Real data makes every scenario count.
A decarbonisation plan can only be as reliable as the baseline it starts from. A target is one sentence, while a plan needs a measured starting point, a set of actions and a way to know which actions work.
Simulation, scenario planning and progress tracking all inherit the quality of that starting point. Raising it means bringing data from the supply chain into the baseline, so that the numbers describe the products a brand actually makes.
Measure first, then choose
A baseline is the reference every action is judged against. When a brand tests a lower-impact material, a different process or a new supplier, the result it reads is a difference: the scenario minus the baseline.
That has a consequence that is easy to miss. If the baseline is off, every difference is off with it, and the plan still looks precise on screen. A reduction of a given size only means something when the starting number describes the product that is really being made.
Why averages cannot steer a plan
An average describes a category of material, not a facility. A database value blends many plants, energy mixes and process routes into one figure, so it tells a brand what a material usually costs the climate. It cannot tell the brand what happens when it changes something.
Three limits show up in practice:
- Averages are flat across suppliers. Switching between two suppliers of the same material changes nothing in the model, so a real difference never appears.
- Averages do not react to effort. A supplier that moves to cleaner electricity or recovers process heat sees no change in its customers' numbers.
- Averages blur the hotspot. An impact concentrated in one step looks the same as an impact spread across the chain, so priorities get set by habit.
Database values still have a job. They let a brand start quickly and cover the whole portfolio. The point is to know where they stop being enough.
What primary data changes
Primary data turns the baseline from a description of a category into a description of your own supply chain. Three things follow.
The levers become real. Differences between suppliers, plants and countries show up in the numbers, so choosing a supplier or a process route becomes a decision with a measurable effect.
Effort becomes visible. A manufacturer that invests in efficiency or cleaner energy sees that work reflected in the footprints of the products it supplies. Database values cannot offer that incentive.
Plans can be checked. Targets and claims will increasingly be read against traceable sources, in the context of ESPR and the Digital Product Passport. A plan built on specific, sourced data is easier to defend than one built on averages.
Collecting primary data without stalling the programme
Start where the impact is concentrated, not where data is easiest to get. A request to every supplier at once tends to bring low response and uneven quality, while a narrow request to the suppliers behind the heaviest materials brings more.
A working loop looks like this:
- Build a first baseline from database values, and see which materials and processes weigh most.
- Ask the suppliers behind those steps for the inputs that drive them, such as energy, water, chemicals and output volumes, at facility or process level.
- Replace the database values for those steps only, and leave the rest untouched.
- Re-read the baseline, then repeat with the next group of suppliers.
Each pass improves the baseline where it counts, and the plan keeps moving between passes. Manufacturers also engage more readily when data is requested once, used in the model and returned as insight on their own hotspots.
From baseline to tested actions
A strong baseline matters because it is what every decision is tested against. In the SBP Simulation tool, a brand models a change on its real products, such as a different material composition or a process moved to another country, and compares the scenario with the baseline before the action enters the plan.
Results are shown per kilogram and for the volume purchased, with the reduction split by SKU and by collection category. A simulation never changes live data, so scenarios can be tried freely.









