For most owners the highest available return is inside the business they already run. They understand it, they can influence the outcome, and the returns are frequently well above anything available externally.
That is also why it is easy to over-commit, and why the discipline of asking for a return matters more here than anywhere else.
Make it a case, not a feeling
Before spending, write down what you expect: what changes, by how much, over what period, and how you will know. Two or three sentences.
This is not bureaucracy for its own sake. Most disappointing business spending is disappointing because nobody ever said what it was supposed to achieve, so nobody could tell that it did not.
The four things worth spending on
Capacity, where you are turning work away or delivering slowly. The return is measurable and immediate — but only where demand genuinely exceeds what you can produce, rather than where you would like it to.
People, where a role would generate more than it costs. The most common high-return investment and the one owners defer longest, usually because the cost is monthly and certain while the return is gradual.
Marketing you have evidence for. A channel that has produced measurable enquiries at a known cost is an investment. An untested one is an experiment, and should be sized as one.
Systems that remove hours. Frequently the best return of all, because the hours freed are usually the owner's, and the owner's hours are the business's scarcest resource.
Start small and measure
Where something is unproven, spend an amount you can lose and find out. A small test that fails costs little and teaches you something; a large commitment on an untested assumption is how businesses acquire fixed costs they cannot carry.
Where it works, scale it deliberately rather than all at once.
Keep the buffer intact
The most common way this goes wrong is spending the reserve on a genuinely good opportunity and then meeting an ordinary bad month with nothing behind you.
Invest surplus. Do not invest the buffer, however compelling the case — the case is always compelling, which is exactly the problem.
Beware the equipment that looks like progress
Premises, vehicles and machinery feel like a growing business and are frequently fixed costs acquired ahead of the revenue to support them.
The test is whether the work exists now. Buying capacity in anticipation of demand you hope to create is a different and much riskier proposition than buying it to serve demand you are already turning away.
Review what you spent
Once a year, look at the significant spending of the last twelve months against what you said it would do. Some will have worked and some will not, and knowing which is what makes the next year's decisions better.
General guidance, not financial advice.