What's actually going on
This is one of the most common snags at this stage of a business, and it's fixable without buying anything or becoming a CRM expert. "Losing track of who I promised what" is a memory problem and a review problem, not a data-volume problem. Five to ten clients is well within what a person can hold in their head on a good day — the trouble is that nothing forces you to check it on a bad one. A promise made on a call, an email you meant to follow up, a small thing you said you'd send "this week" — each one is fine alone. What breaks is that they all live in different places: your memory, an inbox, a text thread, a notebook. Nothing collects them, and nothing makes you look at the collection on a schedule.
A CRM would store all of that in one place. But a spreadsheet, a notes doc, or a plain list stores it in one place too. The part a CRM can't give you — the part that's actually missing — is the habit of checking it. That's worth naming honestly before spending money on the wrong half of the problem.
The smallest fix you can do today
Start one list, in whatever you already open every day, with one row per open promise rather than one row per client: who it's for, what you said you'd do, the date you said it, and what the next action is. Nothing fancier than that.
Then pick a fixed time — Monday morning, first thing, before anything else — and go down the whole list. Update anything that's moved, close anything that's done, flag anything overdue. Ten minutes, same time every week, no exceptions.
Most of the "losing track" feeling disappears once there's one list and one weekly look, because the problem was never really about storage — it was about nothing being reviewed on a schedule.
When it's worth keeping, cleaning, connecting, or converting
Keep the plain list if, once you're actually doing the weekly review, clients stop chasing you for things you forgot. For most people at this stage, that's the whole fix — no further tool is needed.
Clean it if the list exists but is scattered — half in email, half in your head, a notebook that's usually in the wrong bag — or if it exists but nobody's actually running the weekly review. The fix is consolidating into the one list and making the review non-negotiable, not new software.
Connect it if the promise list keeps needing details you already record somewhere else — a booking form, an invoice, a quote — and you're retyping the same client details twice. Pull from the existing record rather than duplicating it by hand.
Convert to a CRM only once the list itself is the bottleneck: several people need to see and update the same client's history at once, or you regularly need to search years of past promises to one client faster than a list lets you. That point is usually well past the client count you're at now, not at it — buying a database before the habit exists just gives the same gap a login screen.
What "done" looks like
Done means you can answer "what did I promise this client, and when is it due" in well under a minute, without checking your inbox or asking them. You'd know it's working if the "sorry, I completely forgot" messages you send drop close to zero, and if you could hand the list to someone else and they could pick up your open promises cold. If that's still not true after a month of doing the weekly review properly, the problem wasn't the list — that's the point where it's worth actually looking at a CRM, not before.