Every controller I know has had the same conversation with their CFO. DSO is creeping up, the board wants it fixed, and the instinctive answer is to throw another body at collections. More calls, more follow-ups, more people chasing the same overdue invoices. It rarely works the way anyone hopes, and it's almost never the actual problem.
I've sat across the table from finance teams who doubled their collections headcount and watched DSO barely move. The issue wasn't effort. It was that the process upstream of collections was quietly generating the very problems collections was hired to clean up.
Where the time actually leaks
DSO isn't one number with one cause. It's the sum of every small delay between the invoice going out and the cash landing in your account: how fast you invoice, how easy it is for a customer to pay, how quickly you catch a payment that's about to go sideways, and how much manual matching happens before anyone even picks up the phone. Add a headcount to the phone-calling stage, and you've fixed the last five per cent of the problem while the other ninety-five per cent keeps generating new late payments every month.
The Credit Research Foundation's most recent data puts median DSO for domestic trade receivables at just under 37 days, and that's an average across companies with wildly different processes. The gap between the best and worst performers in that data usually comes down to how much of the invoice-to-cash cycle is manual versus automated, not how many people are working the phones.
Fix the front end first
The highest-leverage move is almost always earlier in the cycle than people think. If invoices go out late, with errors, or through a channel the customer doesn't actually check, you've built in a delay before collections even starts its job. Offering customers a way to pay instantly the moment they open an invoice, rather than routing them through a separate portal or waiting on a check, closes a gap that most companies don't even realise they've left open. I've seen companies shave real days off DSO just by tightening invoice accuracy and enabling automated payments at the point of delivery, before touching a single collections call.
The next lever is visibility. Most finance teams don't know an invoice is at risk until it's already thirty or sixty days overdue, at which point the recovery conversation is harder and more expensive. AI-powered AR automation changes this by flagging a slipping payment at day five instead of day forty-five, based on patterns in how that specific customer usually pays. A reminder at day five looks completely different to a customer than a terse email at day forty-five. One is a nudge. The other is a confrontation.
Let the system do the chasing that doesn't need a human
Not every follow-up needs a person. A polite reminder three days before a due date, a second nudge on the day it's missed, and a payment confirmation once it clears are all things AI-powered AR automation can handle without anyone lifting a phone, matching payments to invoices automatically as they land. That frees your actual team to spend their time on the accounts that genuinely need a human touch: the disputes, the awkward conversations, the customers who need relationship management rather than a reminder email. Industry benchmarks generally show automated AR workflows reducing DSO by somewhere in the ten to thirty per cent range compared to fully manual processes, and the reason is simple: automation removes the delay, not the diligence.
The real trade-off
Hiring more collections staff treats DSO as a labour problem. It isn't. It's a visibility and workflow problem that happens to show up as a labour problem once things get bad enough. A bigger team can absolutely paper over a broken process for a while, but it's an expensive way to buy time, and it doesn't scale. Every new customer, every new invoice, every new market you enter adds more manual load to a system that was already struggling.
The controllers I respect most didn't fix DSO by adding headcount. They fixed it by making the invoice-to-cash cycle fast, automated, and visible enough that collections became a smaller job, not a bigger team.
By:
Nick Chandi
Published
