A company's public record is rarely one neat folder. Its annual report may sit on an investor-relations site, quarterly slides on a media domain, sustainability material under a separate corporate section, and product spreadsheets several clicks inside a regional website.
The difficult part is not downloading files. It is building a small collection in which every file answers a different question.
A useful company research pack is not the largest pile of PDFs. It is the smallest defensible set of sources that explains the company.
Define the question before collecting files
Start with one company, one research purpose, and one cutoff date. “Understand Example Company's position as of 30 June 2026” is workable. “Download everything about Example Company” is not.
Your question determines the pack. A supplier review may need legal identity, locations, products, certifications, pricing clues, and recent financial condition. Competitor research may emphasize segment revenue, management priorities, product families, and changes in market language. A sustainability comparison needs reporting boundaries, metrics, targets, methods, and historical baselines.
Write the question at the top of a research note and set a time limit. In the first pass, aim for six to ten high-value files. You can always add evidence later; removing twenty overlapping reports is harder.
This workflow uses documents that are publicly reachable. That does not make every document current, official, complete, or licensed for redistribution. Provenance remains part of the research.
Confirm which company you are researching
Before searching by file type, establish the entity. A familiar brand may belong to a subsidiary, several companies may share a similar name, and a group report may describe operations very differently from a local legal entity's filing.
Record the legal name, parent company, important subsidiaries, official domains, headquarters jurisdiction, and any regulator identifier you can verify. For US public companies, the SEC's company filing search supports searches by company name, ticker, and CIK. In the United Kingdom, Companies House search guidance explains how to inspect the register, filing history, and document images.
Use those records to disambiguate the company, not to assume that every registry entry describes the entire commercial group. Put a short company-profile.txt or first row in your manifest with the identifiers and domains you actually used.
Once the identity is clear, search the official website for the main report:
site:example.com "annual report" filetype:pdf
Treat this as an illustrative query, not a promise that a particular company or file is currently indexed. FindFiles.net's site: filter matches an exact hostname, so also try a known investor-relations or media hostname when the main site produces little.
Use filings and annual reports as the factual anchor
The latest annual filing or annual report should usually be the pack's anchor. It provides the reporting period, legal scope, audited financial statements where applicable, accounting notes, business segments, risk disclosures, and management's account of performance. It is the document against which more promotional material can be checked.
For a US issuer, the SEC's guide to using EDGAR for research distinguishes the annual Form 10-K, quarterly Form 10-Q, and Form 8-K reports of material events. Those document classes answer different timing questions; a recent 8-K does not replace the annual context, and an annual report does not describe every event after year-end.
Keep the newest annual source relevant to your cutoff date. Add the prior year only if comparison matters. Add an interim filing when it materially updates the annual picture. If both a glossy annual report and a regulatory filing contain substantially the same financial statements, keep both only when the presentation or additional content has distinct value.
Structured data can be more useful than retyping tables. The SEC's EDGAR APIs expose submissions and XBRL company facts, but structured facts still require attention to units, periods, taxonomy concepts, amendments, and reporting context. Save a spreadsheet or API extract only when you can document what it represents.
Add results presentations for recency and management framing
Results presentations are often faster to scan than a filing. They show which metrics management emphasizes, how segments are grouped, which charts appear repeatedly, and what management says changed during the latest period. They are useful precisely because they are selective.
Search for the presentation itself:
site:example.com "results presentation" filetype:pptx
Many companies publish only a PDF export, so repeat the query with filetype:pdf. Add the quarter or year when the results page contains many periods.
Keep one current presentation if it adds recent metrics or an intelligible summary. Do not save the earnings release, slide deck, webcast handout, and results booklet automatically. Compare their periods, tables, and claims. If three documents repeat the same revenue figures and guidance, retain the authoritative version and note links to the others.
Use the presentation to understand management's framing, then verify material figures and definitions against the filing or accompanying notes. A polished chart can clarify a trend, but it does not gain authority from its design.
Treat sustainability reports as a distinct evidence set
A sustainability, ESG, impact, or responsibility report can add environmental and social metrics, stated targets, governance structures, methodologies, and a reporting boundary that financial documents may not cover in the same way.
site:example.com "sustainability report" filetype:pdf
The GRI Standards organize disclosures about an organization's impacts on the economy, environment, and people. A company may use GRI, another framework, several frameworks, or none. Record what the report actually claims, the entities and period it covers, whether metrics are assured, and where methods or baselines changed.
