Most agencies have one person who is good at getting money in. That is not a strength. That is a single point of failure with a personality.
In short
Ask an agency owner why a particular invoice is sixty days late and you rarely hear about the client. You hear that things got busy, that it fell off the list, that the person who usually chases was on leave. The invoice was not disputed. Nobody decided to let it slide. It simply required a human being to feel like making an uncomfortable call, on a day when there were more pleasant things to do.
That is the actual mechanism behind most late payment at small agencies, and it is fixable — not by hiring someone tougher, but by removing the moment of discretion entirely.
The instinct is to start chasing when an invoice goes overdue. By then two things have usually already gone wrong, and both were preventable.
The first is that the invoice never arrived somewhere useful. It went to a named individual who has left, or to a person who is not in accounts payable, or it needed a purchase order number it does not have. The client is not stalling; the invoice is sitting in a queue it will never leave.
The second is that it arrived but was never approved internally. Many organisations require the hiring manager to approve a supplier invoice before finance can schedule it. That approval is subject to exactly the same inertia as a timesheet.
Both are invisible to you until you ask. So ask early:
This is the part most people get backwards. Escalation feels like it should mean escalating language, so the third email is tense, the fourth is pointed, and by the fifth the relationship is damaged and the invoice is still unpaid.
It does not work because the person receiving those emails is usually in accounts payable and has no authority over the payment run. They are not withholding your money; they are processing a queue. Making them uncomfortable changes nothing they control.
What changes things is a different sender. Each step should widen the audience rather than sharpen the language:
Keep the tone identical at every rung. Facts, amount, date, request. The escalation is carried by who is on the message, and that is legible to everyone without anybody having to be unpleasant.
Get the hiring manager into the sequence earlier than feels natural. They asked for the contractor, they have a relationship with you, and they can walk to finance. Agencies avoid this because it feels like tattling — but a manager who discovers in month three that their supplier has not been paid is usually annoyed at their own finance team, not at you.
The hardest question on a contract desk: at what point do you stop supplying a client who has not paid?
You are paying the contractor weekly regardless. Every week you keep going, your exposure grows by a full week of cost while the receivable ages. The arithmetic is unforgiving.
What continuing to supply an unpaid client costs| Contractors on site with this client | 3 |
| Your weekly cost to keep them there pay plus on-costs, 3 × $1,586 | $4,758 |
| Already outstanding at day 60 | $27,000 |
| Four more weeks of supply adds | $19,032 |
| Exposure at day 88 | $46,032 |
The margin on those four weeks is perhaps $1,300. You are lending $19,000 to earn it, to a client already demonstrating they do not pay on time.
Set the threshold in advance — a days figure, an exposure figure, or both — and write it down before any particular client is in breach. Two reasons. Deciding in the moment means weighing an abstract risk against a relationship in front of you, and the relationship wins every time. And a policy is something you can tell the client about early, without it sounding like a threat: we suspend supply at 60 days, so let's not get there. Said at day 20, that is information. Said at day 59, it is an ultimatum.
A single receivables number tells you almost nothing. The same $80,000 outstanding might be forty clients paying normally, or one client four months late and everyone else current. Those are different businesses with different problems.
Report by client, in buckets, sorted by how late rather than how large — and put it where the people making commercial decisions see it. The person deciding which client gets the next exclusive should know which one pays at 74 days. Usually those two facts live with two different people, and the one deciding is the one who does not know.
In Echo Recruit
A cadence that runs without anyone deciding to run it
Due-date reminders go out on a 15/7/5/3/2/1/0-day schedule rather than when somebody remembers, and the receivables ledger sorts by how late things are rather than how large. Invoices carry part payments, allocations across invoices, refunds and credit notes, so what is genuinely outstanding is what the ledger says. Both parties' identity is frozen onto the document the moment it stops being a draft, which removes the commonest reason an invoice gets bounced back unpaid.
How should a recruitment agency chase unpaid invoices?
On a fixed cadence rather than on mood. Day 2, confirm the invoice reached the right place. Day 20, ask whether it is in the payment run. Day 31, a factual note asking for a payment date. Then days 38, 45 and 60, escalating who sends it rather than how sharply it reads.
Why chase an invoice before it is due?
Because the two commonest failures are already in place by the due date. Either the invoice went somewhere useless — a person who left, someone outside accounts payable, or it needed a PO it does not have — or it arrived and was never approved internally. Both are invisible until you ask, and both are cheap to fix with twenty-eight days left.
When should you stop supplying a client who has not paid?
Set the threshold in days, exposure, or both, in writing, before any particular client is in breach. Three contractors costing $4,758 a week means four more weeks adds $19,032 to a $27,000 balance to earn about $1,300 of margin. Decided in the moment, an abstract risk loses to the relationship in front of you every time.