A business can receive a precise carbon footprint, water footprint or waste total and still be no closer to knowing what to change. The number is useful, but it is an outcome. It describes the consequence of thousands of decisions about materials, equipment, schedules, suppliers and habits, made long before the number was calculated.
Many businesses therefore manage environmental measures as reports: a total is calculated, compared with last year and given a target. The target may be sensible, but a target is not a control system. Improvement comes from moving one level upstream and asking which variables produce the footprint, which of them the business can actually change, and who makes those decisions.
This article explains how to establish a useful baseline, decompose an environmental total into its drivers, distinguish the kinds of measures that matter, assign drivers to the people who control them and check that changes actually caused the improvement. The same method applies to energy, water, waste, materials or emissions.
Start with a baseline
Organisations rarely lack ideas for environmental improvement. What they often lack is a way of deciding which ideas matter. A textbook treatment of sustainability in institutions by Jay Withgott and Scott Brennan describes audits as the starting point for this reason: a quantitative baseline shows what the organisation consumes, emits and discards before any initiative is chosen.
A baseline does three jobs:
- It reveals materiality: which resource flows or activities dominate the total.
- It supports prioritisation: competing ideas can be compared against a common starting point.
- It creates accountability: improvement can be judged on evidence rather than narrative.
The first baseline does not need to be perfect. It needs to be good enough to tell the large items from the small ones. Utility bills, purchasing records, waste invoices, production records and maintenance logs are often enough to start.
From outcome to drivers
A useful way to structure the analysis is a chain:
outcome → contributions → drivers → decision owners → operating response → verification
- The outcome is the total being managed, such as annual electricity use or tonnes of waste.
- Contributions are the activities or areas that make up the total, such as particular machines, processes, products or sites.
- Drivers are the variables that cause each contribution to rise or fall, such as idle running hours, set-points, scrap rates, product mix or delivery frequency.
- Decision owners are the people who can change each driver.
- The operating response is the specific change each owner makes.
- Verification checks whether the driver changed and whether the outcome followed.
Published research illustrates the approach. A study of water use on Irish dairy farms found that a complex water footprint could be predicted reasonably well from a small set of farm variables, such as grass grown, feed concentrates, imported forage and metered water use. A study of the carbon footprint of a large city’s roads broke the total down by life-cycle stage and found that construction materials and maintenance accounted for most of it, which pointed attention towards design, specification and maintenance decisions. The specific results belong to those settings. The method is general: a complex total becomes manageable when it is traced to a few drivers.
Building a simple driver tree
A driver tree is a diagram that breaks an outcome into its contributions and then into the variables behind each. It can be drawn on one page. At the top sits the total, such as annual electricity use. The next level lists the main contributions, such as production equipment, compressed air, cooling, lighting and offices. Beneath each contribution sit its drivers: for production equipment, perhaps running hours, idle hours, product mix and set-points; for compressed air, leaks, system pressure and demand from tools.
Building the tree with the people who run the operation is valuable in itself. Operators and maintainers often know which drivers matter, such as the machine that is left running over lunch or the air line that hisses all weekend, long before any data confirms it. The tree also shows where data is missing, which helps decide where a few weeks of measurement would be most useful.
Choose drivers that earn attention
Not every correlated variable is worth managing. A driver deserves attention when three conditions are met:
- It materially influences the outcome.
- It can be changed through a real decision by someone in the business.
- It can be measured reliably enough to manage against.
Be careful with correlation. A pattern in the data does not prove that changing one variable will change the other. The dairy study, for example, found only a weak relationship between farm profitability and a lower water footprint, which suggests that productivity may influence both rather than one causing the other. Test causal assumptions before building targets on them.
Stock, flow and intensity
Environmental measures come in three forms, and they can move in opposite directions:
- Stock measures describe accumulated quantities, such as stored waste or materials embodied in equipment.
- Flow measures describe absolute amounts per period, such as annual kilowatt-hours or litres of water.
- Intensity measures relate a flow to output, such as energy per kilogram of good product.
