Does invoice chasing software actually work?

Scheduled reminders reliably improve the invoices that were only ever going to be a few days late — the ones where somebody simply had not got to it. They do very little for the invoices that become real problems, because those need a reply read, a promise recorded, a dispute spotted, or a decision about tone. Reminder tools automate the typing, which was never the hard part.

What is the difference between credit control and accounts receivable automation?

Accounts receivable automation describes a set of tasks being performed without a person — sending, matching, posting. Credit control describes a job with judgment in it: deciding which client gets a gentle nudge and which gets a formal letter, telling a real dispute from a stalling tactic, deciding when to accept instalments. The first is administration. The second is the part that recovers money.

Will chasing emails from software damage client relationships?

They can, and the mechanism is specific. Damage comes from mistimed or undifferentiated chases — a formal reminder to a client who paid yesterday, three separate emails on the same morning, or the same firm tone for a twenty-year client and a stranger. Anything sending on a fixed timer without reading replies will eventually do all three.

Should the chase come from my own email address?

Yes, and it matters more than most other features combined. An email that arrives from a software vendor's domain reads as a system, gets filtered as bulk mail, and cannot be replied to naturally. An email from your own address, in your own voice, is a message from a supplier the client knows.

Invoice chasing software or a credit controller? What an agency is actually choosing between

The search that leads here is usually some version of "how do I stop doing this myself". The options look like a spectrum from cheap to expensive. They are not. They are four different jobs wearing similar labels, and three of them do not solve the problem people buy them for.

The question that sorts it

Which of your unpaid invoices are you trying to fix?

Split your overdue list into two piles.

Pile one: nobody has got to it. The invoice is fine, the client is fine, it simply has not been processed. These get paid when somebody reminds them. They represent most of your count of overdue invoices and a minority of the money that goes badly wrong.

Pile two: something is happening. A dispute nobody raised, an approver on leave, a purchase order mismatch, a client whose own cash is tight, a promise made and broken. These are a minority of the count and almost all of the risk.

Almost every product in this market is very good at pile one and does nothing at all for pile two. If pile one is your whole problem, buy the cheapest thing that sends reminders and stop reading. Most agencies discover their real problem is pile two.

The four categories

1. Reminders built into your accounting system

Xero, and equivalents elsewhere, will send scheduled reminders on rules you set once for the whole organisation. Free, already paid for, and worth turning on this afternoon if you have not.

What they do: send a templated email at a fixed offset from the due date.

What they do not do: differentiate between clients, read replies, notice a promise, stop when somebody has paid but the payment has not reconciled yet, or escalate in tone. What Xero's reminders do and where they stop.

2. Dedicated reminder tools

A layer above the accounting system: better templates, multiple steps, sending from your own address, some reporting.

This is a genuine improvement, and it is still fundamentally pile one. The model is a timer. The unit of work is an email that goes out because a date arrived, not because somebody looked at the account and decided it needed something.

The tell, when you are evaluating one: ask what happens after the client replies. If the answer is that the sequence stops and the email lands in your inbox for you to deal with, you have bought a scheduler. That is fine — just price it as one.

3. A person

A credit controller, a bookkeeper doing it one day a week, or the founder on a Sunday. Judgment, phone calls, and someone who knows that Meridian always pays late but always pays.

The constraint is attention, and it is unforgiving. Credit control rewards looking at the same list every single working day and acting on the two accounts that need something that morning. A person can do that. A person one day a week cannot, and a founder in a Sunday burst definitely cannot — which is why the founder option, the one that looks free, tends to recover the least per hour spent. What the hire actually costs.

4. Something doing the whole job

The distinction that matters is not how good the emails are. It is whether the thing looks at every open invoice every day and makes a decision about each one, the way a person in the role would.

Concretely, that means it reads the replies and tells a promise from a query from a dispute; it stops chasing the moment somebody pays or objects; it records the date a client committed to and comes back the morning after if nothing arrived; it knows one client is on instalments and another is disputing a line, and treats them differently; it escalates in tone rather than frequency; and it tells you the two accounts that need you rather than handing you the whole list.

If a product does those things, the comparison is with a salary. If it does not, the comparison is with a mail merge, whatever the pricing page says.

What actually separates them

Six questions, in the order they matter. They are also the questions to ask us.

Does the email come from your own mailbox, in your voice? An email from a vendor's domain reads as a system and gets filtered as one. This is the single most consequential difference in the whole category and it is easy to check: send yourself a test and look at the headers.

Does it read the reply? If a reply just lands in your inbox, nothing has been taken off you. The work in credit control is in the responses.

Does it stop on its own? Chasing a client who paid on Tuesday is worse than not chasing at all, and it is the failure mode that makes people switch these tools off within a month.

Does it record and chase promises? "It'll go out on the 25th" is the most valuable sentence a debtor ever says. Something has to hold that date and act on the 26th.

Does it treat clients differently? One tone for a twenty-year client and a stranger who has never paid on time is how relationships get damaged.

Does it show you what it did? If nobody can reconstruct why a client received a particular email on a particular day, you cannot hand the account to a solicitor and you cannot hold anyone accountable.

Where each one is genuinely the right answer

Fewer than about ten open invoices, all pile one. Turn on your accounting system's reminders. Buy nothing.
Twenty to sixty open invoices, mostly fine, one recurring headache. A reminder tool plus somebody who looks at the exceptions.
A book with real risk in it — concentration, disputes, instalment plans, clients whose own cash is tight. This needs judgment applied daily, which means a person, or something built to do a person's job rather than a person's typing.
Debt already months old and the relationship over. Neither. That is a collections or legal question.

The honest bit

We are in the fourth category and we are not neutral. So use the six questions on us as hard as on anyone else, and be suspicious of any answer to "what happens after the client replies" that is really a description of an inbox.

The category to avoid is the one that prices like an employee and behaves like a timer.

If it is easier to judge from a real book than from a list of questions, here is what the fourth category looks like at one agency — including the part that did not improve.

No vendor comparison table here on purpose: feature lists change monthly and a table written in August is wrong by November. The six questions do not go stale.