How do audit firms stop chasing clients for PBC documents?

Audit firms stop chasing PBC documents by letting the request list do the chasing. Keep one tracked list per engagement, each item with a plain description, a named client owner and a status. AI drafts the follow-ups, checks what arrived against what was asked for and flags gaps, and a senior approves every message before it goes.

#Key takeaways

• Most PBC chasing comes from vague requests and from nobody knowing what has already arrived.
• AI is useful for checking uploads against the request, drafting reminders and summarising status for the manager.
• A person on the audit team approves every message that goes to the client and every item marked as received.
• If your firm already uses a client portal, start there; AI adds the checking and the drafting, not another inbox.
• Measure open items per engagement and how often an item is re-requested, not how many emails were sent.

#Why does PBC chasing eat so much of the audit?

The request list goes out as a spreadsheet. The client's finance team uploads files to a shared folder, emails some, sends others on WhatsApp, and names them all "scan". A senior spends evenings opening each file to work out which request it answers, whether it covers the full period and whether it is the signed version. Then the reminder emails start.

The problem is common. In a survey of North American accountants published by Suralink, a vendor of document-request software, 95% of respondents reported challenges with document requests, including assessing what came in and securing support (Suralink state of audit efficiency report via newswire, checked november 2026). Treat a vendor survey as directional, but most audit managers will recognise the pattern.

#What part of the chase can AI take on?

• Reading each upload and matching it to the request it answers, with a confidence note.
• Checking the obvious: right entity, right period, all pages, signed where a signature is required.
• Drafting a specific follow-up: "we received the march bank statement; january and february are still open".
• Writing a short status note per engagement for the manager: open items, overdue items, items received but not yet reviewed.
• Translating or summarising Arabic documents for a reviewer who reads English, with the original kept on file.

This is the kind of work firms already trust AI with. PCAOB staff reported that firms they spoke to were using generative AI mainly for administrative and research activities, and expected it to augment, not replace, auditors (PCAOB staff, july 2024).

What it should not do: decide that a document is sufficient audit evidence, close an item on its own, or send anything to the client without a person approving it. Those stay with the engagement team.

#How do you set up a request list that chases itself?

1. Write each request so a client clerk can act on it: what, which period, which format, why it is needed.
2. Assign every item an owner on the client side by name, not "finance team".
3. Agree one channel for uploads at kick-off and say that files sent elsewhere will be moved there.
4. Group requests by owner so each person gets one reminder with their own open items, not the whole list.
5. Review status on a fixed rhythm with the client controller and close or re-scope stale items together.

This alone fixes a large share of the chasing, with or without AI. The AI part pays off when volume is high: many engagements, many uploads and a team that cannot read every file the day it lands.

#When is a client portal or a template enough?

For a small practice with a handful of audits, a clean request template and a shared folder with a naming rule is enough. If you already pay for a PBC portal such as Suralink or the request module in your audit software, use it fully before adding anything. AI earns its place when seniors still spend hours checking uploads that the portal only stores.

#Where does Stennir fit?

Stennir starts with AI consulting: the AI Opportunity Audit looks at your engagement workflow and ranks where AI would remove the most work, with request checking usually near the top. If it is the right first build, AI implementation runs it with real users on your own engagements, in Arabic and English, hosted where your client confidentiality rules require.

Every chase and every "received" mark is approved by a named person on your team and logged. After go-live, managed AI operations keeps it monitored and improved. AI training then teaches seniors and associates to use it on their own client files.

#What should an audit partner do next?

Pick one busy engagement and count the open items, the re-requests and the hours seniors spent sorting uploads. That is your baseline. See how other finance teams use AI on the finance page, read how to pick the first process to automate, then take the free readiness assessment or book a 30-minute call.

#Common questions