A business can improve intensity while its total flow rises, simply by growing. Both views are legitimate, but leaders should know which one their strategy aims to change, and report both so neither hides the other.
Give each driver an owner
A footprint without decision owners creates diffuse accountability. Drivers usually sit in existing roles:
- Design and engineering own material choices, equipment specifications and set-points.
- Purchasing owns supplier choice, packaging and material specifications.
- Production owns schedules, idle running and changeovers.
- Maintenance owns leaks, equipment condition and control settings.
The aim is not to give an environmental coordinator authority over every process. It is to embed relevant environmental variables into the decisions that already control the system.
The footprint you influence but do not own
A business’s largest environmental influence may sit outside what it owns and operates. A design consultancy chooses materials and specifications that shape how products are made and disposed of. An engineering firm’s design decisions set the energy use of equipment for decades. A manufacturer’s purchasing requirements can change practices across a network of suppliers. A useful way to think about this uses three levels:
- Control: activities, equipment and processes the business directly manages.
- Influence: designs, specifications, supplier choices, advice and customer decisions the business can materially shape.
- Exposure: environmental conditions that affect the business even though it has little control, such as climate, water availability or regulation.
This model prevents two errors: ignoring indirect effects, and claiming responsibility for outcomes the business cannot deliver alone. For influenced outcomes, identify the mechanism of influence: setting a specification, choosing a supplier, attaching a condition to a contract or advising a client. Where the mechanism is clear, shared measures can be set. Where it is weak, avoid making claims. When asking suppliers to meet environmental requirements, keep them proportionate. Performance data or practical support may achieve more than demanding formal certification from very small suppliers.
Prioritise by impact and control
A simple matrix helps decide where to start:
| High control | Low control | |
|---|---|---|
| High impact | Start here | Work through suppliers, design or partners |
| Low impact | Quick wins, mainly for engagement | Note and monitor |
Low-impact actions can matter for culture and engagement, but they should not consume most of the effort.
Work upstream where you can
For many environmental issues, interventions further upstream have more leverage. Recycling manages waste after it is created. Process changes prevent it. Product and packaging design prevent it earlier still, and purchasing can avoid bringing problem materials in at all. For energy, a similar order applies: remove unnecessary demand, improve efficiency, improve controls and maintenance, and then consider changing supply. Upstream changes usually need more coordination between teams, which is why they are often neglected.
Make it normal work
Improvements last when they become part of normal operating routines rather than campaigns. That means building them into standard work, schedules, maintenance plans, purchasing specifications and design reviews. Technology helps, but efficient equipment still depends on controls, maintenance and people. A new building or machine designed to perform well can lose its advantage within months if nobody owns its settings and upkeep.
Watch for burden shifting
Before turning a driver into a target, ask what could get worse if the target is pursued hard. Research on cropping systems, for example, found that a practice improving yield and several water and carbon measures relied on plastic mulch that raises its own pollution concerns. In a factory, reducing energy by running equipment more slowly might increase labour cost or reduce output. Set guardrails for cost, quality, safety and other environmental effects. The article on avoiding burden shifting covers this in more detail.
Drivers as leading indicators
Environmental totals are lagging measures. They arrive monthly or annually, after the decisions that caused them. Driver measures can be leading indicators: they change first and give earlier warning. Heater-on idle hours, open leak tags, scrap rates, cleaning cycles per week or deliveries per order can be checked daily or weekly by the people who control them. When a leading indicator drifts, the owner can act before the monthly bill or annual report shows the effect. A short weekly review of three or four driver measures often does more for performance than a detailed annual report.
Verify causality
After a change, check two things: did the driver move as intended, and did the outcome respond as expected? If the driver improved but the outcome did not, the causal model is wrong or incomplete, and it should be revised rather than the target intensified. Adjust for volume, product mix and weather where relevant, so improvements are not claimed for changes the business did not cause.
A worked example
This is an illustration. A small plastics moulder spends about $180,000 a year on electricity and wants to reduce both cost and emissions. Its first instinct is to replace lighting and buy a more efficient compressor.
