Q1: Why do credit teams end up with overlapping roles and unclear ownership?

Because most teams are not designed. They start with one person doing everything, a second person is hired to help, then a third, and nobody sits down to decide who actually owns what. Roles overlap, each person assumes the other did the check, and accountability gets blurry.

Six signs your structure has outgrown itself:

• Two people "sort of" own the same task, so each assumes the other did it.
• The same account is worked twice: two calls to the same customer about the same invoice on the same morning.
• Nobody is clearly accountable for the month-end overdue number.
• Every approval waits in one person's inbox.
• Your top accounts live in one person's head (key-person risk).
• The urgent always beats the proactive, so the team is permanently firefighting.

Q2: What are the core roles in a credit function?

There are six core roles. In a small team one person may wear several hats, and that is fine, as long as they know which hat is on at any point in time. Roles should be defined by the role, not the person, because people come and go and your process should not.

The six roles:

• Credit Manager: owns policy, risk appetite, the mandate and the total portfolio number.
• Credit Analyst (Vetting): assesses new applications and sets and reviews limits and terms, using data rather than gut feel.
• Credit Controller (Collections): owns existing accounts, promises to pay, follow-ups and stop supply.
• Cash Allocation / AR: matches payments to invoices, clears unallocated cash and reconciles.
• Disputes and Queries: drives pricing, delivery and purchase order queries to resolution with sales and operations.
• Credit Admin and Data: customer master data, documents, sureties and application records.

Q3: Where do the lines sit between the roles?

The friction usually lives in the handovers between roles, not in the roles themselves. Map each activity to one owner, with the others consulted or informed. Two owners means work gets done twice. No owner means it does not get done at all.

What to do:

• New limits: a controller should never quietly increase a limit because the customer asked nicely on a call. Ask for a review instead.
• Limit increases: the controller knows how the customer really pays, so they must be consulted alongside the analyst or credit manager.
• Stop supply: the controller triggers it under policy. Inform sales and the orders team so nobody keeps offering the customer more stock.
• Disputes: keep the collections team informed so they stop chasing an invoice that is already being resolved.
• Month end: each controller owns their own portfolio number and the credit manager owns the total.

Q4: What are the ways to structure a credit team?

There are three common models, and none of them is the right answer for every business. The best choice depends on your book, your customers and your systems.

What to do:

• By portfolio: each controller owns a set of customers end to end (split by region, industry, alphabet or sales rep). Works for mid-sized, relationship-driven books. Watch out for uneven workloads and key-person risk. Allocate accounts to people's strengths where you can.
• By risk band: effort follows risk. Strong accounts get a light touch such as automated reminders and quarterly reviews, while your most experienced people focus on the watch list. Works for large books with a wide spread of risk. Watch out for handovers when accounts change band.
• By process: specialist teams for vetting, collections, cash allocation and disputes, a bit like a factory. Works for high-volume books of smaller accounts. Watch out for a fragmented customer experience.

Q5: Which model fits our book?

Ask four questions about your own book. Most mature teams end up running a hybrid: a dedicated vetting team at the front door, portfolio or risk band ownership for the ongoing book, and shared services for cash allocation and disputes.

Questions to ask:

• How many active accounts per person? Few enough to know by name suggests portfolio. Thousands suggests process or risk band.
• How wide is the spread of risk? Similar customers can stay simple. A blue chip next to a fragile account should be banded.
• How relationship-driven is your industry? If customers expect one contact, protect continuity. If you rotate contacts, tell customers upfront.
• How mature are your systems? Process splits need clean handovers and notes in your ERP so information does not live in someone's head.

Q6: Should vetting and account management be separate jobs?

Ideally yes, because they are two different jobs with two different mindsets. Vetting is a decision under uncertainty about a business you do not know. Account management is a relationship over time, where circumstances such as business status, directors and trading patterns keep changing. In a small team, separate the time rather than the people.

What to do:

• Vetting needs: a clear application checklist, consistent data-led assessment, documented decisions and conditions, and agreed turnaround times for sales.
• Account management needs: ongoing risk monitoring and alerts, scheduled limit reviews, proactive calls before due date and early-warning signals acted on.
• At handover, pass on context: why the account was approved, the conditions, any concerns and the review date.

Q7: How do relationships help with collections?

Customers do not pay invoices. People in their accounts department pay invoices. Building the relationship before you need it makes you the supplier they pay first when cash is tight, and makes it much harder for a competitor to step in.

What to do:

• Pick up the phone. A call before due date beats ten emails after it, and email is best used to confirm a call in writing.
• Visit key accounts to see the business you are extending credit to, such as how they run their warehouse.
• Meet their teams in accounts payable, procurement and finance, not just the buyer, because people move on.
• Be the first to call. Do not wait for them to contact you.

