Skip to content

Build the dataset
your question needs.

Custom data extractionFull traceability

Custom datasets from sustainability reports.

You bring the question; we turn sustainability reports into the data to answer it. You choose the companies, years and metrics, we agree the definitions together, and then we extract the data. Every value comes with the original PDF and the page it came from, so you and your reviewers can check it.

Honestly, this is the work we enjoy most. Studies that used to take months of manual coding, like long time series or detailed qualitative disclosures, are now practical with AI. Our own extraction pipeline does the heavy lifting, and the evidence behind every observation stays available for review.

What is included

Consultation call

A free call with our research team. We’ll tell you whether it can be done, and if so, send a non-binding quote.

Structured data ready for research

The dataset we agreed on, with the original PDFs, page references and supporting passages.

Research support

We stay available for methodology questions after delivery, including when reviewers come back with questions.

Methodology

We started as researchers at Rotterdam School of Management, and we still work like researchers. Before extracting anything, we agree the sample and indicator definitions with you. We keep the source evidence and document our checks and limitations. And if the reports simply don’t contain what you need, we’ll tell you upfront. We also publish our own research and write about reporting methods on our blog.

We can write methodological notes for your paper covering the sample, definitions, extraction rules, quality checks and source evidence, and help you answer reviewer questions. What we don’t share are our exact prompts, pipeline details and model settings; those stay proprietary.

SRC-Bench v1

Measured extraction quality

In our published benchmark, 226 of 229 reported values matched expert annotations across seven indicators in 50 sustainability reports. That’s one sample, not a promise for every metric or report, but it gives you an idea of what to expect.

Read the benchmark report (PDF)
98.7%value accuracy in the evaluated sample

Getting started

Every project starts with your question. We agree the scope and definitions together before extracting the data.

Request quote

Tell us your research question, companies, years and metrics. Our research team offers a free consultation to check feasibility and help shape the brief.

Pricing is custom and depends on the scope, number of metrics and extraction requirements. You receive a non-binding quote before work starts.

Request quote

FAQ

When should I choose Data Solutions?

When the standard packages don’t answer your question. If you need specific metrics, your own definitions or a particular sample, we build the dataset with you.

How is my dataset produced and delivered?

First we agree the research question, companies, years, datapoints and output format. Then we extract the data from the reports and check it.

You get the dataset in the agreed format, with the source PDFs and a reference for every value. We walk you through the delivery so you can review the results and start working with them.

How is a custom project priced?

The first call with our research team is free. We’ll tell you honestly whether the project can be done, and if it can, you get a non-binding quote before we start. The price depends on the scope, the number of metrics and how hard they are to extract.