At Climate Product Leaders, we know that climate action in product, design, and engineering is best leveraged “in practice”. Case studies are a powerful tool we leverage to make our Playbook more practical and applicable. These examples illustrate what climate-aligned product work looks like in action. Case studies provide social proof of how peers in similar organizations have addressed climate challenges, such as improving efficiency, retiring features, or balancing sustainable design with performance; they’re more likely to act.
We always wanted to include case studies from the very beginning, even for our V1 Playbook release in 2023, but we lacked the resources to properly collect and develop them. We decided then that comprehensive case studies would be a key piece of V2, allowing us to provide the practical, real-world examples that make climate-aligned product practices more tangible and actionable.
Thanks to support from a CAT mini-grant, we were able to dedicate the necessary resources to this case study initiative. The grant enabled us to conduct thorough research, interviews, and documentation that would have been challenging to pursue otherwise, making this comprehensive collection of real-world examples possible.
Our goal was to source case studies to cover all chapters and key themes of the playbook while providing a framework for learning. We wanted to ensure that every major concept in our Playbook had real-world examples that product and engineering teams could relate to and learn from. By collecting diverse stories across different industries, company sizes, and climate challenges, we aimed to create a comprehensive library that would serve as both inspiration and practical guidance for teams embarking on their climate-aligned product journey.
Our Process
While there was initial interest when we launched our call for case studies, the quality of submissions was initially low. Most respondents simply wanted to promote their products rather than share genuine product stories or climate impact data.
With limited quality submissions, we shifted to direct outreach through our network, including peers through the Green IO Podcast & Conference series and our Climate Product Leaders Ambassadors. They helped us identify potential case studies and leveraged their networks for introductions.
Our process evolved from initial outreach to exploratory discussions, then to first drafts, iterations, and final approvals. We aimed to cover diverse industries, company sizes, product roles, seniority levels, and geographic regions. Ultimately, we found the quality examples we needed through our network- organizations that had genuine climate-focused product stories with measurable impact data to share.
What We’ve Collected
Outreach & Collection:
- Organizations contacted: 20+
- Case studies collected: 20 total
- Ready to publish: 10 case studies
- Published: 9 case studies
Key Themes/Focus Areas Covered:
- Data optimization: Minimizing data transfer, reducing page weight/complexity (Brussels Environment, Raptor Maps, Leboncoin)
- Sustainable AI / ML: LLM optimization, server management (Salesforce, Amazon, Baker Hughes)
- Eco-design approaches: Soft disabling, ethical design principles (Vincent Offroy, Designersethiques / French Government)
- Infrastructure optimization: Sustainable hosting, automated server decommissioning (GSF-Vestas, Illumina)
- Traffic / performance optimization: API efficiency, cloud storage solutions
Case Study Deep Dives
We are highlighting the following three case studies because they represent diverse contexts, public sector transformation, B2B SaaS optimization, and enterprise AI implementation, while demonstrating measurable climate impact through different approaches: eco-design principles, performance optimization, and sustainable AI development.
Case Study 1: Brussels Environment
Brussels Environment, the public environmental agency of the Brussels-Capital Region, set out to revamp its website with accessibility and eco-design in mind. They engaged experts early, set clear performance and sustainability KPIs, and benchmarked progress using tools like Greenspector.
Their efforts paid off: the average page CO₂ emissions dropped from 1.36g to 0.25g, and the Greenspector eco-score improved from 64 to 79.6. Data transfer and HTTP requests were drastically reduced, resulting in improved UX and sustainability outcomes. The process highlighted key lessons: start early, engage experts, and balance UX, tech, and sustainability. Though there were unexpected costs, the initiative increased internal awareness and demonstrated leadership across the ecosystem.
Case Study 2: Raptor Maps
Raptor Maps, a software company serving the solar industry, faced challenges as its platform scaled with the exponential growth of solar assets. Their flagship product, Raptor Solar Sentry, was struggling with performance due to the high volume and density of data, especially on mobile devices in low-connectivity environments.
Their team introduced progressive loading, just-in-time networking, aggressive caching, and an offline-first mobile experience. These changes saved 99.25% of the energy used when loading data and improved load speeds by 133x. Users in remote areas could now access the app, and the company avoided product obsolescence. Engagement soared, with over 200% growth year-over-year in map usage, and the product became more valuable, efficient, and climate-aligned.
Case Study 3: Salesforce
Salesforce faced the challenge of developing and deploying generative AI while staying true to its core values, including sustainability. The Large Language Models (LLMs) that power generative AI require enormous compute resources, resulting in significant environmental impacts like carbon emissions, water depletion, and resource extraction. As AI energy consumption grows exponentially, Salesforce needed to balance the compute requirements of LLMs with its sustainability goals.
Their approach focused on five key strategies: developing domain-specific models optimized for efficiency (like CodeGen 2.5, which performs as well as larger models at less than half the size), using efficient hardware like Google’s TPU v4 (2.7x more efficient than TPU v3), choosing low-carbon data centers that emit 68.8% less carbon than global average electricity, implementing sustainable prompt engineering techniques, and tracking usage with rate limits to prevent excessive resource consumption.
The results were impressive: pre-training Salesforce’s AI models resulted in 48 tCO2e, 11 times less than the emissions of GPT-3. They achieved significant energy conservation through efficient hardware and optimized models, while smaller models proved more cost-effective, easier to fine-tune, and faster to operate. The initiative demonstrated that powerful AI can be created with dramatically smaller environmental footprints without sacrificing performance.
What We’ve Learned from the Case Study Process
Gathering case studies isn’t always easy. Many organizations hesitate to share due to greenwashing concerns, confidentiality, or bandwidth. But when we do collect them, the value is immense.
- These stories exist! They can be challenging to source!
- They are high-value tools for enablement and inspiration
- They reveal both patterns and nuance across organizations
- This is only the beginning. We’re just getting started
Key Takeaways from the Case Studies
- Start early: retrofitting is harder and costlier
- Change is hard: building awareness takes time
- Engage experts: sustainable product work is a team effort
- Balance is key: eco-design requires trade-offs across UX, business, and tech
- Simplicity scales: simplifying UX often improves performance and sustainability
- Demonstrate leadership: successful examples build momentum for others
What’s Next
To learn more, please check out the full collection of case studies.
If you have a story to share or want help surfacing one from your team, we’d love to hear from you.