Categorising your financial transactions helps identify activities that may produce greenhouse gas emissions. Sage Earth Carbon Accounting uses information from the transactions in your purchase ledger to estimate emissions based on the amount spent.

We’ve created a comprehensive set of business activity categories to apply to your transactions. These categories indicate the type of emissions generated and allow us to estimate each transaction’s impact.

When your accounting platform syncs with Sage Earth, AI automatically scans and categorises your transactions to estimate your carbon footprint and provide emissions insights. It assigns the most specific category possible, but reviewing and adjusting these categories helps improve accuracy and future results.

The AI model learns from categorisation across all users, so over time this process becomes faster. New transactions sync daily, so it’s a good idea to set aside regular time to check and categorise them. If AI cannot confidently categorise a transaction, it will be marked ‘Categorisation required’. These transactions require a category to be manually applied.

Transactions categorised by AI appear in the ‘To Review’ section. If a more specific category is available, you’ll see ‘Categorisation available’, which you can refine manually.

Once you’ve synced and categorised transactions and entered usage information into modules, Sage Earth Carbon Accounting will link your spend data (e.g., £100 electricity bill) with usage data (e.g., 300 kWh of electricity), counting it as a single business activity.

​
​

Your financial transactions are sorted into three tabs:

This tab shows new transactions that are uncategorised or partially categorised by AI.

Click View details next to a transaction to open the Transaction details panel, which displays:

This tab contains transactions categorised by AI.

This tab lists all transactions where categorisation has been confirmed. Once you confirm a categorised transaction, it moves here.

Bulk categorisation lets you apply the same category to multiple transactions or confirm AI‑suggested categories all at once. To use it, select the checkboxes next to the transactions you want to update and choose “Edit Categories”. In the “To Review” section, you can confirm AI‑applied categories by selecting the relevant transactions and clicking “Confirm Categories”.

Sage Earth Carbon Accounting includes general “unknown” categories, like “Unknown vehicle fuel”, for cases where specific details aren’t available. These can be used when you're unsure how to categorise a transaction more precisely.

Choosing an unknown category applies a less specific categorisation without specifying finer details. For example, under Vehicles > Diesel, you can select:

The first two options offer precise categorisation. Selecting Unknown Diesel applies a general diesel category without identifying the specific type, which will reduce the accuracy of your carbon footprint.

When applying an unknown category to a transaction, you may see a message indicating that more detailed categorisation is available. These transactions will remain in the “To Categorise” tab until a more specific category is selected.

If the AI model cannot suggest a category, or if you manually remove a categorisation, the transaction will remain uncategorised. These items are given a generic label, such as “Unknown electrical goods”, and will appear in the “To Categorise” tab. This prompts you to manually assign a more specific category to improve the accuracy of your carbon reporting.

When your accounting platform is connected and spend data is available, it is automatically linked with matching usage data to prevent double-counting. As usage data is higher quality data than spend data, the spend is overridden by usage.

This link is based on exact category matches. For example, if a fuel transaction is categorised as Diesel > Diesel (average biofuel blend), the usage data entered in modules should use the same category. This ensures both spend and usage are counted as a single business activity.