Choosing accounting software is the last decision, not the first

Ask an owner or an executive director what they need to fix about their finances and the answer is often a software name. Either the current one is wrong, or a better one exists, or a competitor uses something else.

Sometimes that is true. More often the software is the most visible part of a problem that is not about software at all, and switching moves the problem rather than solving it.

Why it feels like the first decision

Because it is the only part of the system that is concrete. It has a name, a price, a website, and reviews. The other decisions are abstract, involve tradeoffs with no obvious right answer, and nobody sells them to you.

So the concrete decision absorbs the energy that the abstract ones needed, and a few months later the same problems reappear in a new interface.

What to settle first

Four things, roughly in this order. None require software to answer.

What you need to be able to answer. Which questions must the books answer, and how quickly. An organization tracking several restricted grants has a hard structural requirement. A business that needs margin by service has a different one. This determines the shape of everything else.

How your accounts are organized. A chart of accounts that mirrors how you actually think about the work makes the reports readable. One inherited from a template makes every report a translation exercise, in any software.

Who does what, and when. Who codes transactions, who reviews, who reconciles, who closes the month, by which date. Most books that fall behind fall behind here, not in the tool.

Where the documents live. One place receipts, contracts, and approvals go. This determines whether a question from an auditor or a funder takes ten minutes or two weeks.

What software actually decides

Less than the marketing suggests, but not nothing. It decides how much manual entry you do, what your bank connections and integrations look like, whether multiple people can work without stepping on each other, what happens as volume grows, and how easily you can get your data out later.

That last one is worth weighing more than it usually is. Every system eventually gets left.

Migration costs more than you think

Moving systems is not a settings change. Historical data has to come across or be deliberately left behind. Account mappings rarely line up cleanly. The first months in a new system produce reports that do not compare to the old ones, right when you most want to check that things are working.

It is usually a full quarter before the new system is producing numbers you trust as much as the old ones. That cost is worth paying for a real constraint and not worth paying for a better interface.

When switching is the right call

There are genuine reasons, and they share a shape: the current system cannot do something you need, rather than does it awkwardly.

• It cannot track a dimension you are required to report on, such as restricted funds by grant.
• It has become the bottleneck as volume grew, in a way that is measurable in hours.
• It cannot support the number of people who now need access, or the separation of duties you need between them.
• It is being discontinued, or has stopped being supported in your region.

If the reason is not on a list like that, the switch will probably be satisfying for about six weeks.

The order that works

Decide what you need to know. Structure the accounts so the books can tell you. Assign the work and the dates. Put the documents somewhere. Then pick the tool that fits what you have decided, which by that point is usually an easy choice and occasionally turns out to be the one you already have.