Portfolio: three pieces of proof
Three small pieces of ESG work that a stranger can check, such as an inventory, a gap analysis or a dashboard. You will document each with a README and publish them on one page.
- Core
- Intermediate
- about 40 hours
Written by the GreenHired editorial teamSources last checked 20 November 2026How we research roadmaps
Finish small, and finish three. Choose the pieces that match the work you want to be hired for, build each on free public data, and write each README as if its reader can never ask you a question. Two finished pieces with clear limits beat six half-documented ones.
Choose by role, not by what looks impressive
For a reporting role, build the inventory (piece 1), the short report (piece 6) and the framework gap analysis (piece 2). For a data role, build the dashboard (piece 4), the data-quality memo (piece 5) and the inventory. For an investment role, build the ratings comparison (piece 3), the dashboard and the short report.
Of the 16 postings we sampled (UK, EU, US and India; how we research roadmaps), 7 asked for Scope 1-3 or GHG Protocol work and 7 for GRI. The inventory and the gap analysis answer those two asks most directly. If you are still choosing a route into the field, see entry routes and adjacent roles.
A README that survives a rerun
A fictional 20-person office has built piece 1 for its 2025 electricity use. Its README states the question (what did the office emit from electricity in 2025?), the data (30,000 kWh from invoices), the factor (0.2 kg CO2e per kWh, with the year and table it came from) and the result: 30,000 x 0.2 = 6,000 kg, or 6 tonnes. The numbers are made up for this example, and the factors are not real ones. The method for the full version is in the inventory project.
A year later the office reruns it with a newer factor of 0.18. The result becomes 30,000 x 0.18 = 5,400 kg, 0.6 tonnes lower, although the office did nothing differently. If the README names no factor and no year, a reader cannot tell a real change from a different table. The same README should list what the piece leaves out, such as travel and commuting, so a missing category reads as a stated limit and not a hidden hole.
Check it cold
Ask someone to open the folder and answer three questions within five minutes: what was asked, which data and which date, and what the answer is and how firm. If they cannot, fix the README before anything else.
Publish, then rehearse
Remove employer data and paid ratings data you may not republish, link the three folders from one landing page and paste your three-line descriptions into your CV. Then practise walking through one piece aloud: why that factor, what you did with missing data, what the biggest uncertainty was.
What you will learn
- Which three pieces give the most proof for the role you want: an inventory, a framework gap analysis, a ratings comparison, a dashboard, a data-quality memo or a short report.
- How to document each piece so a reviewer can rerun it without asking you.
- How to use free public data and keep confidential company data out.
- How to publish the work and describe each piece in two or three lines.
The parts of this topic
Piece 1: a Scope 1 to 3 inventoryabout 8 hours
Calculate a small organisation's emissions with documented factors and a method note, using public data or a clearly labelled example.
Pick a fictional or public organisation, build an activity data table and calculate Scope 1, 2 and 3 totals in a workbook with a check tab. Choose each factor with its year and source, and write the factor year and region next to every number: mixing them without a record is the usual way this piece goes wrong. Free sources include the US EPA's emission factors hub (the 2025 edition, page updated in January 2026) and the UK government's conversion factors (the 2026 set, published 11 June 2026), both as read on 20 November 2026.
- EPA GHG Emission Factors Hub US EPA · Dataset
- UK government GHG conversion factors for company reporting UK government (DESNZ) · Dataset
Piece 2: a framework gap analysisabout 8 hours
Compare one framework with another for a real or imagined company and list what is already covered and what is missing.
Take a short list of ESRS datapoints from EFRAG's IG 3 Excel workbook and a matching set of GRI disclosures. Build a table showing which ones the company's existing report covers and which are gaps, with one sentence and an effort rating for each gap. A mapping with no reasoning tells a reviewer little, so for every partial match show why you called it partial. EFRAG says its guidance is non-authoritative, so say that in your note.
- Download the GRI Standards GRI · Standard
Piece 3: an ESG ratings comparisonabout 6 hours
Compare how several providers rate the same companies and explain why they disagree.
