Designing a Comprehensive Research Plan

Hello! Your research tree now identifies the questions that could change a decision. This lesson turns those questions into a workable research plan: what evidence you need, where you will look, which comparisons you will make, what you will publish, and when you will stop.

The aim is not to plan an exhaustive investment study. It is to give your first company brief enough structure that you can finish it, trace its conclusions to evidence, and explain why you did each analysis. Allow about 40 minutes: roughly 10 minutes for the resources and 30 minutes to work through the example and draft your own plan. If you study in 30-minute sessions, finish the plan in your next session.

A research tree asks questions; a plan assigns work

Recall the illustrative decision from the previous lesson:

Should Company A or one listed peer be prioritized for deeper diligence as a potential growth business, considering disclosed growth, operating margins, cash generation, and principal risks?

Your tree divided this into growth, economics, and reasons to withhold priority. A plan adds a test to each relevant leaf. For instance, “How has revenue changed?” becomes: obtain each company’s consolidated revenue for three aligned financial years, record the source and definition, calculate comparable growth rates, and show the result in an Excel schedule and the brief.

That distinction matters. “Research the financials” gives you no way to tell whether you have enough information or whether an attractive chart answers the question. A useful plan specifies an evidence requirement, a source path, an analytical method, and a decision use.

The Firm Learning video explains why hypothesis-based work begins with a provisional answer rather than an attempt to collect everything. Watch how an initial view becomes a list of investigations—not a conclusion that must be defended.

For the running example, an initial hypothesis might be: Company A merits priority if its stronger growth is supported by acceptable margins and cash generation, and if the principal risks do not overturn that case. This is deliberately conditional. The plan should make it possible for the peer to win—or for the public evidence to prove insufficient.

Specify evidence before choosing sources

For each leaf, first describe what must be established. Only then decide where to search. Otherwise, it is easy to download an investor presentation, find an appealing claim, and build the report around whatever the company chose to emphasize.

For a compact first brief, the essential evidence might be:

• Growth: each company’s reported revenue for the same three financial years; segment figures only if their definitions permit a useful comparison.
• Economics: operating-profit and revenue figures on understood definitions; cash-flow figures and any disclosed explanation for a divergence from reported profit.
• Reasons to withhold priority: material disclosed risks, changes in reporting or business mix, and gaps that prevent a fair peer comparison.

Notice the precision. “Revenue” still needs a reporting basis, period, currency, and unit. “Operating margin” requires a consistent definition of operating profit. A number copied correctly from each annual report can still produce a misleading comparison if one figure is consolidated, the other standalone, or the periods differ.

A public information book is a collection of public company materials assembled to support analysis. Corporate Finance Institute’s short explanation offers a useful menu of document types. Treat it as a menu, not a demand to obtain every item: its examples include US filings and paid research access that may not suit your first publicly sourced case.

For a publicly listed Indian company, your first search path can be the company’s investor-relations materials and its stock-exchange disclosures: annual reports, financial results, presentations, and material announcements. Use a credit-rating report if debt or liquidity is decision-relevant and such a report exists. Reputable industry reporting can add context, but it should not silently replace a company filing as the source of a reported financial figure.

The Plain Bagel’s walkthrough shows how business understanding and financial analysis call for different evidence. It describes a longer stock-research process than you need for this brief, so borrow its source-selection logic rather than its full scope.

Build a small evidence-to-output matrix

The matrix below is a filled example of a plan, not findings about real companies. It retains the previous lesson’s question and gives every analysis a job.

The method column describes what you will do, not what you expect to find. The output column prevents “interesting but unused” research. If a source or calculation will not affect a finding, chart, limitation, or decision, it probably does not belong in this first release.

Give the tasks an order. Priority 1 is obtaining comparable revenue, margin, and cash-flow evidence and checking the peer fit. Without that foundation, attractive strategic commentary cannot rescue the comparison. Priority 2 is investigating the growth drivers and the strongest contrary evidence. More detailed segment work is conditional on usable disclosures; it is not a reason to delay a defensible brief indefinitely.

A plan can also state what it excludes. For this decision, a DCF would require a forecast and valuation assumptions beyond the promised rapid brief. You will build that skill later. Excluding it here keeps the first case honest and publishable sooner, rather than treating an unfinished valuation as evidence of rigor.

