Digitising without breaking anything

Digitising is where the younger family member most often overreaches, because the tools are cheap and the current system is visibly held together with a register and a memory. The failure is almost never the software. It is switching several things at once, in a busy month, without anybody but you knowing how the new thing works.

Run it as a sequence, and only after the process is mapped, because putting software on top of a process nobody has understood produces a faster version of the same confusion. Take the smallest piece first — digital payment records, or billing, or stock — and run it in parallel with the old method for a fortnight, deliberately doing both. Parallel running feels wasteful and it is the entire safety mechanism: when the new thing is wrong, and it will be in week one, the business does not stop.

Three questions before adopting anything. Who else can operate it if you are away for a week — and can they, demonstrably, not in principle? What happens on a day the internet or the power is out, because there will be one? And can you get your own data out if you leave the tool, which is the question nobody asks until they want to leave. Then write the one-page instruction before you go live, not after, and have somebody else follow it while you watch and say nothing.