The arithmetic is different
At two hundred staff, saving four minutes each is a real number. At ten, it is nothing you can redeploy.
So the target is concentrated work: the person who spends six hours a week on one repetitive thing.
What usually qualifies
- Whoever handles the shared inbox
- Whoever keys documents into the accounting or job system
- Whoever writes up notes after calls or visits
- Whoever assembles quotes or reports from templates
Ask who in the business would be most relieved by a specific task disappearing. That is usually the right project and they are usually right about it.
Buy before building
At small scale, subscriptions usually win on total cost, and the integration between products is where a small build earns its money.
Buy the capability, build the connection. That is a few thousand pounds rather than tens of thousands.
Keep the ownership realistic
- One named owner, with a few hours a month
- A system simple enough that they can understand its numbers
- Alerts routed to a person, not a channel
- A supplier arrangement that does not depend on urgency
Complexity you cannot maintain is worse than a smaller thing you can.
What to leave alone for now
| Not yet | Why |
|---|---|
| Forecasting | Not enough history or volume |
| Recommendation engines | Not enough behavioural data |
| Multi-system orchestration | Maintenance burden outweighs benefit |
| Custom model work | Never justified at this scale |