Berg, Kölbel and Rigobon compared six rating providers and, in the published paper, split their disagreement into scope (38%), measurement (56%) and weight (6%). An earlier working paper gave different shares, so quote the published ones. For a small version, pick three companies, collect public ratings or methodology notes and explain the differences under those three headings. The EU's rules on ESG rating activities (Regulation 2024/3005, applying from 2 July 2026) give useful context. Do not copy paid rating data you may not republish.
- Aggregate Confusion: The Divergence of ESG Ratings (Review of Finance, 2022) Oxford University Press / Review of Finance · Guide Open access under CC BY 4.0; the publisher's site may refuse automated link checkers
- ESG rating activities (European Commission) European Commission · Law or regulation
Piece 4: a small ESG dashboardabout 6 hours
Turn an open dataset into a clear dashboard with one question, three charts and a note on the data.
A dashboard built around one question beats one that shows everything. Take Our World in Data's CO2 repository (CSV, Excel or JSON, with a codebook) and build three charts in Power BI, Excel or any tool you know: one trend, one comparison, one per-person or per-output measure. Microsoft Learn has free Power BI training paths. State the data version, since the repository's README says its files are no longer updated and points to newer files in OWID's data catalogue.
- OWID co2-data repository Our World in Data · Dataset
- Microsoft Learn: Power BI training Microsoft · Course
Piece 5: a data-quality memoabout 4 hours
Write a two-page memo that rates the quality of each data source in an inventory and says what would improve it.
The GHG Protocol Scope 3 Standard compares primary data (for example, supplier-specific figures) with secondary data (industry averages). Primary data represent the value chain better but may be costly and hard to verify; secondary data are cheaper but less representative. Use that frame to grade each input in your inventory: if every input gets the same high-quality label, the memo tells the reader nothing. Then list the largest uncertainties and recommend the next improvement.
- GHG Protocol Scope 3 Standard (full PDF) GHG Protocol · Standard Chapter 7 includes a table comparing primary and secondary data
Piece 6: a short reportabout 4 hours
Write a four-page report that a manager could read in ten minutes: summary, table, method and next steps.
Write up your own work, for example pieces 1 and 5 combined, as a four-page report: a one-page plain-language summary, a results table with units and boundaries, a method note and a list of next steps. Report your result, not the standard: a reader who wanted the numbers does not need a tour of the GHG Protocol. Writing guidance for government communicators is a good free model for the summary. Ask a friend outside the field to mark anything they do not understand.
- Plain language guide series Digital.gov (US GSA) · Guide
Piece 7: publish where a recruiter can read itabout 4 hours
Put your three pieces on one simple page with a short description of each and a link to the data and method.
GitHub Pages hosts a site directly from a repository, supports plain HTML pages and Markdown with Jekyll, and allows a custom domain; its page tells you to create a public repository. Any simple host works. A link that opens a messy folder with no explanation gives the reader nothing to check, so give each piece its own folder with a README, link them from one landing page and describe each in two or three lines: the question and the result. Remove confidential or paid data first.
Practice task
Pick three pieces from the parts below. Finish each in its own folder with a README that gives the question, the data source and date, the method, the result and the limits. Publish the three on one page and write a three-line description of each that you could paste into a CV or a LinkedIn profile.
Skills you build
Portfolio buildingInventory calculationGap analysisData visualisationTechnical writingPublishing
Resources for this topic
Every source here is free. Each link opens the provider's own site in a new tab.
- CO2 and greenhouse gas emissions (Our World in Data) Our World in Data · Dataset OWID's own data are under CC BY 4.0, free to reuse with credit; third-party data on the page keep their own licence
- GHG Protocol Corporate Standard GHG Protocol · Standard
- pandas: 10 minutes to pandas pandas · Guide
- SQLBolt: interactive SQL lessons SQLBolt · Course The page does not state its price; it offers interactive lessons in the browser
Live ESG analyst jobs
Roles whose title matches this career path.
Spot a problem with this topic?
Tell us what is wrong, missing or unclear here. Your note is attached to this topic.
Notes are sent from this page with JavaScript on. Without it, write to us on the contact page.