What a Tax Planning Report Should Contain: The Deliverable Anatomy

A tax planning report gets read twice. The client reads it once, in the planning meeting, and decides whether to act. The second read comes later, sometimes years later, when a successor preparer, a partner, or an examiner asks where a number came from and who decided it.

Most reports pass the first read. Plenty fail the second. The recommendation is there, but the baseline it was measured against isn't. The assumptions lived in the preparer's head. The figures carry no source and no date, and nobody signed.

This piece lays out the anatomy of a plan deliverable that survives both reads: six parts, a copy-ready outline you can paste into your own template, and a sample source manifest row. It's written for firms that deliver planning to a recurring S-corp book, and it applies whether you call the document a report, a plan, or a memo.

Report vs proposal: two documents, one anatomy

A tax planning proposal and a tax planning report answer different questions. The proposal comes first. It tells the client what the firm will analyze, which levers are in scope, and what the finished deliverable will contain. The report comes after the work. It holds the analysis and the recommendation.

The two should match. If the proposal promises a baseline, three scenarios, and a signed recommendation, the report should open with exactly that. A proposal that describes the six parts below is also the cleanest way to set expectations about what the client is paying for. Pricing the engagement is its own question, covered in how to price tax planning services.

The six parts of a defensible plan deliverable

Each part below has a job. Each also has a failure mode that shows up when it's missing.

1. Baseline

The baseline is the current-year projection if the client changes nothing. It names its inputs: the prior-year return, year-to-date books, and current payroll. Every scenario is measured against it.

Without a baseline, a report can claim a client "may save" some amount, but it can't show what that amount is saved against. A reader in year three has nothing to reproduce.

2. Scenarios

A scenario changes one or a few named levers and shows the projected difference from the baseline. Typical levers for an S-corp owner are the officer salary level, a retirement plan contribution, or reimbursements under an accountable plan. Two to four scenarios is usually enough.

Show the delta and the lever, not just the end number. A single "optimized" figure with no visible lever is the most common way a planning report becomes impossible to check.

3. Recommendation

The recommendation picks one scenario and says what the client should do, in a sentence the client can act on. Then it lists the actions with dates. Estimated payments are the easiest example: the 2026 individual due dates are Apr 15, Jun 15 and Sep 15 2026, and Jan 15 2027 (IRS Form 1040-ES, as of 2026-09-23).

For an S-corp owner, the recommendation is usually organized by topic: salary, distributions, retirement, and so on. The owner memo format covers those sections one by one.

4. Assumptions

Every projection rests on facts that could change. Write each one down, with who confirmed it and when. "Owner confirms no new hires in 2026, confirmed by email on [date]" is an assumption. "Business as usual" is not.

Some assumptions are choices between rules. The estimated tax safe harbor is one: 100% of prior-year tax, 110% if prior-year AGI was over $150,000 ($75,000 if married filing separately), or 90% of current-year tax (IRC §6654(d)(1), as of 2026-09-23). The report should say which one the plan uses and why.

5. Sources

Every figure in the report maps to a row in a source manifest: the figure, its value, the source, where in the source it appears, and the date it was confirmed. That habit is the same one that makes an IRS notice response traceable. The next section shows a sample row.

6. Sign-off

The last part is the named Reviewer, the sign-off date, and the scope of what was reviewed. The Reviewer controls the conclusion. The sign-off is how the report shows that on its face. If the plan changes midyear, the revised recommendation gets a new sign-off with a new date.

Sample source manifest row

Here's what a source manifest looks like with real 2026 figures in it. The third row is illustrative and uses placeholders.

What each column does:

• Figure names the number in plain words, the way it appears in the report.
• Value is the number itself, exactly as used.
• Source is the document or authority, specific enough to find again.
• Where in source points to the line, table, or section, so nobody has to reread the whole document.
• As of is the date the figure was confirmed. Figures change every year, and a figure without a date can't be checked.
• Pulled by records who entered it.
• Used in ties the row back to the parts of the report that depend on it.

Copy-ready deliverable outline

Paste this into your firm's template and fill it in per client. It works for a full report or a shorter plan letter.

1. Cover: client name, entity, tax year, prepared date, preparer, Reviewer.
2. Baseline: current-year projection with no changes, and the inputs it was built from.
3. Scenarios compared: a table with one row per scenario. Columns: scenario name, levers changed, projected change against the baseline.
4. Recommendation: the chosen scenario in one sentence, then an action calendar (action, owner, due date).
5. Assumptions: a table. Columns: assumption, confirmed by, date confirmed.
6. Source manifest: one row per figure, in the format shown above.
7. Sign-off block: the Reviewer's statement, name, and date.
8. Client acknowledgment: the client's confirmation that they received the plan.
9. Scope note: what the plan covers, what it doesn't, and that it is not tax advice outside the engagement.

For item 7, a sign-off block can be this short:

Reviewed and signed off: [Reviewer name], [date] Scope reviewed: baseline, scenarios [A-C], recommendation, assumptions, source manifest Changes since prior version: [none / list]

Why sign-off and retained workpapers matter at exam time

Planning decisions get questioned long after the planning meeting. An examiner asks why the officer salary was set where it was. A new partner inherits the client. The owner forgets what was agreed and asks again. In each case the question is the same: what did you rely on, and who decided?

The PDF the client received rarely answers that. It shows the conclusion. The firm's workpaper file answers it when it holds the baseline inputs, the scenarios that were rejected, the source manifest with dates, and the Reviewer's dated sign-off. That file belongs to the firm. It shouldn't depend on a vendor account staying open or a preparer staying at the firm. The same logic drives reasonable compensation documentation and workpapers.

A signed record doesn't guarantee any outcome. What it does is let the firm show its reasoning without reconstructing it from memory. That's also the pattern behind the Reasonable Compensation Study: traceable evidence, a Reviewer who signs off, and workpapers the firm owns.

What a plan deliverable should not contain

• Guaranteed savings, or any savings figure without "may" attached.
• Figures with no as-of date.
• Scenarios with no baseline to compare against.
• A recommendation nobody signed.
• Dollar examples that aren't labeled as illustrative.

Illustrative example: one scenario line, filled in

Illustrative example. Scenario C in a report for an S-corp owner adds a 401(k) elective deferral for 2026.

• Lever changed: owner elective deferral of $24,500, the 2026 limit (IRS Notice 2025-67, as of 2026-09-23). Source manifest row 2.
• Assumption: the plan allows elective deferrals for 2026 and the owner's salary covers the deferral. Confirmed by [owner], [date].
• Projected change against baseline: [computed in workpapers].
• Signed off by: [Reviewer name], [date].

The line doesn't claim a savings figure on its own. The number lives in the workpapers, next to the inputs that produced it. Actual results vary based on income, industry, and state.

Conclusion

A plan deliverable is only as defensible as its least-sourced number. Give the report a baseline, compare scenarios against it, write down the assumptions, source every figure, and have a named Reviewer sign it. Then keep the workpapers.

TracePrep ships the Reasonable Compensation Study today. It covers the salary input that most S-corp scenarios start from. A tax-planning workflow is in development. To see how a signed, sourced Study is put together, see how a Study is built.

Not tax advice.

TracePrep is a software product from TracePrep Inc., not a CPA firm. This content is operator experience for educational purposes; it is not tax advice and has not been reviewed by a licensed CPA. Consult a qualified tax professional for guidance on your specific situation. Source citations in TracePrep workpapers support audit defense but do not guarantee IRS outcomes.

Get started or See how a Study is built.