Ask any bookkeeper what they like least about the work and you will get some version of the same answer. Not the reconciliations. Not the deadlines. The chasing.
It is unpleasant for a specific reason: you are asking someone to do something they agreed to do, they are not refusing, and yet nothing is happening. There is no conflict to resolve and no way to move it forward except asking again, which makes you feel like a nuisance for wanting to do the job they hired you for.
The good news is that chasing is nearly always a design problem, not a client problem. Here is what actually reduces it.
Most chasing is the second, third and fourth request. Those exist because the first one was unclear or arrived in pieces.
The pattern that generates the most follow-up is asking as you go — a question on Tuesday, another on Friday, a third the following week. Each one is small, so it feels considerate. But each one also restarts the client's decision about when to deal with you, and small requests are the easiest to defer.
One consolidated request, with everything on it, is a single task the client can schedule. Six small ones are six things to postpone.
A generic checklist is an invitation to opt out. The client reads items that clearly do not apply, decides the list was not really written for them, and mentally files the whole thing as approximate.
Build the request from what you know about them. If you did their books last year, you know what they have. Naming the actual bank, the actual accounts, the actual months does more for completion rates than any reminder schedule.
And say explicitly what to do about items that do not apply. "If you did not have any of these this year, reply and say so" turns a stalled item into a two-word answer.
"When you get a chance" produces exactly what it asks for. Compare:
Could you send the bank statements when you get a chance?
with
I need the bank statements by Thursday 14th to have your GST return filed before the deadline on the 30th.
The second one is not more aggressive. It contains the information required to prioritise — the date, and what happens if it slips. Most people who miss deadlines are not ignoring you; they simply had nothing to weigh your request against.
This is the part that removes the emotional cost, and it is why chasing feels so much worse for a small practice than a large one.
In a bigger firm, an assistant sends the reminder. It is procedural, nobody takes it personally, and it goes out whether or not anyone felt like sending it. On your own, you are both the person owed the work and the person asking for it, so every reminder is a small decision about whether to be a bother today.
Automated reminders remove the decision. Set a schedule — say three days before the deadline, on the day, and three days after — and let it run. The message is polite and identical every time. You are not sending it; the system is.
Clients read automated reminders as organisation, not nagging. It is only awkward when it comes from a person who is visibly reluctant to send it.
The reason chasing sprawls is that the information about what is outstanding lives in your memory and your inbox, and both are unreliable in February.
You want to be able to answer, in five seconds, "what is still missing for this client." Not "have I emailed them," which is a different question and the one an inbox can answer.
A checklist per client with items ticked as they arrive does this. So does a portal that shows uploaded and outstanding. The mechanism matters less than having one place where the answer lives.
It also changes the tone of your follow-up, because you can name what you already have:
I have the Q1 and Q2 statements and the payroll summaries. Still need Q3 and the vehicle log.
That reads as competence. "Any update on those documents?" reads as a complaint.
Every step between intending to send and having sent is a place the task dies.
If your client has to find the email, remember the password, log into a portal they use twice a year, and navigate to the right folder, some fraction of them will decide to do it later. Later is often much later.
A direct upload link that opens to the right place, works on a phone, and does not require a login they will have forgotten removes most of that. Most clients photograph documents on their phone. Meeting them there rather than assuming a desktop scanner is worth more than any reminder.
The best time to set expectations is before there is anything to chase.
A short clause in the engagement letter, mentioned once out loud at the start:
Records are needed by the 10th of the month following each quarter. Where records arrive late, filing deadlines cannot be guaranteed and additional work may be billed at [rate].
You may never invoke it. Its job is to make the deadline real from day one, so that when you do follow up you are referring to something already agreed rather than introducing a new demand.
Some clients are late regardless. The system helps, but it will not fix them.
At that point it is a commercial question rather than an administrative one. The hours you spend chasing are hours you are not billing anyone, and that cost is real even though it never appears on an invoice.
Pricing that reality in is legitimate. So is deciding the relationship is not worth the drag. Neither is a failure of process — it is what process is for: making it obvious which problems are fixable and which are not.
Read next: How to Collect Documents from Clients at Tax Season