How to Collect Climate Data from Borrowers Without Losing Your Mind

Every bank and SACCO we talk to says the same thing: "We understand the requirements. We know we need to report. But how on earth do we get climate data from our borrowers?"

It's a fair question. Your borrowers didn't sign up to fill out emissions questionnaires. Your relationship managers didn't train to ask about carbon footprints. And your core banking system definitely doesn't have a field for "Scope 1 greenhouse gas emissions."

But here's the thing: you don't need perfect data from day one. You need a system that starts collecting useful data now and improves over time.

The Data You Actually Need

Let's demystify this. For most Kenyan financial institutions, the climate data you need from borrowers falls into five categories:

Sector and activity: What does the borrower do? (Agriculture, manufacturing, transport, energy, construction, etc.) You probably already capture this — but how granular is it?
Physical location: Where are the borrower's operations? County-level is a minimum. Specific coordinates are ideal. Why? Because Kenya's climate risks vary enormously by location. A farmer in Turkana faces different risks than one in Kiambu.
Energy profile: What energy sources does the borrower use? Grid electricity, diesel generators, solar, biomass? This is the foundation for estimating their emissions.
Physical risk exposure: Is the borrower in a flood zone? Drought-prone area? Exposed to heat stress? You can estimate much of this from location data and existing hazard maps.
Transition readiness: Has the borrower taken any steps toward climate adaptation or mitigation? Solar installation, water harvesting, crop diversification, energy efficiency improvements?

A Practical Collection Strategy

Here's what actually works:

Phase 1: Use what you already have (Month 1-2)

You have more climate-relevant data than you think. Your existing loan files contain sector codes, business descriptions, and location data. Cross-reference this with:

Kenya's 47-county climate hazard profiles
Kenya Green Finance Taxonomy sector classifications
PCAF data quality guidance (to estimate emissions from proxy data)

This alone gets you to a baseline — rough, but real.

Phase 2: Add questions to existing processes (Month 3-4)

Don't create a separate climate questionnaire. Instead, add 5-10 climate questions to your existing credit application and annual review processes:

What county/sub-county are your operations in?
What's your primary energy source?
Have you experienced business disruption from weather events in the past 3 years?
Do you have any climate adaptation measures in place?
What percentage of your revenue comes from agriculture/manufacturing/transport?

Your relationship managers can ask these questions during normal interactions. No special training needed for the basics.

Phase 3: Build a digital data collection layer (Month 5-6)

This is where technology comes in. A platform-based approach allows you to:

Send standardised digital questionnaires to borrowers
Auto-populate fields from existing data
Calculate emissions estimates from proxy data
Score climate risk at the borrower level
Aggregate data for portfolio-level reporting

What About Data Quality?

The PCAF (Partnership for Carbon Accounting Financials) framework has a brilliant concept: a data quality ladder from 1 (best — verified primary data) to 5 (worst — estimated from sector averages). The point isn't to start at level 1. The point is to start somewhere and improve.

Quality 5: Estimate using sector and country averages. You can do this today with zero borrower input.
Quality 4: Use revenue or asset-based estimates with sector emission factors.
Quality 3: Use borrower-specific energy data to calculate emissions.
Quality 2: Use verified borrower-specific data.
Quality 1: Audited and verified emissions data directly from the borrower.

Most Kenyan institutions will start at quality 4-5. That's fine. What matters is having a plan to move up the ladder.

Stop Waiting for Perfect Data

The biggest mistake we see is institutions waiting until they have "complete" data before they start reporting. That's like waiting until you're an expert swimmer before getting in the pool. You learn by doing.

Start with proxy data. Add real data as you collect it. Improve your quality scores over time. Your first report won't be perfect — and that's exactly what regulators expect. What they don't expect is silence.