Today is day 2,132 of my daily barefoot-style running streak. Today's run was a grind. I would have happily handed it to Claude Code. I can't, so I ran, and I used the miles to think about what has changed in my business over the last few months.

I started using Claude as a chat tool. I asked questions and got answers. Then I gave it basic tasks. Then I gave it larger tasks, and then I let it run some of them on a schedule without me. Now I use Claude Code and Claude Cowork to build bespoke solutions to problems that used to sit on someone's task list for weeks.

The latest step is the one I want to write about. I have introduced Claude Code to my team.

What happened

I run a small back office of four people, based in the Philippines. Over the last few weeks they have started building their own projects with Claude Code. One of them came to me this week with something to review.

We already had a content process for Runpreneur, the daily vlog behind this mission. It was sophisticated, but it still needed people. She rebuilt it from scratch with Claude and connected the tools together. The flow now works like this. I record the video and save it into a Google Drive folder. The software she built picks it up, edits it, adds captions, makes thumbnails, writes the descriptions and publishes in the right format on each platform.

It is not finished. There are tweaks to make and issues to fix. But if it works as designed, it replaces work that used to take a share of two or three people's week, roughly half of their time. I am careful with that number. It is my estimate of the time spent, not a measured figure.

Why a second person's ideas matter

When I use Claude Code on my own, it gets used to me. It learns my habits, my phrasing and my way of framing problems. That is useful, but it also narrows the output. When someone else sits down with the same tool and brings different ideas, the results change. She did not build what I would have built. She built something better suited to the work she knew from the inside.

That is the main lesson for me. Giving the tool to your team does not only add more hands. It adds more perspectives pointed at the same tool.

The cost is real

I will be straight about the money. Free usage runs out quickly. The first paid tier gives access to Claude Cowork and Claude Code, but those tools use more of your allowance, so you hit the cap even sooner. I am on the plan at around £100 a month. It covers nearly everything I need, but scheduled tasks and my team's projects are using the allowance faster than I expected.

My plan is to move up a tier once I have replaced other software subscriptions with what we are building. The saving pays for the upgrade. The higher tier gives me about 20 times the usage of what I have now, which would cover what we are doing.

The cost is not the point, though. The gain in output is far larger than the monthly fee.

A practical tip on usage limits

My team works in the Philippines, so we flipped our hours. While I sleep, they use the allowance. During my working day, it is mine. Usage sessions cap out in roughly four-hour blocks, so understanding that rhythm matters. Nothing is more frustrating than making real progress and hitting the limit halfway through. Until we upgrade, a simple rota solves most of the problem.

How I think about my team

I describe myself as an ethical entrepreneur. I am not using this to cut my team. I am training them to build systems for different parts of the business, then to look for new opportunities. The aim is to move people away from fixed, repetitive daily work and towards creating something new. I value what they do, and I want them to keep adding value in roles that get better as the tools improve.

My estimate is that Claude Code and Cowork multiply my own productivity many times over. I would put it around twenty times, though that is a judgement, not a measurement. If four other people reach a similar level, the combined effect on what we solve across the business is hard to overstate. I would rather say I do not know the exact figure than pretend I do.

The lesson I took from today

If you run a business and you are not yet experimenting with tools like Claude Code and Claude Cowork, you are falling behind. There is a financial cost, but I think the cost of waiting is higher. Once you have started, bring your team in early. Let them build. Review what they bring you. You will probably be surprised by the outputs, as I was.

The link to the run

A daily streak and a team-built system have something in common. Both compound through repetition. I do not run fast. I run every day, and the distance builds. Today's run took me to 16,383.42km in total, which leaves 23,691.58km to complete a lap of the world at 40,075km. A good system works the same way. You improve it a little at a time and it keeps working while you sleep.

Why this matters beyond the business

Every efficiency we find gives me more time and energy for the mission. This vlog is part of a fundraiser to save the lives of children, with a goal of raising £1M for children's causes, including Great Ormond Street and BBC Children in Need. The more people see these episodes, the more we raise. A content process that runs with less manual effort helps more people find the vlog, and that is the practical link between a team building with Claude Code and the children this run is for.

What I would like you to do

If you have a team, ask one person to pick a single repetitive task this week and try to build a solution with Claude Code. Review what they bring back. If you want to follow the mission, watch the daily vlog, subscribe, share an episode or leave a comment. Each one puts the run in front of more people, and that is how we raise more for children's causes.

I will be back out tomorrow for day 2,133.

Watch the full episode: https://youtu.be/PArp_OEwTaU