Two separate jobs here. Getting answers to the list you have, and making next month's list shorter. Most bookkeepers only work on the first.

Getting the list answered

Batch, never drip

Sending questions as you find them trains clients to ignore you — five emails in a week from their bookkeeper, each about $40, is noise. One list, once, at a predictable point in the month, is a task they can complete in five minutes.

Make each line answerable in four words

The reason lists come back half-finished is that some lines require thought. Reduce every question to recall, not analysis:

Subject: 6 quick questions before I close March

Hi {{first name}},

Almost done with March. I just need to know what these were for — a few words each is plenty:

• 03/04 — AMZN MKTP — $187.42
• 03/09 — SQ *THE COFFEE HO — $64.10
• 03/15 — TRANSFER TO CHECKING — $2,000.00
• 03/22 — HOME DEPOT — $412.88

Business or personal, plus roughly what it was, is all I need.

Thanks,
{{your name}}

Give them the date, the merchant and the amount

"Transaction #4471" means nothing. The client remembers a Tuesday, a shop and a price — give them all three and recall does the work.

Offer a default

For recurring charges you can usually guess: "I've assumed the $64 at the coffee shop was a client meeting — tell me if not." A client who only has to correct you replies far more often than one who has to compose an answer.

Cap the list. A forty-line question list gets abandoned at line six. If you have forty, something upstream is broken — code the obvious ones on a documented assumption and ask about the ten that genuinely matter.

When they don't answer

  • Code what you reasonably can with a documented assumption, and tell the client what you assumed. Most people correct an assumption faster than they answer a question.
  • Use a holding account for genuinely unknown items, and keep it visible — never bury an unknown in a plausible-looking expense account.
  • Never guess on anything material. A large or unusual item stays open until answered, and gets flagged to whoever prepares the tax return.
  • Track the pattern. A client with a persistent list of unanswered questions is an engagement pricing conversation, not a categorisation problem.

Making the list shorter

This is where the actual win is. A transaction is only a question because you have a bank line and no context — and context normally arrives as a receipt. Collect the receipt and the question never gets asked.

  • Collect receipts during the month, not after the close, while the client still remembers what they bought.
  • Make submitting a receipt a one-tap phone action so it happens at the till.
  • Build vendor rules for the repeat offenders — the same coffee shop does not need asking about twelve times a year.
  • Push clients toward separate business cards; mixed personal and business spending is the single biggest generator of these lists.
  • Ask about categories once, as a policy, rather than transaction by transaction: "how do you want me to treat client meals?" answers thirty future questions.

The honest calculation

A ten-line question list costs you the time to build it, the time to chase it, and the delay to the close. Collecting receipts as they happen costs the client two seconds each. The reason most practices do the former is not that it is cheaper — it is that the receipt collection process was never set up.

Fewer questions, because more receipts arrive

ClientClose collects receipts through a no-login link during the month, reads each one with AI, and posts it to QuickBooks Online or Xero attached to the transaction. The transactions that still need a question are the genuinely ambiguous few.

Start your free week →