Manual invoice chasing vs. automation: a decision in plain terms

The debate between manual invoice chasing and automation is not a moral one. It is a volume, variance, and policy problem. Manual works when the number of open items is small enough that a thoughtful human can know each story, the relationships are high-touch, or the work is legally sensitive. Automation works when the process is well understood, the failure mode of “someone forgets a step” is expensive, and the messages should be consistent so customers can predict you. The dangerous zone is a middle where you automate a mess: same wrong PO line, ten times, faster.

When manual chasing is the professional choice

• A tiny set of large accounts where a mistimed bot email could nuke a renewal.
• A dispute you should not dunning-ify until someone senior has read the thread.
• Markets or regulators where a canned sequence could create compliance risk without review.
• A founder who, today, is still fixing invoice data quality: automation only speeds up a wrong ask.

When automation wins: fewer heroics, fewer dropped balls

Once you are past a handful of open invoices, memory becomes the bottleneck, not politeness. Automation should mirror a policy the team has already written down: who gets reminded when, and what to do for disputes. If you do not have the policy, buy a coffee for the team, write a one-page, then return to the automation question. The accounts receivable automation resource walks the “what to systematize first” angle; this article is the fork in the road before you implement.

How bad automation is worse than a slow human

Bad automation nags a customer who has already said they are insolvent. It sends the same text after a public dispute. It emails the wrong person because your CRM field was stale. A good system has pause, dispute, and “promise date” as first-class ideas. If a vendor cannot describe those, you are not shopping for a receivable product; you are shopping for a very loud calendar.