Q8: How can AI help a credit team?

Think of AI as a diligent assistant that reads every account every day and never gets tired. It decides where to look, and your people decide what to do. This helps you catch the account that is quietly paying a little later each month before it slips through the cracks of a top-down ageing report.

The four steps:

• Watch: monitor every account for risk changes, payment pattern shifts and broken promises to pay.
• Alert: flag accounts that need attention today, not just those already overdue.
• Prioritise: rank the call list by value at risk and likelihood of late payment.
• Prepare: summarise account history and draft follow-ups before the call.

Q9: What should we measure for each role?

Measure each role on what it actually controls. Hold people to outcomes, and use activity measures for coaching only. Targets are yours to set, based on your terms, industry and book.

Q10: How do we stay on top of performance without micromanaging?

Make the scoreboard obvious, then get out of the way. Think of the petrol gauge in your car: you glance at it and keep driving, and only stop when it drops. Your credit policy should set out expectations so people do not need to run to the manager for every decision.

What to do:

• Give everyone their own number, so each person sees their portfolio daily and manages it.
• Use dashboards rather than status emails. If you are asking "where are we?", it is not visible enough.
• Review exceptions, not everything. You do not need to check all 900 accounts, only the 20 that moved the wrong way. Agree what counts as an exception and make it your weekly agenda.
• Coach on activity, hold to outcomes. When a number slips, look at the behaviours together, such as accounts contacted before due date and promises to pay captured.
• When a target is missed, ask why, and use the answer to set better goals next time.

Q11: How should approval limits and mandates be set up?

A mandate is a simple written statement of who can approve what, up to what value, under what conditions. It sounds bureaucratic, but it is what lets you move quickly: when the analyst has checked everything at level 1, level 2 does not need to check it again. Every level needs a named backup.

An illustrative ladder:

• Level 1, Credit Analyst: standard limits up to a set value, within policy and supported by data, and chasing missing documentation.
• Level 2, Credit Manager: larger limits and exceptions to standard terms.
• Level 3, Financial Director or CFO: major exposures and higher-risk policy exceptions.
• Level 4, Credit Committee or Exco: exposures that could materially hurt the business.

Q12: How do we make sure a mandate holds under pressure?

The test of a mandate is the hard day, not the easy one. When a customer threatens to cancel unless an order ships now, the process has to hold. Pressure like this has also been used in fraudulent applications where the paperwork was never finished. If sales can release stock to an account on stop, your mandate is decorative.

What to do:

• Write it down and get it signed off at the top. A mandate in someone's head bends the first time it is leaned on.
• Sales can escalate, not override. Challenges go up the credit ladder, never around it. Credit should respect how hard sales worked to land the customer, and sales should know that a lot of thought went into a no.
• Log every exception: who, why and on what basis. An auditor will ask, and it shows where policy needs updating.
• Review limits yearly. Mandates set for a smaller book push everything upward.
• Back the people who say no. Overrule them every time and they stop declining.

Q13: How should the team change as the book grows?

Build the team in stages and move before it breaks, not after. Most teams only move up a stage after a big write-off, a resignation or an unexplained month end.

Four common stages:

• Stage 1, the all-rounder: one person does everything. Write down a basic policy and process.
• Stage 2, small team: two to four people. First split vetting from collections and write the first mandate.
• Stage 3, specialised team: dedicated roles, a deliberate structural model and per-role scorecards.
• Stage 4, multi-team function: team leads, bands or regions, a credit committee and full management information.

Q14: What are the signs we need to add capacity?

Watch leading indicators so you add capacity before it breaks. Pick two or three, set your own thresholds and review them quarterly.

Leading indicators:

• Accounts per controller: they cannot reach every account before due date.
• Vetting turnaround: applications are taking longer and sales feels it first.
• Overdue or DSO trend: moving the wrong way two or three months running.
• Ageing disputes: queries stay open longer and longer.
• The manager's diary: more time approving and chasing than improving.
• The leave test: one person is away and things grind to a halt.

Q15: What can we do in the next 30 days?

You do not need a budget or a new system to start. Structure is what turns hard work into results: everyone knows what they own, can see how they are doing and has the authority to act.

Your 30-day checklist:

• Map your current roles against the activity table.
• Name one owner per activity and put it in your policy.
• Pick up the phone and book one customer visit.
• Write down or review your mandate, with backups.
• Pick your growth triggers and thresholds, and check that you are playing to the strengths of your people.

Need more help?

The support team is ready to assist

support@tradeshield.ai

+27 (0) 10 753 1630

www.tradeshield.ai