How to stop chasing clients for tax documents

The chase is treated as a discipline problem, which is why the usual remedies are about nagging harder: a portal, a reminder cadence, someone junior assigned to follow up. Those help at the edges. They do not touch the reason the chase is long, which is that most of it is caused before the first request goes out.

Five changes, in the order they pay off. None of them needs software you do not already have.

1. Ask for documents, not categories

The standard organizer asks for things like “investment income” and “business expenses”. A client reading that has to translate a tax concept into an object on their desk, and the translation is where the delay lives. They are not ignoring you. They do not know whether the thing they are holding is the thing you meant.

Ask for the artifact. “Your 1099-B from Fidelity”, not “investment income”. “The closing statement from the Elm Street sale”, not “real estate transactions”. If you prepared last year’s return you already know most of the specific documents by name, which is the whole basis of the next change.

2. Build the list from last year, not from a template

A generic request list asks every client for everything, which trains them to skim. A list built from what actually appeared on their prior year return is shorter, obviously personal, and answerable.

Practically: pull the prior year, list the source documents that fed it, and start there. Then add the deltas you know about, and mark the items you expect to have gone away. A client who sees “we do not expect a K-1 from the Harrison partnership this year, tell us if that is wrong” will correct you if it is wrong, which is a chase you never had to run.

3. One channel, and it is not email

The most expensive part of the chase is not asking. It is reconciling what arrived. When documents come by email, portal, text and a folder somebody dropped off, the question “what are we still waiting for?” cannot be answered without a person going and looking.

Pick one channel and be rude about it. If a client emails a document anyway, put it in the channel yourself rather than letting the exception stand, because the second exception is what ends the system. The point is not the channel’s features. It is that the answer to “what is outstanding” lives in one place and can be read without asking anyone.

4. Name the intervals before the season starts

“Follow up if you have not heard” is not a rule, and under pressure it becomes whatever the least busy person has time for. Decide the intervals in advance and write them down: first ask, then three days, then a week, then a week, then it escalates to a named person.

The specific numbers matter less than their being fixed. Fixed intervals can be delegated, automated, or simply followed. A judgement call cannot be any of those, and in March it will not be.

5. Escalate to a person, on a date, by name

Every request list needs a defined end. Not “keep chasing”, but “on the 12th, this becomes the relationship partner’s problem, and here is what we tell the client about the filing date”.

Two things happen when the escalation is dated. The chase stops consuming indefinite staff time, and the conversation with the client changes from a reminder to a decision, which is the conversation that actually moves documents.

What good looks like

A firm that has done these five can answer, for any client, on any day: what did we ask for, what has arrived, what is late, and whose problem is it now. Most firms cannot answer that without someone going and finding out, and the going-and-finding-out is a large fraction of what the season costs.

If it is the first change you want to make, there is a free request list builder on this site that produces one for a 1040 or an 1120-S, named by document rather than by category. No sign-up, and the output is yours.

Docenta automates the middle of this, which is the part that is pure bookkeeping: building the list from the prior year, sending on the fixed intervals, tracking what has landed, and escalating on the date. The five changes above are worth making regardless of what software you use, and four of the five are policy rather than tooling.

Once documents do arrive, the next problem is turning them into a workpaper somebody can review, which is the subject of a separate piece.