The owner builds a baseline from a year of electricity bills and two weeks of logging on the main circuits. It shows that presses account for about 55% of use, compressed air about 18%, chillers about 12%, lighting about 5% and offices and other loads about 10%.
The team traces the large contributions to drivers:
- Presses: barrel heaters left on during breaks and long changeovers; many short production runs requiring repeated heat-ups.
- Compressed air: leaks, and system pressure set higher than any equipment needs.
- Chillers: set-points lower than the process requires.
Owners are assigned. The production manager owns a shutdown routine for breaks and changeovers and a schedule that groups short runs. Maintenance owns a quarterly leak survey and repairs. The engineer owns a review of pressure and chiller set-points.
The business tracks three kinds of measure: total kilowatt-hours (flow), kilowatt-hours per kilogram of good parts (intensity) and leading indicators such as heater-on idle hours, open leak tags and system pressure.
After six months, energy per kilogram of good parts has fallen by about 14%. Output has grown by about 6%, so total electricity use has fallen by about 9%. The lighting upgrade still happens, but later and with a clearer view of its small share. The business reports both the intensity and total figures, so neither hides the other.
How this applies to a small Australian business
Small businesses often have good data available without realising it:
- Use your bills: electricity, gas, water and waste invoices give a baseline.
- Ask your electricity retailer or distributor whether interval data from your meter is available, to see when energy is used.
- Log key equipment for a few weeks if bills are not detailed enough.
- Trace the largest totals to a few drivers.
- Give each driver an owner in an existing role.
- Track flow, intensity and a few leading indicators.
- Build changes into standard work, maintenance and purchasing.
- Make claims carefully: environmental claims must be accurate and supportable under the Australian Consumer Law, and the ACCC publishes guidance.
The articles on sustainability as an operating choice and choosing your decision measures before the design cover related ideas.
Signals worth watching
- Environmental reporting that cannot be traced to operational owners.
- Many small initiatives with no ranking by impact.
- Improvement claims without a baseline.
- Intensity improving while total use rises unnoticed.
- Efficient equipment performing poorly because controls or maintenance are weak.
- Waste measures improving only because material moved to another category.
- Purchasing decisions undoing operational gains.
Common mistakes
- Starting with solutions before measuring.
- Counting initiatives rather than impact.
- Managing the total without identifying drivers.
- Confusing correlation with cause.
- Reporting intensity or totals alone.
- Leaving drivers without owners.
- Running campaigns instead of changing routines.
Frequently asked questions
How detailed should our baseline be? Detailed enough to separate the few large contributions from the many small ones. Start with bills and records, and add measurement where the largest uncertainties sit.
What if we cannot measure a driver directly? Use a reasonable proxy, such as machine running hours for energy, and improve measurement over time where the driver matters most.
Should we set targets on totals or on drivers? Both have a role. Totals show the outcome that matters. Driver measures show whether the actions that should improve it are actually happening, and give earlier warning.
Who should coordinate the work? One person should own the baseline, the driver tree and the review rhythm, but the drivers themselves should be owned by the people who control them in normal operations.
How often should we review? Monthly for leading indicators, and quarterly or annually for totals, adjusted for volume and product mix.
Questions to ask
- Which few drivers explain most of our main environmental total?
- Who can change each of those drivers, and when?
- Are we tracking totals, intensity or both, and have we confused them?
- What could get worse if we pursue our current target hard?
- Which of our measures describe the past without informing any decision?
- How will we know that an improvement came from our actions rather than from volume, weather or product mix?
Bringing it together
An environmental footprint makes an invisible consequence visible, but its value begins after measurement. Establish a baseline, decompose the total into contributions and drivers, choose drivers that are material, controllable and measurable, give each an owner, track flow, intensity and leading indicators, build changes into normal work and check that results were caused by your actions. A business that can explain why its footprint moved is far better placed than one that can only report that it moved.
Source: KEVOS notes, drawing on J. Withgott and S. Brennan’s textbook treatment of institutional sustainability, and published research on dairy farm water use, urban road carbon footprints and cropping practices. Examples and figures in this article are illustrations.
