EcoVadis questionnaires and ESG data requests: how to answer them with an AI agent
EcoVadis questionnaires, client requests and bank questions almost always ask for the same data. What they ask, where to find the answers and how an AI agent for sustainability finds them with their source.
When you answer an EcoVadis questionnaire, a client's data request or a bank's ESG questions, the longest part of the job is not writing the answer but finding the right figure and its evidence. An AI agent for sustainability like EVA looks for the answers in your company's data and documents and shows where every number comes from: whoever fills in the questionnaire checks and sends, without starting from scratch with every request.
Why ESG data requests are arriving more and more often
Sustainability data requests no longer come only from large listed companies. They reach SMEs from three directions, each with its own format and deadline.
- Clients and lead companies in the supply chain. Large companies measure Scope 3 and monitor their supply chain: they ask suppliers for emissions, policies and certifications, often through platforms like EcoVadis.
- Banks. For financing, they ask for ESG indicators through their own questionnaires. We cover this in the article on bank ESG questionnaires.
- Tenders and calls for funding. Environmental and social criteria reward those who document consumption, emissions and policies.
The questions are very similar, but each request phrases them differently: that is why, without a single data source, every questionnaire turns into a small project.
What the EcoVadis questionnaire is and what it asks
EcoVadis is a sustainability rating platform used by many large companies to assess their suppliers. The supplier fills in an online questionnaire, attaches supporting documents and receives a scorecard with a score from 0 to 100, which the client can view. The questionnaire is tailored to the company's sector, size and country and covers four themes, for a total of 21 criteria.
| Theme | What it assesses | Typical data and documents |
|---|---|---|
| Environment | Energy, emissions, water, waste, product impacts | Bills, Scope 1 and 2 calculation, waste registers, environmental policy, ISO 14001 |
| Labor and human rights | Health and safety, working conditions, training, diversity | Headcount by gender, accident register, training plan, HR policies |
| Ethics | Anti-corruption, fair competition, data protection | Code of ethics, 231 compliance model, privacy policy |
| Sustainable procurement | How you assess and engage your suppliers | Supplier code of conduct, supplier questionnaires, selection criteria |
Two rules matter more than the others. Answers only count if they are backed by documents: a policy that is stated but not documented carries little weight. And the scorecard is valid for twelve months: every year the questionnaire comes back, with updated data.
The questions that come up in every questionnaire
Whether it is EcoVadis, a client's supplier questionnaire or a bank's, some questions always come up. Almost all of them match a disclosure in the VSME standard: if you already have a sustainability report, you already have most of the answers.
| Typical question | Where to find the data | VSME disclosure |
|---|---|---|
| Scope 1 and 2 emissions for the last year | Gas and electricity bills, fuel invoices, maintenance contractor reports | B3 |
| Share of renewable energy | Supply contracts, Guarantees of Origin, solar PV systems | B3 |
| Waste generated and sent for recovery | Waste transfer forms and waste register | B7 |
| Employees at year end, by gender | Payroll register, personnel records | B8 |
| Workplace accidents | Accident register and accident reports | B9 |
| Training hours per employee | Training plan and records | B10 |
| Environmental, ethics and procurement policies | Approved policy documents, code of ethics, certifications | B2 |
For emissions, the method is the GHG Protocol: you will find the steps, with a complete example, in the article on calculating Scope 1 and 2.
Why every questionnaire seems to start from scratch
If the data exists, why does every questionnaire take days of work? Almost always for the same reasons.
- Scattered data. Bills sit with accounting, headcount with HR, maintenance contractor reports with production, policies in a shared folder.
- Questions phrased differently. “Total energy consumption”, “purchased energy”, “MWh per site”: the same figure, asked in three ways.
- Sources you cannot trace. Last year's number is there, but nobody remembers which file it came from.
- Inconsistent answers. Different clients and banks receive slightly different numbers for the same indicator.
The solution is not to fill in questionnaires faster, but to keep your data in one place and be able to find it, with its source, when a question arrives.
How an AI agent for sustainability works
An AI agent for sustainability is an assistant that works on your company's sustainability data and documents: it understands questions in natural language, looks for the answer in the data on the platform and shows where it comes from. It is different from a generic chatbot, which does not know your consumption and gives generic answers.
- Your data is in one place. Bills, invoices, registers and reports are uploaded to the platform and linked to sites and years.
- Ask the question as you would ask a colleague. You can paste the questionnaire question as it is, without knowing how to write prompts.
- Get the answer with its source. Every number comes with the document or calculation it is based on.
- Check and send. The decision on what to disclose stays with you.
EVA, the Envify Virtual Agent, does exactly this: it helps you answer EcoVadis questionnaires, requests from clients and lead companies in the supply chain and questions from banks, answers questions on regulations and methodology and analyzes emissions data, such as trends, comparisons between sites and consumption anomalies.
It works with the same safeguards as the platform: data hosted and processed in the EU, in compliance with the GDPR, access limited to your organization and a complete log of every answer and its sources.
An example: three questionnaire questions put to EVA
Here are three questions taken from real questionnaires, put to EVA exactly as they are. The answers are illustrative: the numbers are those of the sample company.
| Questionnaire question | What EVA returns |
|---|---|
| “What are our Scope 1 and 2 emissions for the last year?” | 312 tCO₂e in 2025, broken down by source and by site, with the calculation and the bills it is based on |
| “What share of renewable energy did we use in 2025?” | 38%, based on the consumption recorded by source, with the documents that support it |
| “How many employees did we have at year end, by gender?” | 48 employees, 21 women and 27 men, with the document the figure comes from |
In a few seconds you have the answer and the evidence to attach. And next time, when another client asks for the same figure in different words, the answer will be the same.
What stays with people
An AI agent takes away the manual searching and checking, not the responsibility. Three things stay with people.
- Checking. Every number has its source: check it before sending.
- Deciding what to disclose. Some answers require judgment, for example on targets and future plans.
- Avoiding claims you cannot prove. An environmental claim you cannot document is a risk, even in a questionnaire. Learn more in the article on how to avoid greenwashing.
EcoVadis questionnaires, client requests and bank questions largely ask for the same data: emissions, energy, waste, people and policies. If your data is in one place, an AI agent for sustainability like EVA finds it, with its source, in a few seconds. You check and send, without starting from scratch with every request.
Try EVA for free
Want to see how EVA helps you answer your next questionnaire? Fill in the form: we will get back to you and show you EVA, for free and with no commitment. Try EVA for free.
