There has never been more help on offer for running a small business. Whatever the job — quoting, scheduling, chasing invoices, writing the marketing — there’s software built for it, and now an AI that will draft it before you’ve finished describing what you want. When I started my agency in 2001, finding out how other firms handled a problem meant asking around, or buying a book and hoping. That scarcity has gone. Information is abundant, tooling is abundant, and yet most of us are still carrying a queue of decisions that none of it has made any shorter.

That queue points at what the scarce resource has become. It isn’t knowledge any more, and it isn’t capability; it’s execution discipline: doing the right things, in the right order, with what you already have. The order fits on an index card — stabilise the process before you buy the tool, and prove the tool before you add the person — and most owners have had that playbook on the shelf for years. Owning it was never the constraint.

I’m partial to this argument because I lived the untidy version of it. Seventeen years running an agency on Gerber’s principles (build the system, document the work, stay on the business rather than in it) and they worked well enough that the business survived my leaving. Sequence was still the part I got wrong most often. Buying the tool is the most enjoyable job in the building; fixing the process the tool is meant to serve is the least. Given a spare afternoon in a heavy week, I know which one I picked more often than I should admit. The classic version is the CRM bought to fix follow-up while the follow-up process itself was still a set of habits in one person’s head — the software gets the blame when it’s abandoned a quarter later, but the order was the problem.

Start the order with cash

Cash goes at the front, and I’d push it harder than most of the advice does. Discipline is far easier with money in the bank. The operator who calmly takes things in sequence can usually afford to wait; at three weeks of runway, the correct next step loses to the immediately profitable one nearly every time, and it probably should. So when “be more disciplined” gets aimed at a business that’s already tight, I’m suspicious of the advice rather than the business. If cash is what stands between you and the ability to sequence at all, then the buffer is the first process to stabilise, because it buys the right to run the rest of the order.

The banking research on the small-business execution gap arrives at the same place from the data. Its three rules (stabilise cash flow before increasing volume, improve systems before adding complexity, strengthen retention before scaling acquisition) are one rule applied three times: earn the next move before you make it.

The six-month test

To find out whether the order is holding, ask what would happen if you disappeared for six months. Would customers get the same experience? Would decisions keep being made, or would they queue behind your absence? Dennis Kuipers, who scaled his own company across 35 countries before exiting, uses that question to test whether a founder has let go, and he’s honest about failing it for most of his journey: more customers meant more oversight, more staff meant more approvals, and growth stayed chained to his diary.

The distinction underneath the test is between delegating activities and transferring ownership of outcomes. Most of us do the first and call it the second: the work goes out, the approval stays home. Gerber told a generation of us to build the system and step back, and the stepping back is the harder half of the instruction. Selling taught me that more thoroughly than running ever did. The buyer was purchasing a business that had to keep delivering after I’d left it, and the two-year handover was largely a tour of outcomes I still owned without realising. I can’t tell you when the balance finally tipped, either; there was no week in which the approvals stopped being mine, which is partly why I rate a blunt test you can ask yourself on any ordinary morning. Passing it comes down to what exists while you’re present: someone else holding a decision outright, and a written answer to “what does good look like” that works when you’re not there to be asked.

What the discipline is worth

Aaron Leibtag’s exit puts a price on it. His company, Pentavere Research Group, employed fifteen people; its software digs clinical information out of the free text and PDFs where hospital records bury it, and it did that whether or not a founder was in the room. HealWELL AI acquired the business at a valuation of around $15 million (roughly a million dollars per employee) and, as Leibtag tells it, the offer was built on where the company could go under owners who weren’t him. What a buyer prices in that situation is delivery that repeats and decisions that don’t all route through one person’s diary. EINEdge’s coverage of founder discipline says the same thing from the funding side: what gets backed is what already runs.

You don’t have to be selling for any of that to matter. A business that passes the six-month test is a better place to spend your week long before anyone offers money for it.

Three things worth doing this week

All three fit in an afternoon, and none of them costs anything.

  1. Write the queue down. List every decision currently waiting on you — approvals, sign-offs, choices between suppliers or quotes — on a single page. I’ve suggested auditing your whole week before; this is the narrower, faster version. The list is usually longer than expected, and a few entries will have waited so long that the delay has become the decision.
  2. Sequence by who it unblocks. The five-minute approval four people are waiting on beats the two-hour job nobody is waiting on. It’s the same instinct as scaling through leverage rather than headcount: find where a small amount of your attention releases the most capacity elsewhere.
  3. Hand over one decision, properly. Pick a category — spend under £1,000, quotes below a threshold, holiday approvals — and give someone the outcome to own, with a written note on what good looks like and an agreement on how you’ll both find out if it goes wrong. Some of their calls will be worse than yours; accept that going in, because the queue behind you shrinks either way.

Start with whichever of the three you’ve been avoiding longest, give it a fortnight, and let the next tool on the shortlist wait until the process it’s meant to serve has earned it.