Office automation is an old phrase for a simple idea: taking the repetitive administrative work of an office and letting software do the repeating. In a large company it means a suite. In a small company office, where one person is the whole administration, it means something much more specific and much more useful: picking the handful of jobs that happen often enough, and the same way often enough, that software can carry them, and deliberately leaving the rest alone. This page is written for the person doing that picking. It sets out what the term covers, the three tests a job has to pass before automating it pays for itself, and the jobs that look automatable and are not.
What office automation actually covers
Four families, and most small offices only ever touch three. Document handling: scanning what arrives on paper, filing it somewhere it can be found, and getting rid of it on a schedule. Records: the suppliers, the contracts and their renewal dates, the equipment and who holds it, kept somewhere other than one person's memory. Routing and approval: an expense, a purchase or a request that needs somebody to say yes before it goes further. And communication: reminders that go out on a date without anyone remembering to send them. The fourth family, workflow between departments, is what the enterprise products are sold on and what a twenty person office almost never needs.
The three tests a job has to pass
Frequency first: a job done twice a year is not worth automating however tedious it is, because the setup will cost more than the job. Then repeatability: the job has to happen the same way each time, because a rule that has exceptions half the time becomes a rule somebody has to check, which is the job you were trying to remove. Then judgement: if the step requires someone to decide something that is not written down anywhere, software cannot carry it and pretending otherwise produces confidently wrong output. A job that passes all three, filing, reminding, routing a fixed approval, is worth the setup. A job that fails any one of them is worth leaving as it is.
What looks automatable and is not
The office move, because every one is different and the value is in the thinking rather than the sequence. Supplier negotiation, because the software can tell you the renewal is in six weeks and cannot tell you what to say. Anything that turns on retention rules, because how long a given record must be kept and in what form depends on what it is and where you trade, and the Internal Revenue Service's own guidance is that the period varies by the document and the situation rather than being one number. And anything you have not done manually enough times to know how it actually goes. Automate the fourth version of a process, never the first.
Questions people ask about office automation
Is office automation the same as going paperless?
Going paperless is one part of it, the document handling family, and it is usually the part with the clearest payback because the cost of the paper can be counted. The free worksheet on this site counts it: documents a month, the minutes each takes to handle and to find again, and the floor the cabinets stand on. The other families, records, routing and reminders, have real value too and are much harder to put a number on.
Do we need one system or several?
Several is the normal answer for a small office and there is nothing wrong with it. The thing to watch is the seams: work is lost between tools, not inside them. If a renewal date lives in one place and the supplier lives in another, somebody has to keep the two in step, and that reconciliation is a job you have created rather than automated.
What does office automation cost a small office?
The subscriptions are rarely the expensive part. The expensive part is the week somebody spends getting the existing mess into the new thing, and the month afterwards when both the old way and the new way are running. Budget for that time explicitly, and automate one family at a time so the disruption is bounded.