Predefine sources, calculations, and publication artifacts

Your plan needs a source path that is repeatable. For each required item, write a preferred source and a fallback, then log what happened. For example:

FY revenue: annual-report consolidated income statement; if the year is not covered there, check the relevant filed financial results. If the figures differ, record the reason before using either. Do not fill the cell from an unsourced finance-site snapshot simply to complete the table.

Maintain the source register you created earlier: document title, URL, publication date, access date, relevant page or section, claim or figure supported, and a reliability note. A separate search log can record where you looked and what you did not find. That makes “not publicly established” a documented limitation rather than an unexamined omission.

Cornell University Library’s guidance concerns evidence searches in another field, but its advice to document search locations, terms, dates, and selection rules transfers well to a public-company research plan.

Write the metrics and methods into the plan before opening Excel: revenue change over aligned periods, a consistently defined operating margin, a peer comparison, and a review of operating cash flow alongside profit. The next lesson will work through the growth and comparison calculations and their common denominator and period errors. Here, the important planning question is whether the inputs can be obtained and compared.

Specify deliverables just as explicitly. For the first case, the planned package is a two-page, source-linked company brief, an Excel workbook with raw sourced figures and separate calculations, a source register, and three decision-relevant charts if the evidence supports them. Plausible chart candidates are a revenue-growth comparison, a margin comparison, and a cash-flow-versus-profit view. They are candidates, not a quota to fill with weak data; a chart should earn its place by clarifying a decision test. The brief will state a conclusion, its strongest counterargument, and its limitations.

Decide in advance what “enough” means

Stopping criteria protect both quality and speed. “Stop after three hours” is too mechanical: you might stop before checking the peer’s reporting basis. “Stop when I understand the company” is too vague: there will always be another presentation to read.

Use evidence gates and a bounded search together:

1. The core comparison is traceable. Each figure used in the decision has a recorded source, period, unit, and definition; calculations can be reproduced from the workbook.
2. The decisive branches have an answer or an explicit limit. Growth, margins, cash generation, and material risks each have a supported finding, or a clearly documented reason they cannot be established.
3. The initial hypothesis has faced a real challenge. You have checked at least the strongest plausible counterargument from the disclosed risks or contrary performance, and revised your conclusion if warranted.
4. The publication package agrees with itself. The brief’s numbers match the workbook and charts; its conclusion does not claim more certainty than the evidence permits.

For a missing but important item, specify a short fallback search in advance: check the annual report, relevant filed results, and the investor presentation or report notes. Record the search and then either qualify the conclusion, choose a more suitable peer, or narrow the question. Repeated searching without a defined next source is not progress. Conversely, a brief that says “the comparison is inconclusive because the disclosed measures are incompatible” can be a credible outcome if it explains the incompatibility and what diligence would resolve it.

A timebox can help you manage 30-minute daily sessions—for example, reserve one session for obtaining core filings and another for filling and checking the historical schedule. Treat that as a scheduling aid, not permission to publish before the evidence gates are met. If the gates fail, reduce scope or report the limitation rather than inventing certainty.

Put the plan in your company folder

A good working artifact is a short research_plan.md in the relevant company-study folder. Keep it readable enough that you could show an interviewer how the finished brief was designed. This structure includes all six elements of the learning outcome:

Decision question:
Initial, falsifiable hypothesis:
Company and peer; reporting periods:

For each decision-relevant research-tree leaf:
- Required evidence:
- Preferred source and fallback source:
- Metric or analytical method:
- Planned finding, chart, or workbook output:
- Priority and response if the evidence is unavailable:

Shared deliverables:
- Source-linked Excel workbook
- Source register and search log
- Two-page brief and supported charts

Stopping criteria:
- Core figures are traceable and comparable, or limitations are explicit
- Growth, economics, and material risks are addressed
- Strongest counterargument has been checked
- Brief, charts, workbook, and citations reconcile

Out of scope for this release:

The best plan is not the longest one. It makes the proposed work testable, finishable, and auditable. You should be able to point from any planned chart or conclusion back to a research-tree leaf, an evidence requirement, and a target source.

Key takeaways

A research plan is the bridge between a decision question and a published case. It specifies required evidence, where to seek it, metrics and methods, outputs, and stopping rules before findings tempt you to change the question. For your first brief, prioritize a reproducible peer comparison and an honest counterargument over an oversized collection of company facts.

Next, you will calculate and interpret growth, margins, market share, concentration, and indexed measures—turning the planned evidence into comparisons without period or denominator mistakes.