Building a startup product with 10 pilot customers and a boring tech stack

Executive overview

Carbon-intensive companies face growing pressure to disclose emissions but lack software that makes the process easy or financially attractive. Gravity Climate solved this by launching with a cohort of 10 pilot customers, building in real time against their actual needs, and converting over half to paying customers within four months.

Two compounding lessons shape their approach: niche down to a specific customer before building, and pick a boring tech stack so engineering calories go toward the product, not the tools.

Fall in love with the problem, not the solution — define your identity by what you're solving, not what you're building.

The pilot program approach

• Recruited 10 customers across different organisation sizes, including direct industrials and private equity firms
• Offered a three-month engagement: measure emissions, generate a report, no purchase commitment
• At launch, had only a few lines of code and many hypotheses — built in real time against customer questions
• Focused on scope one and scope two emissions measurement, shaped entirely by the first cohort
• Risk: a non-representative cohort builds a product for nobody else — mitigated by finding lowest common denominators
• Over 50% of the pilot cohort converted to paying customers at an early-partner rate
• Revenue within four months; case studies used to recruit future customers and sales reps

Lessons from ClearGraph on niching down

• ClearGraph's ambition was making data accessible to all business users — too broad
• A horizontal product built for everyone added value for no one
• Microservices architecture early on slowed the team significantly
• Lesson: niche down earlier, specialise the product, keep the architecture simple

Picking boring technology

• Startups have too few resources to experiment with unproven databases or novel architectures
• Experimenting with powerful new tools is "a luxury for people with more time and money to burn"
• Gravity's boring tech stack lets a small team scale to large data volumes and significant revenue
• Ships code dozens of times per day; fixes problems in minutes, not hours
• Predictable feature development because the team is never "fighting tools"

Hiring for urgency, agency, and customer obsession

• Urgency and agency: people who take a high-level objective and drive results without hand-holding
• Customer obsession: self-awareness to recognise when a rabbit hole isn't serving the user
• Look for candidates who spotted a problem on their own, devised a solution, and shipped it
• Speed of execution can be learned, but for some people it is innate

Engineer as product person

• Engineering as an identity can be limiting — the job is to build a useful product, not software
• Maximise the rate of change of product to get the most attempts at building something useful
• Software engineers should think of themselves as product people who build product through code
• A product mindset means falling in love with the problem and the person who has it
• Carbon accounting seems dry, but deep engagement with the problem makes it interesting