Mapping the ESG Data Gap in Deutsche Telekom's Supply Chain

Context

Deutsche Telekom wanted to explore AI-supported ESG supplier screening for a complex, hard-to-measure supplier group. The underlying goal was greater precision in Scope 3 accounting: much of the upstream footprint is calculated using industry-average emission factors, and Telekom wanted to understand whether supplier-level data like reduction targets, base years, target years and reported emissions could be sourced at scale to replace averages with company-specific figures. Alongside existing data sources like CDP, this meant assessing the sustainability maturity of a long tail of smaller German suppliers, many of which had little to no public ESG disclosures.

Challenge

This supplier category is genuinely difficult to screen. Public sustainability documentation is sparse for smaller suppliers, and manual research across hundreds of companies is slow and resource-intensive. Telekom wanted to find out how far automated, public-data screening could go in a segment like this, and where its limits would show up.

Approach

Briink and Telekom worked iteratively rather than as a one-off delivery. Over the course of the engagement, the two teams tested and refined:

• Public-data extraction across more than 500 supplier screenings
• Questionnaire logic and structured data handling
• Benchmarking and table output formats
• Excel export workflows
• Output delimiters and formatting details surfaced through real usage

Every round of feedback from Telekom was treated as input into the product itself, not a one-off fix.

What Telekom valued

Three things stood out to Telekom through the pilot:

Speed of feature development. Feedback moved from conversation to shipped product change within days. As Telekom put it: "The speed at which Briink turned our feedback into working product improvements was exceptional. We saw changes land on the platform within days."

Collaboration and responsiveness. Open questions were resolved quickly, and Telekom's input visibly shaped the direction of the product. "The Briink team was highly engaged throughout the pilot. Open questions were clarified quickly, and we felt that our feedback genuinely shaped the product."

Durable, scalable product functionality. The improvements built during the pilot were not bespoke, throwaway work. "The pilot did not just produce a one-off analysis. It helped develop platform functionality that can support more scalable enterprise screening and benchmarking workflows."

Data quality and outcome

Across more than 500 supplier screenings over the course of the pilot, Briink was able to identify relevant ESG or sustainability documentation for roughly a third. Importantly, this was not a gap in extraction accuracy: where relevant public sources existed, Briink's extraction held up reliably. The limiting factor was that for a large share of smaller suppliers, no public sustainability documentation exists at all.

That distinction turned out to matter more than raw coverage. For every supplier without available documentation, Telekom now had a clear, confirmed answer rather than a lingering open question, telling the team exactly where research would be futile and where to focus supplier outreach instead. As Alexander Freudenberg, ESG Strategy Manager, at Telekom Deutschland GmbH noted: "Where relevant public sources were available, the extraction quality was strong. The pilot also helped us understand where public-data screening reaches its limits."

For Scope 3 accounting, that is a strategically useful finding in its own right. It tells Telekom where supplier-level emission factors can realistically replace industry averages, and where other routes will be needed instead, such as direct supplier engagement, purchased datasets, or lighter-weight disclosure programmes. Rather than investing in a broad rollout on an untested assumption, Telekom reached that insight in weeks.

Telekom now works with an evidence-based view of what public-data ESG screening can and cannot deliver in a difficult supplier segment at this point in time, while helping shape platform functionality built for broader enterprise use.