ERP for $5–25M businesses

The situation

What actually changes things at this size.

Not revenue. Four specific conditions, and you need at least two of them before an ERP conversation is warranted.

A second or third entity

The single strongest signal. Consolidation in a spreadsheet is fine at two entities and stops being fine at four, and the workbook becomes infrastructure nobody else can run.

More than 200 bills a month

Around this volume the manual work justifies a hire. That is the point at which automating it is cheaper than staffing it, and it is a workflow fix rather than an ERP.

Contracts got complicated

Multi-element deals, ramps, or usage billing mean ASC 606 stops being a formality. For software companies this is usually the forcing event.

Somebody external is coming

A first audit, a lender covenant, a growth round, or an acquisition approach. Internal adequacy and external scrutiny are different standards.

One person holds the close

Not a size issue but a risk one. If nobody but your controller could produce the statements, the system is a person and that is worth fixing regardless of revenue.

Reporting rebuilt monthly

Department or project P&L reconstructed in Excel every month means dimensions were never captured — fixable at entry, not in the report.

The two fixes that cost a fraction

Most of what companies at this size want from an ERP is one of two things, and neither requires replacing the ledger.

Automate accounts payable. If the pain is keying volume, that is a workflow problem. Bills read, coded, matched, and approved on top of QuickBooks or Xero, writing finished bills back.
Build reporting above the ledger. If the pain is department, project, or entity reporting, that is a dimensions problem. Where the data is recoverable we produce the reporting from a graph reading your existing system.

Both are live in weeks, both are reversible, and together they cost a small fraction of an implementation. If they resolve the pain — and at this size they frequently do — the ledger conversation postpones for a year or two, which is a better outcome than an unnecessary project.

Revenue is the worst predictor of whether you need an ERP. Entity count, close duration, and bill volume predict it far better, and none of them correlate reliably with turnover.

What we will tell you to do instead

Roughly a third of assessments at this size conclude that the right answer is to change nothing structural. Usually the advice is unglamorous: tighten the AP cut-off, enforce dimensions at entry, document the close so it survives a holiday, and revisit in a year.

That advice costs us a sale and it is the correct one. A company that implements an ERP at $12M because a salesperson framed revenue as the trigger tends to spend the following year regretting it, and the implementation gets remembered as a failure despite being delivered as specified.

The exception worth naming

If you are growing quickly and expect to be $40M within two years, factor that in. Fixing the chart of accounts and the entity model at $15M is materially cheaper than at $40M, because there is less history to restate and fewer people whose habits have formed around the current arrangement.

That is not an argument for a full ERP now. It is an argument for getting the foundations — chart, dimensions, close discipline — right early, which we do as an engagement without selling you a platform.

Questions

What people ask.

At what revenue should we consider an ERP?

Revenue is the wrong trigger. Entity count above two, close duration above ten days, bill volume above 200 a month, or imminent external scrutiny — you want at least two of those before the conversation is warranted.

Are you really telling us not to buy?

About a third of the time at this size, yes. It costs us a sale and it saves you a year of regret, and the referrals from having said it have been worth more than the deals.

What should we spend money on instead?

Chart of accounts design, enforcing dimensions at entry, documenting the close, and AP automation if volume warrants it. All of it is cheaper than an implementation and none of it is wasted if you later move.

We are growing fast — does that change it?

It argues for fixing the foundations early, because restating a chart of accounts at $40M is much harder than designing it at $15M. It does not argue for a full platform now.

Do you work with companies this size at all?

Yes, mostly on reporting, AP automation, and foundations rather than on ledger replacement. And we say no reasonably often, which is why the yeses are worth something.

At this size you probably do not need an ERP