Client communication for small accounting firms: the chasing problem
Client communication for small accounting firms is mostly chasing documents. See what to automate, how to handle deadline season and what stays human.
Rebecca PearsonUpdated 11 min read
On this page
- What we'll cover
- Key takeaways
- Why chasing is the whole problem
- Making the request specific
- The chase sequence that works
- Deadline season
- Onboarding a new client
- The queries that interrupt the work
- Where professional judgement has to stay
- A chasing calendar that works backwards
- Frequently asked questions
- Related reading
The best client communication tool for a small accounting firm is an automation that chases documents item by item, escalates in tone, stops when the item arrives, and hands persistent cases to a person. An accounting practice's communication is mostly one activity: asking clients for things they have not sent, before a deadline that will not move.
Advice, planning and the other work clients value sit behind that chasing. A practice that handles chasing well has capacity for the valuable work. A practice that does not spends January on a task no client thanks it for. This page covers what to automate and where a practice should not automate at all.
What we'll cover
Key takeaways
- Document chasing is the first task to automate, because it takes the most communication time and is the task practices defer most often.
- Track chasing per document, not per client, so each reminder lists only what is still outstanding.
- Escalate the tone of the reminders, not their frequency, and stop the sequence the moment the item arrives.
- Never automate advice or client identification checks: automate the collection and the chasing, and keep the judgement with a person.
- Work backwards from the deadline, starting the request sixteen weeks out, so the final two weeks are for production.
Why chasing is the whole problem
Chasing is the whole problem because four features of accounting work make generic automation advice a poor fit.
The deadline is external and fixed. You cannot negotiate with a filing deadline, which means the practice absorbs every client's lateness.
The work cannot start without the client. Unlike most services, the practice is blocked by the person it serves, and has to ask repeatedly.
Chasing is socially awkward. Nobody enjoys sending a fourth email to a client they like. That awkwardness is why chasing gets deferred, and deferral is why January is what it is.
The volume is per client, per item. Each client needs not one chase but a bank statement, a mileage log, a dividend voucher, a signature, each with its own state.
Automation removes the awkwardness by removing the decision. Nobody has to send a reminder that sends itself, and clients receive a consistent process instead of a reflection of how busy their accountant felt that week.
Making the request specific
A specific request makes chasing effective, because vague original requests cause most failed chases.
Ask for a named list, not "your records". Bank statements for these months, the mileage log, the dividend vouchers, the loan statement. A client who receives a list can work through it; one who receives a general request does not know when they are finished.
Say what each thing is for, briefly. Clients supply documents faster when they understand why, and fewer arrive in the wrong format.
Give one obvious route to send it. A portal, a link or a reply: pick one route. Offering four means documents arrive across four channels and somebody has to reconcile them.
Track item by item. The unit is the document, not the client. A client who has sent three of five things needs a reminder about two things, not a repeat of the original list, and clients notice the difference.
Acknowledge what arrives. A client who sends something and hears nothing assumes it did not arrive, and either sends it again or worries. Automatic acknowledgement removes follow-up in both directions.
The chase sequence that works
Start earlier than feels necessary. Send the first request well before the deadline, because clients are slow and the practice absorbs the delay.
Escalate in tone, not in frequency. A gentle reminder, then a clearer one naming the deadline, then one setting out what happens if it is missed. Three well-spaced messages that change work better than six identical ones.
Name the consequence at the right point. Not in the first message, and clearly before it is too late. Clients respond to a specific date and a specific consequence in a way they do not respond to a general prompt.
Escalate to a person for the persistent ones. After a set number of unanswered chases, a phone call from someone at the practice. Some clients need a conversation, and the automation's job is to identify which clients need the call, not to replace it.
Stop when the item arrives. The most common complaint about automated chasing is a reminder for something already sent, which undermines trust in everything else the practice sends.
Report the state internally. Which clients are outstanding, what is missing, how long it has been. A weekly view lets a practice manage the season instead of discovering problems late.
This sequence (tracking items per client, escalating tone, stopping on receipt, escalating to a person) is simple to describe and slow to assemble across a practice's systems. On CodeWords you describe it in plain language and Cody, the automation builder, builds it, connects it to the systems the practice already uses, and deploys it. Automations connect to 3,000+ integrations. The Free plan covers light use. Pro costs $39 per month ($33 per month billed yearly) and Business costs $100 per month (CodeWords pricing page, 2026); details are on the pricing page.
Deadline season
Deadline season is where automation earns its cost, and five practices make it manageable.
Segment by risk, not alphabetically. Clients who were late last year will be late this year. Starting with them, earlier, works better than working through the list in order.
Know your capacity before you need it. Count how many returns the practice can complete in the final fortnight. Comparing outstanding work against that number weekly from November turns a January crisis into a November decision.
Make the internal view the shared one. Everyone in the practice should be able to see what is outstanding without asking. Otherwise three people each hold part of the picture.