Keep a sustainability report when those subjects matter to your question. Do not keep it merely because every research pack appears to need one. A report and its separate data appendix may both deserve a place if the appendix contains the usable time series. A highlights brochure that repeats ten selected figures from the full report usually does not.
Use product, pricing and data files for operational detail
Filings explain the reporting company. Product brochures, catalogues, technical sheets, price lists, and spreadsheets can show what the company actually offers: product families, model numbers, regional ranges, technical positioning, packaging, and price structure at a particular date.
Try a product document:
site:example.com "product brochure" filetype:pdf
Then look for structured or tabular material:
site:example.com "price list" filetype:xlsx
Change XLSX to XLS, CSV, or ODS if necessary. A regional price list may exclude taxes, use a distributor's terms, or have been superseded. A catalogue may represent only one market. Record currency, geography, effective date, customer class, and revision instead of treating a public price as universally available.
| Document type | What it adds | Duplication risk | Practical keep rule |
|---|---|---|---|
| Annual report or filing | Legal scope, financials, risks, segments | Glossy and filed versions may overlap | Keep the authoritative annual source; add the other only for unique content |
| Results presentation | Recent metrics and management framing | Often repeats the release | Keep the clearest current source with distinct information |
| Sustainability report | Impacts, targets, methods, boundaries | Highlights may repeat the full report | Keep the full report and a separate data appendix when useful |
| Product brochure or catalogue | Offer, terminology, models, positioning | Regional editions may be near-identical | Keep the edition relevant to the market and date |
| Price list or spreadsheet | Structured products, prices, effective dates | Frequent revisions and mirrors | Keep the latest relevant version and material historical changes |
Keep historical versions that explain change
History matters when you need to know what changed, not when you are collecting years for their own sake. A prior annual report may expose a segment reclassification. An old catalogue may show when a product disappeared. Two price lists can reveal a revision that the current page no longer mentions.
Choose comparison points deliberately: the current version, the version at your research cutoff, and perhaps one baseline before an important event. Record why each historical file remains in the pack.
Current file search cannot retrieve a document that is no longer publicly reachable. If you know the former page or file URL, the Internet Archive's Wayback Machine guidance explains how to inspect captures. Archives are incomplete, captures can omit attachments, and an archived copy still needs provenance and date context.
Do not silently replace an older file with today's download under the same filename. Add the period or effective date to your local name and keep the original source URL in the manifest.
Stop collecting the same story repeatedly
Company communications are designed to travel. The same table may appear in an annual report, results release, slide deck, fact sheet, and media kit. Twenty files can therefore contain only five pieces of evidence.
Use two levels of deduplication. First, identify exact duplicates with file size and, when useful, a SHA-256 checksum. Two differently named downloads with the same checksum are the same bytes. Keep one copy and both source URLs if the provenance matters.
Second, identify near-duplicates by comparing the reporting period, table of contents, page count, key tables, and revision date. Ask one question before retaining either file: what could I support with this document that I could not support with the files already kept?
If the answer is “nothing,” add the URL to a duplicate or superseded field rather than saving another copy. If one file is easier to read but another is the formal source, keep both and state the role of each.
A useful pack preserves disagreement too. Do not discard an older claim merely because a later report changed it. Mark it as superseded, retain its date, and explain why the difference matters.
Build the pack and its manifest
A simple folder structure is enough:
01-filings02-results03-sustainability04-products-pricing05-data06-historymanifest.csv
For every retained file, record its document class, reporting or effective period, publication or revision date, source URL, access date, and provenance. Add an optional SHA-256 checksum, a one-sentence “unique contribution,” and a link to any file it supersedes or duplicates. This turns a directory of downloads into an auditable source pack.
A focused first pass can fit into 60 to 90 minutes:
- Define the question, company identity, cutoff date, and jurisdiction.
- Collect the latest annual source and relevant regulator record.
- Add the latest results material that contributes new information.
- Add one sustainability, product, pricing, or data source only when it serves the question.
- Choose meaningful historical comparison points.
- Deduplicate, rename files consistently, and complete the manifest.
- Write a short gap list instead of filling uncertainty with more downloads.
Official registries and investor-relations pages should remain your first stops for authoritative filings and current corporate material. General web search helps find pages and context. FindFiles.net complements both when the downloadable PDF, presentation, spreadsheet, or archived-looking file is the object you need.
Search public files on FindFiles.net
The finished pack should let another person see what you used, where it came from, when it applied, and why it earned a place. That is far more valuable than a folder containing every document with the company's name on it.