Credit control that chases consistently and escalates like a human.
Most late payment is not malice. It is invoices that slipped down someone's list because nobody chased on time, in the right tone, with the right information. Automation fixes the consistency. People keep the relationships.
Credit control automation means running the chase ladder on rails: reminders before and after the due date, statement runs, escalating tone and a stop-list, all applied to every invoice without exception. AI helps with the drafting, the promise tracking and the prioritisation. The calls, the disputes and the decision to escalate legally stay human.
The process in one paragraph
Define credit terms and a chase ladder: a pre-due courtesy note, an on-due reminder, escalating chasers at fixed intervals, monthly statements, then a stop-list and formal escalation. Automate the cadence, the drafting and the ledger prioritisation. Keep relationship calls, dispute resolution and legal escalation with named people. Measure debtor days, promise-kept rate and disputes, and feed the results into the cash-flow forecast.
Late payment is a persistent problem for UK small businesses, and the government publishes guidance and maintains statutory protections for suppliers, including the right to interest on late commercial payments. See GOV.UK's guidance on late commercial payments. The process below is Clerq's operational synthesis, not legal advice.
Every invoice, every time
The ladder runs on all invoices without exception, so nothing depends on someone remembering.
Escalation, not aggression
Early reminders are service messages. Tone hardens gradually, and only as far as the facts justify.
Humans own the exceptions
Disputes, sensitive accounts and legal steps are decisions with names attached, never automated.
Why late payment deserves a system, not a person's spare time
In most founder-led businesses, credit control is a part-time job done in the gaps: someone exports the aged debtors report when cash feels tight, sends a burst of chasers, then goes back to their actual role. The result is chasing that is sporadic, inconsistent between customers and driven by anxiety rather than by terms.
Customers learn from this. A supplier who chases politely, accurately and on schedule gets paid before one who chases in occasional bursts. Consistency is the single biggest lever, and consistency is precisely what software is good at. The relationship cost people fear from automation comes from bad automation: wrong amounts, chasing disputed or already-paid invoices, or tone that jumps straight to threats. A well-built process avoids all three by design.
The chase ladder
A chase ladder is the agreed sequence of contact from invoice issue to escalation. The exact intervals should fit your terms and your customer base; what matters is that the ladder is written down, applied to every invoice and only ever paused by a human decision with a reason attached.
Before the due date
A short courtesy note a few days ahead: invoice number, amount, due date, payment details. Framed as service, not chasing. This one message catches the invoices that never reached the right inbox, which is a far more common cause of lateness than refusal to pay.
On and just after the due date
A polite reminder on or immediately after the due date, then a firmer follow-up at a fixed interval. Each message states the facts, references previous contact and offers an easy route to flag a query. Any reply that raises a dispute exits the ladder immediately and routes to a person.
Statement runs and the phone call
Monthly statements give customers the full account picture and catch allocation errors on both sides. Once an account, rather than a single invoice, is materially overdue, a person calls. The call is a relationship act: it surfaces the real reason, secures a dated payment promise and keeps the customer.
Stop-list and formal escalation
Defined triggers, agreed in advance with whoever owns the customer relationship, move an account to stopped supply or formal escalation: a letter before action, statutory interest, third-party recovery. These are human decisions. The system's job is to present the facts and the history that justify them.
Where AI helps, and what stays human
The cadence itself needs rules, not intelligence: dates, intervals and triggers are deterministic. AI adds value in three specific places on top of that scaffolding.
- Drafting personalised chasers. Reminders that reflect the account's actual history, what was promised on the last call, which invoices are queried, how the customer usually pays, read as human and get answered. Drafts are reviewed before sending until the templates have earned trust for routine rungs.
- Payment-promise tracking. When a customer says "Friday", that promise is logged and checked against receipts. Kept promises pause the ladder; broken ones escalate with an accurate, specific message. This is the detail human teams drop first under pressure.
- Prioritising the ledger. Ranking overdue accounts by value, age, promise history and risk signals tells a small team where the next hour of calls does the most good, instead of working the report top to bottom.
