Improving Scope 3 data has become an established practice. For a large share of the corporate world, it is simply a given. Yet beneath all this activity sits a question that almost no one is asking.
The state of play
CBAM and the forthcoming Digital Product Passport call for increasingly detailed, product-level data. Companies with validated Science-Based Targets are pressing their suppliers for GHG numbers. Reporting platforms such as EcoVadis are widely used, not least to help populate CSRD disclosures. The result is a great deal of genuine work, much of it directed at a single objective: obtaining better supplier data.
The questions everyone is asking
Unsurprisingly, the conversation has organised itself around two practical questions: How do we obtain the data? And will we use the carrot or the stick? Do we demand data from suppliers, or engage and support them? These are all sensible questions, and most projects (and most vendor propositions) are built around those questions.
But are these the right questions? It is worth pausing before accepting that framing. Before asking how to collect primary data, a company should ask whether it needs it at all. In my view, three questions come first: is the data required, do the tools actually reduce the burden, and is it needed at all?
First, is it required? Primary data is not mandatory for your inventory. Companies are required to account as accurately as is practicable and to disclose their methodology and data sources. But neither the GHG Protocol, nor the SBTi, nor CSRD requires the use of primary supplier data across Scope 3. Spend-based estimation is permitted. Specific regimes do compel detailed data for the goods they cover, like CBAM and the DPP. CBAM effectively requires actual embedded emissions data for imported covered goods, with punitive default values as the fallback, and the Digital Product Passport is product-level by design. But having this level of detail for your complete Scope 3 (category 1, purchased goods and services) isn’t required.
Second, does it actually reduce the burden? The tools promise less work and an easy path. But the two problems that are causing headaches are rarely the ones they tackle:
- Revising the base year. Better current-year data is of limited value for tracking progress unless the base year is restated on the same basis. Suppliers readily provide current-year figures, but base-year data of equivalent quality rarely exists, so the restatement falls back on proxies. The base year becomes the weakest point in the series.
- Hybrid Scope 3 Category 1 inventories. As suppliers migrate to primary data at different speeds, Category 1 becomes a permanent mix of primary and spend-based figures. Managing that mix of data while avoiding double counting and maintaining consistency across years is often ongoing manual work that no collection tool removes.
Third, and most fundamentally: is it really needed? That depends on the business and its reduction strategy, not on the market’s enthusiasm for data.
A decision, not a default
A useful question to test the need is : does a concrete reduction lever depend on this data? If improving the data for a given supplier or category would not change a single decision or action, the case for collecting additional data is weak. Data that informs a reduction is time well spent.
For a B2C retailer selling complex, multi-material products to consumers, think of thousands of SKUs and a long supplier tail that will never report to a useful standard within time, primary product data across the value chain is often the wrong route. Continuing with spend-based accounting and setting an SBTi supplier-engagement target is a more reasonable path, because it reflects where real influence lies.
It is worth being clear that an engagement target is not an alternative to data; it is often a route towards it. Suppliers that adopt their own Science-Based Targets will, in time, hold their own emissions data, and can provide a far more representative figure than a generic spend-based factor. A practical step is a cost-of-sales emissions ratio: the supplier’s total emissions (Scope 1 and 2, plus upstream Scope 3) divided by its revenue, applied to your cost of sales with that supplier. This remains an allocation, but one grounded in the supplier’s actual emissions intensity rather than the sector average. A steppingstone between spend-based estimation and full primary data.
For other companies, particularly those sourcing homogeneous materials, data improvement is genuinely worthwhile, because better data feeds directly into identifiable reduction levers. For example, purchasing low-carbon materials and/or materials with recycled content.
The point is not that data improvement is unnecessary. It should be a deliberate decision, taken based on a company’s reduction options and climate strategy.
Be pragmatic and focus on what matters
Where data improvement does make sense, the following approach can be helpful. Most companies have an 80/20 supply chain: roughly 20% of suppliers account for around 80% of emissions. Concentrate on primary-data effort there, and hold the remainder on spend-based data. Map the tail of suppliers nonetheless, and revisit the boundary periodically, since suppliers move in and out of materiality as the business changes. Finally, sequence the work so that base-year revision and target updates are consolidated, for instance with the mandatory five-year SBTi review, rather than disrupting the time series every year.
The most important starting question
The question that often remains untouched is neither how to collect the data nor whether to use a carrot or a stick. It is the prior question, whether primary data serves the company’s strategy, together with the consequences, such as the base-year revision and hybrid inventories. Until that question is answered, better data is just a more precise way of measuring the same inaction.