A tax practice has something unusual: a list of people who will need the service again, on a schedule everyone knows in advance. The business rarely uses it that way.
The three automations that pay for themselves
1. Reactivating past filers before the season. The list exists, the reason to contact them is obvious, and doing it early converts capacity planning from a guess into a number. Most practices contact clients when the client contacts them, which means the busiest weeks decide who gets served.
2. Qualifying new enquiries immediately. In a seasonal business, speed matters more than usual: someone looking for a preparer in February is choosing this week. A lead that waits for someone to check the inbox has often already booked elsewhere.
3. Removing the steps between enquiry and engagement. The intake questions, the document request, the reminders. Identical every time, repeated hundreds of times, and currently done by a person who should be doing the work.
Segment by what they filed
A simple individual return, a return with a business attached, and a client who needed planning are three different conversations, three different values and three different capacity needs.
A single reactivation message across all three underperforms and, worse, fills the calendar with the wrong mix.
The seasonal constraint that decides everything
Build off season. Without exception.
During the rush nobody has attention to specify, test or fix anything, and a half configured system in a seasonal business is worse than no system: it fails at the exact moment failure is most expensive.
The practical calendar: design and build in the quiet months, run a small live test before the ramp, and enter the season with something that already worked once.
Where the phone beats the inbox
For past clients, a call reactivates the relationship faster than an email. Where the list is larger than the team can call, an AI first contact covers the volume and transfers the ones worth a person.
Two constraints specific to this industry. Messaging in the United States is gated by carrier registration and is blocked until approved, so plan voice first. And client data here deserves stricter handling than most sectors, which should shape tool choice before the tools are chosen.
What the system should leave behind
- A list that is cleaner than it was, with dead contact details routed out
- Every conversation written onto the client record, so whoever picks it up is not starting over
- A count of committed returns before the season rather than during it
- A repeatable configuration, so next year is a setting change rather than a project
The measure that matters
Not calls made. Engagements committed before the season opens, and hours of preparer time returned to preparation. Both are countable, and both are the argument for doing it again next year.