In short
- Doing scope 3 means three different things, and vendors rarely say which one they mean.
- You need four or five categories, not fifteen. Test the shortlist on those only.
- A larger factor library is not the same as better scope 3 coverage.
Nearly every carbon platform claims scope 3, and the claim covers everything from a spend-based estimate across three categories to primary data collected from named suppliers and held with its evidence. The gap between those is most of the work and most of the cost. Anyone facing a customer questionnaire, a tender or a standard that asks for value chain emissions needs to know which one they are buying. The first move is to decide which of the fifteen categories actually apply to you, then test each shortlisted tool against those categories and ignore the rest of the demo.
What does "does scope 3" actually mean?
Three different capabilities travel under the same phrase. Sort every vendor into one of them.
Estimation. The tool applies published factors to your spend or your activity data and produces a number for the categories you enable. This is legitimate, fast, and enough for a first inventory or a modest questionnaire. It is what most tools mean.
Inventory management. The tool holds a mixed-method scope 3 where some categories are spend-based, some activity-based and some supplier-specific, keeps track of which is which, and can restate a year when factors change. This is what you need for an annual cycle you intend to defend.
Supplier engagement. The tool sends requests to your suppliers, chases them, receives their data and folds it in. This is a distinct product surface and it is usually a distinct line on the quote.
A vendor answering yes to scope 3 might be offering any of the three. Ask which, in those words.
Which of the fifteen categories do you actually need?
Almost nobody needs all of them. The GHG Protocol expects you to screen for relevance and say what you excluded and why, and a focused inventory with a clear boundary beats a padded one.
| Category | What it covers | Who it bites hardest | Usual method |
|---|---|---|---|
| 1. Purchased goods and services | Everything you buy | Almost everyone, usually dominant | Spend-based first, supplier data later |
| 2. Capital goods | Machinery, buildings, vehicles bought | Asset-heavy manufacturers | Spend-based |
| 3. Fuel and energy related | Upstream of your own energy | Everyone, small but easy | Activity-based, automatic |
| 4. Upstream transport | Freight you pay for | Manufacturers, wholesalers | Tonne-kilometres or spend |
| 5. Waste in operations | What you send for treatment | Everyone, small | Tonnage by stream |
| 6. Business travel | Flights, rail, hotels | Professional services | Booking data |
| 7. Employee commuting | Getting to work | Office-based employers | Survey or headcount |
| 11. Use of sold products | Energy your product consumes in use | Equipment makers, often dominant | Product data and use assumptions |
If you are a services business, categories 1, 6 and 7 are most of your footprint. If you make things, 1 and 11 usually are. Screening this properly at the start saves more effort than any software feature, and it is the first thing a good scope 3 consultant does with you.
How do you test a platform's scope 3 in a demo?
Bring your own data and refuse the sample dataset. Five tests, thirty minutes.
Load one month of real purchase lines and see how much of it the tool categorises without human help, and what it does with the rest.
Ask it to switch one category from spend-based to activity-based and watch how the prior period behaves. If the history silently changes, your year-on-year comparison is unreliable.
Add a supplier-specific factor by hand. You will receive these by email from customers and suppliers, and you need somewhere authoritative to put one.
Exclude a category deliberately and see whether the tool records the exclusion and the reason. Screening is part of the standard, not a workaround.
Export the category-level breakdown. Customers ask about their own category, not your total.
Is a bigger factor library a better scope 3 tool?
Not by itself, and this is worth understanding because factor counts are the easiest number for a vendor to lead with.
EcoHedge advertises 277,000 or more emission factors on its comparison page, read on 27 August 2026. That is a larger library than the 20,000 or more spend-based and activity-based factors Hedgehog publishes, and we are not going to pretend otherwise.
What a library count does not tell you is whether the right factor exists for your material, in your country, for your year, and whether you can see which one was applied to a given entry. A hundred thousand factors for regions you do not operate in add nothing. One correct factor for the alloy that dominates your bill of materials changes your whole number.
So compare on three things instead of on the headline count: coverage of your actual categories and geographies, whether you can add your own supplier-specific data, and whether an entry can be traced back to the factor behind it. That last point is one our own customers have pushed us on, and it is the right question to ask everyone.
What about primary data from suppliers?
This is where scope 3 programmes either mature or stall, and it is worth planning before you buy.
Spend-based estimates get you a complete number quickly and cannot show a reduction you actually achieved, because they move with price rather than with physics. Replacing your largest suppliers with their real figures is the only route to a scope 3 number that responds to what you do.
Two practical points. First, most suppliers cannot answer yet, so budget for a low response rate in year one and target the handful that dominate your spend. Second, the requests you send should be proportionate. There is a cap on what large reporting companies may demand from SME suppliers, and being the customer who ignores it is a poor way to get data back. We set out where the line sits in refusing ESG questions.
Where does Hedgehog sit, and where does it not?
The platform is an inventory tool in the second sense above. It carries over 20,000 spend-based and activity-based factors, lets you add organisation-specific and supplier-specific CO2 data, and guides GHG Protocol setup, inventory building, data upload and reporting through an AI assistant. Entity management across locations and sites with roles for data owners, auditors and managers lets category owners load their own data. Reporting covers the GHG Protocol, PPN 006 and the CO2-Prestatieladder, with workflows for SBTi, B-Corp and EcoVadis. A free account needs no sales call and Pro starts at EUR 1,200 a year.
Three limits worth stating for a scope 3 buyer specifically.
Loading the data is the work. A customer said so on G2 in August 2026: once the data is in it works perfectly, and getting it in is the challenging part. Scope 3 is where that is most true, because the data starts in other people's systems.
Product footprints are a service. If category 11 is your dominant line, or a customer wants a per-product declaration, that is LCA, EPD, MKI or PCF work delivered as consultancy. The platform does organisational footprints.
Not a broad ESG suite. A customer rated us 3.5 out of 5 in June 2026 and said the platform is less complete if you want a broader ESG data and reporting system, with no data source management feature and no decarbonisation target monitoring.
What should you do first?
Screen the fifteen categories down to the four or five that are material to you and write one line for each on why it is in or out. That document is worth more in a customer conversation than any dashboard, and it takes an afternoon.
Then run one category end to end in a real tool before you commit to a shortlist. You can start a free account and try category 1 on a month of purchase data without talking to anyone.
Sources: Hedgehog platform, Hedgehog on G2, ecohedge.com/compare/hedgehog. Verified 27 August 2026. Check GHG Protocol category definitions against the standard itself before relying on the summaries above.
Facts on this page were last verified on 2026-09-17.
