agentleFS
Faaz17Faaz17/Agents-and-Skills/skills/research-writing/secondary-data-sources/SKILL.md

secondary-data-sources

Skill1 starsChanged 2 months agoMITView on GitHub

Where to find and how to verify secondary data for the MBS4820 Capstone — FedEx SEC filings and investor relations, GCC/UAE logistics and e-commerce market data, Statista/Passport/IBISWorld, World Bank LPI, UAE government statistics, and academic databases. Includes the source-quality hierarchy, the data-log protocol, and rules against fabricated figures. Use when researching market size, company financials, competitor data, industry trends, or any quantitative claim.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Never invent a figure: If a number cannot be traced to a real, named, dated source, write [UNVERIFIED — needs source] and leave it in the …

    line 10
  2. Distinctive:

    Always state which basis you mean.

    line 40
  3. Distinctive:

    Always capture the DOI at the moment you find the paper.

    line 93
  4. Very specific:

    Every figure enters 02-Research/data-log.md before it enters a draft:

    line 98
  5. Distinctive:

    Prefer the primary source: If Statista attributes a figure to the World Bank, cite the World Bank.

    line 118
  6. Distinctive:

    Triangulate anything load-bearing: If a market-size figure drives the financial model, find a second independent source.

    line 127
  7. Distinctive:

    Every data point should either justify a decision in the business plan or be cut.

    line 141

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Secondary data — sourcing and verification

The rule that outranks the rest

Never invent a figure. If a number cannot be traced to a real, named, dated source, write [UNVERIFIED — needs source] and leave it in the draft where the reader can see it. A visible gap is a task; a fabricated number in a Turnitin submission is misconduct.

This applies to market sizes, growth rates, market shares, revenue figures, headcounts, delivery volumes, cost benchmarks, and every percentage. It also applies to "approximately" and "around" — a vague fake is still a fake.

Source-quality hierarchy

Work down this list. Stop at the highest tier that has the number.

  1. Primary company disclosure — SEC filings (10-K, 10-Q, 8-K), annual reports, investor presentations, earnings-call transcripts. Audited, dated, legally accountable.
  2. Peer-reviewed journal articles — the only tier that counts toward the 50% quota.
  3. Licensed market databases — Statista, Passport (Euromonitor), IBISWorld, Mintel.
  4. Multilateral and government statistics — World Bank, IMF, WTO, UNCTAD, UAE FCSC, Dubai Statistics Centre, Dubai Chamber.
  5. Industry bodies — IATA, IRU, Global Express Association, Dubai Logistics Corridor.
  6. Consultancy research — McKinsey, PwC, Deloitte, BCG, Bain. Credible but promotional; read the methodology note.
  7. Quality business press — FT, Reuters, Bloomberg, Gulf News, The National, Arabian Business. For events and colour, not for headline statistics.

Never cite: blogs, SEO listicles, "top 10" content-marketing pages, Wikipedia (follow its sources instead), AI-generated summaries, or undated pages.

FedEx — where the real numbers are

FedEx Corporation, NYSE: FDX. Fiscal year ends 31 May — FY2024 covers June 2023 to May 2024. Always state which basis you mean; mixing fiscal and calendar years is a classic error a logistics-literate marker will catch.

WhatWhere
Revenue, operating income, segment resultsForm 10-K, Item 7 (MD&A) and Item 8
Segment breakdown (Express, Ground, Freight, Services)10-K segment reporting note
Average daily package volume, yield per package10-K MD&A operating statistics tables
Fleet size, aircraft, vehicles, facilities10-K Item 2 (Properties)
Strategy, DRIVE programme, Network 2.010-K Item 1 (Business), investor presentations
Risk factors — genuinely useful for competitor analysis10-K Item 1A
International/Middle East exposure10-K geographic disclosures
Quarterly trendsForm 10-Q, earnings releases

Access: investors.fedex.com and SEC EDGAR (sec.gov/edgar, full-text search). EDGAR is free, authoritative, and permanently linkable — prefer it.

Note the structural change: FedEx has been consolidating Express and Ground under Network 2.0 and spinning off FedEx Freight. Segment definitions have shifted across years — do not build a time series across a redefinition without saying so.

