Automate invoicing is two decisions rather than one. First, which side: raising invoices to send is billing, receiving and verifying them is payables, and the two share a phrase and almost nothing else. Second, if it is the receiving side, which step, because keying, waiting and exceptions each need a different remedy and none substitutes for the others.
Decide the side
Billing automation generates invoices from contracts, subscriptions or deliveries, chases them, and is measured in cash collected. Payables automation captures, verifies, routes and pays, and is measured in elapsed time and in not paying twice. State which you mean in the first sentence of any vendor conversation and you save both sides a fortnight.
Then measure the step
On the receiving side, time ten invoices and record minutes of work against days elapsed, then count last month's exceptions by reason. Keying minutes point at capture. A large minutes-to-days gap points at approval routing. Exceptions dominated by missing orders and receipts point upstream, where the fix costs nothing.
Sequence what you do
Upstream discipline first, because it raises the ceiling on everything else and is free. Then workflow, because approval waiting dominates elapsed time in most teams. Then capture, which now has somewhere to deliver. Each step delivers on its own, which is what makes the sequence low risk rather than a single large project.
Questions people ask about automate invoicing
What if we need both sides?
Treat them as two projects with two products, or one suite evaluated separately on each side. They rarely share requirements.
How long does the measurement take?
An hour a day for a week, or a sample of fifty invoices. Either answers the question well enough to direct the spending.
Can we skip straight to capture?
You can, and it is the most common way to improve the part that was already cheap.
What is the commonest sequencing mistake?
Buying capture when the constraint is approval waiting. It produces accurate data that then sits for a week, and the elapsed time a supplier experiences does not change at all.
Does volume change the answer?
It changes the size of the prize rather than which step is the constraint. A small team and a large one usually have the same three failure modes in the same order.