Leading AI adoption

After tax season comes training season: the 90-day plan for an AI-fluent firm

The weeks after a filing deadline are the only quiet a firm gets all year, and most firms spend them recovering instead of rebuilding. Here's the 90-day sequence that turns the lull into the reason next busy season runs differently.

Adrián RidnerAdrián Ridner·July 10, 2026·4 min read

In short

What a firm does after tax season determines what the next one feels like: the quiet weeks after a filing deadline are the only window to build AI skills that stick by busy season.

  • Month one: run the busywork audit and train the partners and seniors who carry the biggest drafting loads, on real engagements.
  • Month two: extend to staff, build the firm's prompt and workflow library, and lock in verification habits as policy.
  • Month three: rebuild the recurring deliverables, the close, client communication, workpapers, around the new workflows and measure the recovered hours.
  • Firms that use the lull this way walk into busy season with the workflows already boring. Part of our advisory pivot series.
The window

Why the weeks after tax season are the only time to train

The weeks after tax season are the only stretch all year when a firm has the slack to build new skills with the team still intact. Accounting has a scheduling advantage most industries would envy: a built-in annual window when the work recedes. Slack is when new skills take hold, because nobody learns a new workflow during the close. Firms that treat the lull as recovery time get the same busy season every year; firms that treat it as training season get a different one.

The math of the window is what makes it urgent instead of merely nice to have. Skills built in the quiet weeks compound through every month that follows, and the accountant shortage means the hours they recover are hours no recruiter can replace. Wait until fall and you're learning under load; start now and busy season runs on workflows that are already reflex.

The plan

What does the 90-day plan look like?

The 90-day plan runs in three moves: audit and train the heaviest drafters in month one, extend to staff and systematize in month two, then rebuild the recurring deliverables and measure in month three. Days 1 to 30: audit and seed. List the firm's recurring deliverables, mark what AI can first-draft, and train the people carrying the heaviest drafting loads first, usually partners and seniors, on their own live engagements rather than exercises. The early wins need to land on the most visible calendars in the firm, because partner behavior is the only change management that works in a partnership.

Days 31 to 60: extend and systematize. Staff trains on their own work, the firm builds its shared prompt and context library, engagement templates, client-voice notes, review checklists, and verification habits get locked in as policy, not preference. Days 61 to 90: rebuild the recurring deliverables around the new workflows and measure, hours recovered per role, turnaround on client requests, where review catches issues. The AI time-savings calculator gives you the baseline to measure against, and the AI readiness scorecard shows where the firm stands before you start.

None of this requires a transformation office, the same finding as our SMB 90-day playbook, and the firm-level program is exactly what AI for accounting firms runs, partner to staff. The deadline that matters: be through month one before the extension wave builds.

The sequence

How to run the 90-day plan, week by week

The order matters more than the speed. Run it in this sequence.

  1. 1

    Weeks 1-2: run the busywork audit

    List the recurring deliverables, mark what AI can first-draft, and baseline the hours each role spends drafting today. That baseline is what you measure against in week 12.

  2. 2

    Weeks 2-4: train partners and seniors on live engagements

    Put the heaviest drafters first, working on their own real engagements instead of practice exercises, so the early wins land on the most visible calendars in the firm.

  3. 3

    Weeks 5-8: extend to staff and build the firm library

    Staff trains on their own work while the firm builds its shared prompt and context library: engagement templates, client-voice notes, review checklists.

  4. 4

    Weeks 5-8: make verification policy, not preference

    Lock in the habit that a licensed professional reviews and signs every AI draft. Write it down as policy so it survives the busy season.

  5. 5

    Weeks 9-12: rebuild recurring deliverables on the new workflows

    Rework the close, client communication, and workpapers around the new workflows so the saved time is built into how the firm operates.

  6. 6

    Week 12: measure recovered hours and pick advisory pilots

    Measure hours recovered per role against the baseline, turnaround on client requests, and where review catches issues, then choose the advisory pilot clients.

FAQ

Common questions

When should an accounting firm train on AI?

In the lull after a filing deadline, the only window when the team has slack and the work recedes. Skills built then are reflex by busy season; skills attempted during the close don't take. The 90-day plan fits comfortably between deadlines.

Who in the firm should train first?

Partners and seniors, for two reasons: they carry the heaviest draftable loads, so the recovered hours are biggest, and partner behavior is the only change management that works in a partnership. Staff follows in month two.

What should a firm measure after training?

Hours recovered per role against a pre-training baseline, turnaround on client requests, and where review catches issues. Candova AI's firm program builds the measurement in, so partners see progress in numbers they can check themselves.

Spend the lull on next busy season

Ninety days from now this is either done or deferred a year. Train the firm while the calendar allows it.

Power users save 10+ hours a week. Learn how.

The practical AI habits behind it, one a week.

Adrián Ridner

Written by

Adrián Ridner

Co-founder of Candova, founder of Study.com, and O'Reilly AI author

Adrián has spent two decades as a serial entrepreneur opening the doors to the life-changing impact of education. Before Candova, he founded and scaled Study.com into the largest platform for online college-credit courses, certification prep, and career-aligned degree pathways, helping millions of learners earn credentials for the modern workforce.

Future-proof your work with AI

On-demand · start today

Get started