In short
- Buy judgement from a consultant and repetition from software. Most companies need both, in that order.
- If you will do this again next year, the licence is cheaper than the second engagement.
- Product footprints are a project, not a subscription, whatever your platform claims.
The real question is not consultant against software. It is whether the scarce thing in your situation is judgement or repetition. A consultant is the right buy when you need decisions made well once: a boundary set, a method chosen, a tender answered next month, a product footprint built. Software is the right buy when the same exercise repeats every year with new numbers in it. Most companies facing a first customer request need a little of the first and a lot of the second, and the way to tell which is to answer one question: will you have to do this again next year?
We sell both a platform and consultancy, so this page has an obvious conflict of interest. It is written to be useful anyway, because the way we lose money is selling somebody the wrong one of our two products and watching them leave a year later.
What does a consultant actually do that software cannot?
Four things, and they are all judgement rather than calculation.
Set the boundary. Which entities, which sites, operational control or equity share, what happens to the joint venture and the leased warehouse. Get this wrong and every number downstream is wrong in a way no tool will flag.
Screen scope 3. Deciding which of the fifteen categories are material to your business, and writing the justification for excluding the rest, is an argument you have to be able to defend rather than a calculation.
Handle the unusual. Process emissions, refrigerants, a treatment route with no sensible published factor, a group structure that changed mid-year. These are the cases where a generic factor library runs out.
Answer the specific document. A tender response, a carbon reduction plan in a prescribed format, a target that has to survive validation. That is drafting work with a deadline, and it is genuinely faster to buy.
There is also a fifth, less flattering to everyone involved: a consultant creates a deadline and an owner. Plenty of internal carbon projects fail for want of both.
What does software do that a consultant cannot?
Three things, all of them about the second year onwards.
Repeat cheaply. The first inventory is expensive however you build it. The fifth should be routine, and it only becomes routine if the method, the factors and the structure persist in one place between years.
Spread the work. Six data owners entering their own site data beats one coordinator chasing six spreadsheets. That needs roles and permissions, which is a product feature rather than an engagement.
Answer quickly. When a customer asks in April for a number you built in February, a platform gives it to you the same day. An engagement that ended in February gives you a PDF and a calendar invitation.
The underlying economics are simple. Consultancy is priced by time and does not get cheaper when you repeat it. Software is priced by subscription and the marginal cost of the second year is close to zero.
Which one does your situation need?
Find the row that matches. Most companies match two.
| Your situation | Usually the right answer | Why |
|---|---|---|
| First customer questionnaire, one entity, unlikely to repeat | Software, starting with a free tier | The calculation is small and the learning is worth keeping |
| Annual reporting across several sites | Software | Repetition is the whole cost, and it compounds |
| One tender closing next month | Consultant | You are buying a deadline you will actually hit |
| A customer wants a per-product footprint | Consultant, as a service | LCA, EPD, MKI and PCF are projects, not platform outputs |
| Setting a science-based target | Consultant first, software after | The target is judgement, the tracking is routine |
| Complex value chain, unclear materiality | Consultant to design, software to run | Screening once, measuring forever |
| Group consolidation heading for assurance | Both, plus a separate assurance budget | Auditors ask about structure and evidence, not totals |
The pattern in that table is consistent. Judgement is a purchase you make once and rarely repeat. Measurement is a purchase you make every year for as long as you have customers.
What do the two actually cost, and how do you compare them?
Not on the invoice. Compare them over three years, on the same defined scenario.
Software has a visible floor in this market, though barely. On 27 August 2026 we read the pricing pages of eleven carbon accounting vendors, including Coolset, Greenly, Watershed, Persefoni and Seedling, and found no numeric price on any of those five. EcoHedge, read the same day, publishes a free tier and a paid plan at GBP 990 a year. Hedgehog publishes a free account and Pro from EUR 1,200 a year on its G2 profile. So the visible entry point for a paid licence is four figures a year, and almost everything else in the category is quoted.
Consultancy is quoted too, always, because the price depends on the state of your data rather than on the size of your company. A firm cannot honestly price a footprint before seeing what you have.
Which is why the comparison has to be structural rather than numerical. Ask both sides the same three questions: what does year one cost in total, what does year two cost for the same output, and what do I own at the end. That last one separates the two purchases more sharply than price does. From an engagement you own a report. From a licence you own a maintained inventory and the ability to answer the next question yourself.
When is a consultant the wrong answer?
Three cases, and the first is the most common expensive mistake in this market.
When the exercise repeats and you have not built any internal capability. Paying a firm to rebuild your inventory annually, from your data, in their model, is the most expensive way to run a routine process. You will be quoted again next year and you will have learned nothing.
When the real problem is data collection. No amount of expertise makes your fuel invoices arrive. If your obstacle is getting six departments to send numbers, a consultant becomes an expensive chaser.
When you need an answer in a week, repeatedly. Questionnaires arrive with short deadlines all year. An engagement cannot be on standby.
When is software the wrong answer?
Two cases, and we sell software, so take these as the ones we would rather you knew before signing.
When what you actually need is a product footprint. An LCA, an EPD, an MKI or a PCF is a modelling project with supplier data behind it. At Hedgehog that is delivered as a service, and our platform does organisational footprints. If a customer has asked for a number per article or per part, a licence will not produce it.
When nobody internally has time to own it. A platform gives you a place to put data. It does not collect it. One of our customers put this plainly on G2 in August 2026: once the data is loaded everything works perfectly, and getting it loaded is the challenging part. If you have no one who can spend days on that in year one, buy the first year as a service and take the licence over afterwards.
How do most companies actually end up?
Doing both, in a particular order, and it is worth planning for that sequence deliberately rather than discovering it.
Year one, someone helps set the boundary, screen the categories and build the first inventory properly, either a consultant or a vendor's onboarding team. Year two onwards, the internal owner runs the cycle in a platform and buys advice only for the unusual: a new science-based target, a scope 3 expansion, a product declaration, an assurance readiness review.
That sequence works because it puts the judgement purchase where judgement is scarce and the subscription where repetition is expensive. It also means the second year costs a fraction of the first, which is the only way these programmes survive a budget round.
If you would rather have the first footprint built with you than by you, that is exactly what carbon footprint consultancy is for.
What should you do first?
Answer the one question honestly: will you be asked for this again next year? If yes, whatever you spend in year one should leave you owning a maintained inventory rather than a document.
Then get a rough first number before you commission anything, because the state of your data is what both quotes will actually be priced on. You can start a free account with no sales call and find out in an afternoon how bad the data really is. We go through the rest of the shortlist criteria in choosing carbon accounting software for SMEs, and if you want to talk the sequence through, you can book a meeting.
Method: each named vendor's own pricing page was opened on 27 August 2026 and checked for a currency figure attached to a plan. Sources: Hedgehog platform, Hedgehog on G2, ecohedge.com/pricing. Verified 27 August 2026.
Facts on this page were last verified on 2026-08-27.