Decide the cut-off and say it early. A stated date after which you cannot guarantee completion, communicated in advance, is fairer to clients and better for the practice than absorbing everything and working through the night.
Protect the last week for work, not chasing. By then, chasing has largely failed. The automation should front-load the chasing so the final period is production.
Onboarding a new client
Onboarding sets the tone for the client relationship, and most small practices run it inconsistently.
Collect what you need in a structured way. Identity verification, authority to act, previous accountant details, the systems they use, who to contact about what.
Handle the professional obligations properly. Client identification, engagement letter, professional clearance. These steps are mandatory and slip when somebody is busy, so automating the sequence and the chasing reduces risk.
Set expectations in writing at the start. What you need, when, how, and what happens if it is late. A client who was told this at onboarding responds differently to the third chase in January.
Get the recurring requests scheduled immediately. If you will need quarterly figures, set that up at onboarding rather than remembering each quarter.
The queries that interrupt the work
Routine client queries are a smaller drain than chasing, but automation still helps.
The same questions recur: when is my deadline, how much do I owe, have you received my records, what do I need to send, what is my reference number. The practice already holds the answers to several of these.
Route by type rather than by who opened the email. A technical query, a deadline question and a fee query go to different people, and routing them automatically removes a triage step.
Acknowledge with a realistic timescale. A client who knows they will hear within two days does not send a follow-up on day one.
Keep the substantive answers with people. The next section covers this boundary.
Where professional judgement has to stay
Professional judgement must stay with a person, because accounting is a regulated profession and over-automation carries real risk.
Never automate advice. Anything about what a client should do, how something should be treated, or what a figure means is professional judgement that carries professional liability. An automated message that reads as advice is a serious problem however carefully it is worded.
Keep the boundary visible in the messages themselves. An automated message should be recognisably administrative (a request, a reminder, a confirmation) and should route anything else to a person rather than attempting an answer.
Client identification and anti-money-laundering checks are yours. Automate the collection and the chasing of documents; the assessment is a professional obligation and stays with a person.
Be careful with client data. Practices hold unusually sensitive information. Check any system that touches it: where data is processed, who can reach it, how long messages are retained, and what your professional body and your insurer expect.
Check your engagement terms cover the channel. If you are going to communicate by messaging, that belongs in the engagement letter rather than being assumed.
A chasing calendar that works backwards
A chasing calendar starts from the deadline and works backwards, instead of starting when someone notices.
Sixteen weeks out: the request. The named list of what you need, why each item matters, and one route to send it. Most practices send this later, which compresses the rest of the sequence.
Twelve weeks: the first reminder, listing only what is still outstanding. A client who has sent three of five things should see two items, not the original list again.
Eight weeks: the clearer reminder, naming the deadline explicitly and what remains.
Six weeks: segment by risk. Anyone who was late last year, or who has sent nothing, moves to a phone call from a person. The automation has done its job by identifying who, early enough that a conversation still changes the outcome.
Four weeks: the stated cut-off. After this date you cannot guarantee completion. Said in advance and applied consistently, this is fairer to clients than absorbing everything and it protects the final fortnight.
Final two weeks: production only. Chasing has already succeeded or failed. This period is for doing the work.
The weekly internal view runs throughout: who is outstanding, what is missing, how long it has been. That report turns the season into something managed instead of endured.
Frequently asked questions
What should a small practice automate first?
Automate document chasing first, tracked per item rather than per client. Chasing is the largest single use of communication time and the task practices defer most often because it is awkward. Deferring it is what makes deadline season difficult, so start there.
Will clients mind automated reminders?
Clients mind less than practices fear, provided reminders name exactly what is outstanding and stop when the item arrives. Clients dislike a generic repeat of the original request, or a chase for something they sent a fortnight ago. Item-level tracking avoids both.
Can we use WhatsApp with clients?
Yes, many clients prefer WhatsApp and respond faster. Check consent, cover the channel in your engagement terms, apply your retention policy to conversations, and make sure replies reach a person. Use WhatsApp for administrative contact and keep advice in your normal channels.
How do we handle clients who never respond?
Escalate to a phone call after a set number of attempts, and state the point at which you can no longer guarantee the deadline. The automation's value is identifying which clients need the call early enough for the call to work.
Does this help with the January crunch?
Yes, because automation moves work earlier, which is the only change that helps. Chasing that starts in October and escalates properly produces a smaller January than chasing that starts when someone notices the problem. Start the request sixteen weeks before the deadline.
What about anti-money-laundering obligations?
Automate collecting and chasing the documents, and keep the assessment with a person. That split is compliant and practical, because the chasing is the tedious part and the judgement is the regulated part. Confirm the details with your professional body.
Should we build this or buy practice software?
Check what your practice software already does, because many products now include document requests and chasing. The gaps are usually the escalation sequence, the internal view of what is outstanding, and a connection to the channel clients actually read. Compare those three gaps against your current software before building.