What stays human is equally specific: relationship calls, dispute resolution, negotiating payment plans, stop-list decisions and any step towards legal action. Automating those is how you lose a customer over a query that a five-minute call would have solved.
The statutory lever: late payment interest
UK business-to-business contracts sit under the Late Payment of Commercial Debts (Interest) Act 1998, which provides for statutory interest and recovery costs on qualifying late payments. Check the current rates and conditions on GOV.UK rather than hard-coding them into templates, and take advice before relying on the Act in a specific case.
In practice, the Act is most useful as a calm, factual escalation step: a late-stage letter noting that statutory interest is accruing changes the conversation without theatrics. Many businesses choose never to invoice the interest itself; the point is that the option exists and the customer knows it. Whether to deploy it on a given account is exactly the kind of relationship judgement that stays with a person.
Feeding the cash-flow forecast
Credit control and cash forecasting are the same discipline seen from two ends. The receipts line of a 13-week cash-flow forecast is only as good as the collections data behind it: which invoices are promised for when, which customers habitually pay ten days late, which disputes will delay settlement.
A working credit-control system produces those inputs as a by-product: promise dates, behavioural payment patterns and dispute status flow straight into the weekly receipts forecast, replacing the optimistic assumption that everyone pays on terms. The mirror-image process on the payables side is covered in our guide to accounts payable automation.
Controls and measurement
Automated chasing needs guardrails, because a wrong message sent consistently is worse than no message. And it needs a baseline, because otherwise improvement is a feeling rather than a fact.
- Accuracy gate. The system chases only invoices that are genuinely open, undisputed and correctly allocated. Receipts post before each run, and disputed items are excluded automatically.
- Suppression rules. A human can pause any account with a recorded reason and a review date. Pauses expire; they do not silently become permanent.
- Audit trail. Every message, call note, promise and escalation decision is logged against the account, so the file justifies any formal step later.
- Measures. Debtor days (DSO), ageing by bucket, promise-kept rate, disputed invoices and their resolution time, and the share of invoices collected without escalation. Track them monthly against the pre-automation baseline.
An implementation path
As with any finance automation, stage the build so each layer proves itself before the next depends on it.
- First, write the ladder down. Agree terms, rungs, intervals, tone and stop-list triggers with whoever owns customer relationships. This costs nothing and improves collections on its own.
- Second, automate the cadence. Reminders and statements run on schedule from accurate ledger data, with dispute routing and suppression rules in place from day one.
- Third, add the AI layer. Personalised drafting, promise tracking and ledger prioritisation, each reviewed by a person until its accuracy is demonstrated.
- Fourth, connect the forecast. Promise dates and payment behaviour feed the weekly receipts line, and the measures above go on the monthly dashboard.
If you are unsure whether collections is your highest-value first build, a short structured look at the numbers settles it. That is what the two-week Diagnostic is for.
Frequently asked questions
What is credit control automation?
It is the use of software to run a consistent chase process: reminders on a defined cadence, statement runs, escalating tone, prioritisation of the debtor ledger and tracking of payment promises, under human oversight.
Will automated chasing annoy customers?
Not if it is designed well. Consistent, polite and accurate reminders are more professional than sporadic manual chasing. Relationship calls, disputes and sensitive accounts should always stay with a person.
Where does AI help in credit control?
Drafting personalised chasers that reflect account history, tracking payment promises against receipts, and prioritising the ledger by risk and value. Cadence and stop-list rules are better handled deterministically.
Can a UK business charge interest on late commercial payments?
Business-to-business contracts fall under the Late Payment of Commercial Debts (Interest) Act 1998, which provides for statutory interest and recovery costs. Check the current position on GOV.UK and legislation.gov.uk before relying on it.
What should credit control measure?
Debtor days, ageing by bucket, promise-kept rate, disputed invoices and their resolution time, and the share of invoices paid without escalation. Measure before automating so improvement is provable.
Primary and authoritative sources
This article is practical operating guidance. The following sources support the statutory and cash-management principles referenced above.
Find out how much cash your ledger is sitting on.
The two-week Diagnostic baselines your debtor days, maps the chase process and hands you a costed roadmap. It is yours to use with Clerq or another builder.