Comparators worth benchmarking: UPS (10-K), DHL / Deutsche Post (annual report), Aramex (DFM-listed, Dubai — especially relevant for a GCC venture), SMSA, Naqel, and the e-commerce captives (Noon Logistics, Amazon Logistics).

GCC / UAE market data

NeedSource
UAE e-commerce market size and growthPassport (Euromonitor), Statista, Dubai Chamber of Commerce
Courier, Express & Parcel (CEP) marketPassport category report, IBISWorld
Logistics performance benchmarkingWorld Bank Logistics Performance Index
UAE macro, population, tradeUAE Federal Competitiveness and Statistics Centre; Dubai Statistics Centre
Free zones, business setup costs, licensingDMCC, JAFZA, Dubai South, DED published fee schedules
Air cargo capacity and ratesIATA, Dubai Airports, Emirates SkyCargo
Cross-border trade rules, customsDubai Customs, WTO Trade Facilitation database
SME landscapeDubai SME, Ministry of Economy

Free-zone fee schedules are published and citable — they make start-up cost assumptions defensible rather than invented, which directly serves marking criterion 5.

Academic literature — where to search

Middlesex library discovery (mdx.keylinks.org), Google Scholar, ScienceDirect, Emerald, Wiley, Taylor & Francis, JSTOR, SSRN. The module reading list is at https://mdx.keylinks.org/new-ui/hierarchy/list/21326.

Journals that fit this project: Journal of Business Venturing, Entrepreneurship Theory and Practice, Strategic Entrepreneurship Journal, Academy of Management Review, International Journal of Physical Distribution & Logistics Management, Supply Chain Management: An International Journal, Journal of Business Research, Transportation Research Part E, International Journal of Logistics Management.

Always capture the DOI at the moment you find the paper. Retrofitting DOIs across 20 references the night before submission is miserable and error-prone.

The data-log protocol

Every figure enters 02-Research/data-log.md before it enters a draft:

FieldMeaning
IDD001, D002… — cite this ID in draft comments
ClaimWhat the number asserts, in one line
ValueThe figure, with units and currency
PeriodThe year or period the figure describes — not the publication year
SourceFull Harvard reference
URLDirect, stable link
AccessedDate retrieved
Confidenceverified / secondary / estimate / unverified
  • verified — read it in the primary source myself
  • secondary — a source I read cited it to someone else; note both
  • estimate — derived or triangulated; show the arithmetic in the notes column
  • unverified — placeholder only; must not survive to submission

Verification discipline

  • Prefer the primary source. If Statista attributes a figure to the World Bank, cite the World Bank.
  • Check the period. A 2025 publication often reports 2023 data.
  • Check the definition. "E-commerce market size" can mean GMV, revenue, or transaction count. Say which.
  • Check the geography. UAE ≠ GCC ≠ MENA. These get conflated constantly and the differences are large.
  • Check currency and units. AED vs USD, millions vs billions. State the basis and any conversion rate used, with its date.
  • Triangulate anything load-bearing. If a market-size figure drives the financial model, find a second independent source. Where they disagree, report the range and say why — that is exactly the critical engagement criterion 5 rewards.

Using secondary data well — what earns marks

Markers reward interpretation, not collection. For every figure, answer so what?

Weak: "The UAE e-commerce market was worth USD X billion in 2024 (Statista, 2025)."

Strong: "At USD X billion in 2024 and growing at Y% annually, the UAE e-commerce market is expanding faster than incumbent last-mile capacity, which grew Z% over the same period (Source A, 2025; Source B, 2024). That gap is the opening this venture targets."

Every data point should either justify a decision in the business plan or be cut.

Secondary-data limitations to acknowledge

State these in the methodology — acknowledging limitations is explicitly rewarded:

  • Data collected for another purpose may not fit the research question exactly
  • Publication lag means figures are historical
  • Aggregator definitions and methodologies are often opaque
  • Company disclosures are self-presented and selectively framed
  • Consultancy reports may serve a commercial agenda
  • Licensed databases raise access and reproducibility limits
  • Coverage of GCC SME logistics is thinner than for US/EU markets — a real constraint worth naming rather than hiding

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.