Practical thinking for business owners.

Most poor decisions are not caused by bad information. They come from starting to compare options before agreeing on what is being decided. Before any major decision, such as hiring, investing, opening a location or changing an offer, we suggest answering seven questions in writing.

• What are you building toward?
• Over what horizon?
• What can you not afford to lose?
• Who decides, and how?
• What must be true for this to work?
• What would make you change course?
• What will you do on Monday?

The first three set the frame: the goal, the time you have and the limit of what you can risk. The next two make the decision accountable and expose the assumptions. The last two turn it into action and keep it honest.

Our point of view is simple: a decision written on one page is better than a perfect analysis nobody finishes. If you cannot answer one of the questions, that is where to spend your next hour.

You do not need a transformation programme to benefit from AI. Five well-chosen automations cover most of what a small business needs first.

• Replying to routine customer questions, with hand-off to a person for complex cases.
• Drafting quotes and proposals from incoming requests, reviewed by a person before sending.
• Producing a monthly management report from your accounting and sales data.
• Filing and classifying invoices, contracts and supplier emails.
• Summarising meetings and tracking the actions that follow.

Before choosing, ask three questions. How many hours does this task take each week? What happens if the tool makes a mistake? Who reviews the result? Start with tasks where a mistake is cheap and a person can check the output.

Our point of view: the best first project is the most boring one. It saves time quickly, it is easy to measure and it builds your team's confidence for what comes next.

When a customer offers to pay in crypto, or a friend mentions stablecoins, the discussion often jumps straight to which asset. That is the wrong first question. This article is general information, not a recommendation about any asset.

Start with the purpose. Is it to accept payments, to hold a treasury reserve, to pay suppliers abroad or simply to learn? The answer changes everything that follows.

Then consider the categories of risk that apply to any digital asset use:

• Market risk: the value can change quickly.
• Custody risk: who holds the keys, and what happens if access is lost.
• Counterparty risk: the platforms and providers you depend on.
• Regulatory and tax risk: rules differ by country and change over time.
• Accounting risk: how the activity is recorded and audited.

Finally, write a short policy: what is allowed, up to what amount, who has access and who reviews it. Then speak to your accountant and a regulated provider.

Our point of view: a business can say no, yes or not yet. What it should not do is decide without a written reason.

Structure is not bureaucracy. It is what lets a company grow without everything depending on one person. Three signs tell you it is time to add some.

• Decisions wait for one person. Employees ask before acting, and the owner is the bottleneck for small matters.
• Knowledge lives in heads. When someone is away, work stops, and nobody can say how a task is really done.
• New people take months to become productive, because there is no simple description of how things work here.

If you recognise one of these, do not start with a new organisation chart. Start by writing down who decides what, for the five or six decisions that come up most often.

Our point of view: add structure only where a problem repeats. The lightest rule that removes a real problem beats any grand reorganisation.

Succession problems rarely come from the plan. They come from the conversations that never happened. Before involving advisers, write down answers to five questions. This article is general information, not legal or tax advice.

• What does continuity mean for us: ownership, management, values, or all three?
• Who will do what, and by when?
• How will disagreements be settled?
• What does the founder keep, and what does the founder let go?
• What happens if the plan does not work?

Writing the answers separately, then comparing them, often reveals differences that were never spoken. That is the point: better to find them on paper than in a crisis.

Our point of view: succession is a project with a calendar, not a moment. The earlier the questions are written, the calmer the answers.

A good AI project starts as a clear sentence, not as a tool. Five decisions are enough to scope a first one.

• Pick one task. Not a department, one task, such as drafting quotes or answering routine customer questions.
• Define what good looks like: the time saved, the error rate you accept, the response time you want.
• Decide what data the tool may see, and what it must never see.
• Decide who checks the result, and how often.
• Pilot on a small part of the work for a few weeks before extending.

If you cannot write the task in one sentence, it is too big. Split it.

Our point of view: scope small, measure honestly and extend only what works. Technology is the easy part. Clear ownership is the hard part.