where I'd start. List every initiative your company is running right now. Next to each one, write the constraint it's clearing. Any initiative where you can't name the constraint, or where the constraint isn't in your top three, is a candidate to stop.

Stopping things is uncomfortable. Someone championed each of them, and someone's going to feel it. But every initiative you stop gives time back to the few things that are actually holding the goal back, and in my experience that's the fastest way any company has of speeding up. Not doing more. Doing less, aimed better.

It's also why we think the job of an operating model in the AI era isn't to help you plan more. It's to help you see what's in the way, and clear it, again and again. The next belief, The machine asks, people decide, is about who should do what in that loop.