Omnibus lets multi-site employers use estimates, proxies and secondary data for defensible Scope 3 Category 7 commute reporting.
Yes: for Scope 3 Category 7, you can now build reporting with estimates, secondary data, proxies, and internal records instead of treating a full staff survey as the default. That is the main takeaway from the article.
If you report under ESRS, this means you can take a model-first route when employee commute data is hard to collect across many sites. The article’s core point is simple: the job is no longer “survey everyone first”. The job is to build a clear evidence mix, separate source data from assumptions, and check the final EFRAG wording before I publish.
Here’s the article in one view:
Bottom line: if you run a large, multi-site employer, a model-first process is often easier to keep current than a blanket survey, and the article explains why in plain terms.
| Topic | Main point |
|---|---|
| Omnibus change | More room to use indirect data for Category 7 |
| Main reporting issue | Commute data is hard to keep current at scale |
| Best starting point | Internal data first, outside data second |
| Surveys | Useful, but not always the first step |
| Best fit for multi-site firms | Model-first with a clear audit trail |
| Final step before disclosure | Verify final EFRAG wording |
Read the rest of the article as a guide to building that evidence mix without making a full survey the starting point.
The big change here is flexibility. Under the simplified ESRS wording, you can build your Scope 3 Category 7 case using estimates, secondary data, proxies, and documented assumptions when that is easier to defend than trying to collect full primary data from every single site. That’s why the Omnibus matters in this area.
Treat this as a practical reading, not final guidance, until you verify the EFRAG wording. Before you rely on this interpretation, confirm the final EFRAG text. This is general information only and is not legal or tax advice.
If the final wording confirms this direction, you don’t need to depend on full primary-data collection at every site. You can combine internal records, secondary sources, proxies, and documented assumptions instead. For commuting data, that shift makes a model-first reporting setup far more workable.
Employee commuting is a strong fit for a model-first approach because direct collection is hard to keep complete and current across a large organisation. People move, work patterns change, office attendance shifts, and site-level data can go stale fast.
That’s why a flexible evidence mix becomes the practical default for Scope 3 Category 7.
Start with the records you already have in-house. Then fill the gaps with secondary inputs where direct data is missing. That shifts employee commuting from a data chase into a modelling exercise.
Your internal records give you the base layer for the estimate: employee headcount, site locations, and any travel or HR data already in your systems [1]. If you calculate anything from those records, label it separately from the source data.
Bring in external secondary inputs only when they improve coverage or make the model more current. Keep them clearly separate from internal records. It also helps to add a short note for each one that explains why you used it and what gap it fills.
When you document the model, separate data pulled from your systems from data inferred through assumptions or modelling [2]. That line matters. It shows exactly where judgement enters the estimate. Before you publish, verify the final EFRAG wording.
That mix of evidence is why a model-first approach tends to work better than a blanket survey.
Once you allow estimates and secondary data, the job changes. It’s no longer just about collecting more data. It’s about calibrating the estimate well.
Surveys can go out of date fast when attendance, headcount, or site patterns shift. And across multiple sites, another issue shows up: consistency.
One site team may run a survey in March, another in June. One may ask detailed commuting questions, while another keeps it short. That makes it hard to roll everything up into one company-level figure without mixing apples and oranges.
A commute model uses the same estimation logic across every site you run. That matters.
When attendance patterns or headcount change, you just update the inputs. The estimate changes with them. No need to start a full survey cycle again. You also get a steady baseline for repeat site updates, which makes multi-site reporting much easier to handle.
The table below compares the three approaches across the points that matter most for multi-site reporting.
| Dimension | Full Survey | Hybrid (Survey + Model) | Model-First |
|---|---|---|---|
| Data effort | High | Medium: surveys at selected sites, model elsewhere | Low: one model using internal and secondary inputs |
| Coverage | Varies by site and response quality | Strong where surveys are run, modelled elsewhere | Consistent across all sites |
| Update frequency | Low: a new survey cycle is needed | Medium: model updates, surveys refresh periodically | High: update inputs as workforce data changes |
| Consistency across sites | Variable | Moderate | High |
| Audit trail | Depends on rollout and documentation | Clear when survey data and model assumptions are documented | Clear when all inputs and assumptions are traceable |
In practice, surveys work best as calibration inputs inside a model-first process.
Before you publish, verify the exact EFRAG provision that supports this approach.
After you pick a model-first approach, check the final EFRAG text before you publish anything. The simplification may give you more room to use estimates and secondary data in Scope 3 Category 7. But you need to confirm the final wording before you rely on it in a public disclosure.
Flexibility doesn’t remove the need for rigour. Your documentation should show what you used, why you used it, and what each input can and cannot prove. It should also spell out the level of uncertainty that comes with each input.
Once the audit trail is clear, the next question is simple: which inputs can support the model across many sites? A model-first approach works well when commute patterns are structured enough to estimate from site-level data. Postal codes and shift patterns can give you a baseline that you apply in a consistent way across sites.
If the final wording supports this route, the next step is a repeatable modelling workflow. triply models real employee commutes from minimal inputs, such as postal codes and shift patterns, without surveying your full workforce. You can simulate emissions and cost impacts before you invest, and produce audit-ready Scope 3 Category 7 output for CSRD and ESRS E1 reporting.
Book a demo to see how it builds a Category 7 model from the data you already have.
