There’s an Automated System for That — and Your Team and Auditors Will Notice the Difference
The easy fix is to quickly set up some automation and get the data out. This might be through an OCR tool or AI agent. The method is to find a vendor, set up the tool, get the data and move on to the next vendor. Every vendor is a thread and hopefully the sum of the threads add up to a complete report.
But customers come to GLYNT.AI frustrated with this method. They find their tooling constantly breaks or introduces errors, even with AI Agents. The source files keep changing, there are always new vendors and layouts change too. Plus, the LLMs keep changing! This too breaks automated flows.
One might easily spend 100 hours setting up a system and 40 hours per month maintaining it. Plus inaccurate data will creep in, perhaps at a 5% – 10% error rate. Further, with so many separate threads of workflow, it is difficult to ensure that results are coherent.
Threads Approach to Sustainability Data
Here’s what sustainability teams get with the threads approach:
- The reported data looks pretty clean
- Each row is a separate set of code or hand keyed
- The total report is the sum of individual work streams, e.g. the threads
Here’s the report card for the Threads approach:
GHG Protocol Compliance Report Card — The Threads Approach
| Status | Data Quality Principle | Note |
|---|---|---|
| Not met | Boundaries of Reporting | Not clear how to check the data. |
| Not met | Complete Data | Worrisome that there could be partial data for some sites, but no way to check this. |
| Met | Accurate Data | Looks pretty good. |
| Not met | Consistent Data | Very difficult to check, as the incoming files vary so much each month. Hard to manage the inventory of threads. |
| Unclear | Relevant Data | The data is done and will be used in reporting, but it may not be ready for use in operations. |
| Not met | Reliable Data | Would another person, using the same method, arrive at the same results? Hard to say. |
| Not met | Sufficient Time to Build Insights and a Strong Narrative | This is not a GHG Protocol principle, but it is certainly on everyone's mind. |
Act Two
Now consider a different method, one based on a system. Here’s what the system delivers:
- Double Verification. Every file coming in and out of the system is tracked, leading to 100% compliance
- Four-Eyes. Each month has a review and approval process so that two people approve the data, just like in financial accounting
- Read Reports. Sustainability managers have a best-in-class data quality test and reporting system. Instead of checking endless rows of data, managers read the accuracy, completeness and gap reports that show system performance
- Ready for Operations. With automated updates to reporting software or customized monthly reporting, senior management sees the data in context and approves it for operational use
- Ready for Audits. With a complete system approach, more reports are added to the Audit Archive each month. The clean end-to-end audit trail reduces the time and cost of internal and external audits
The Systems Approach to Sustainability Data
Here’s what sustainability teams get with the systems approach:
- A small number of hours are used to set up a system, and countless hours are saved over the reporting year
- ‘Manager’ back into the sustainability job description
- The team delivers verified highly accurate data
- A monthly review process solves issues before they drag on, just like financial accounting
- The end-to-end system and workflow meets compliance and audit requirements
GHG Protocol Compliance Report Card — The Systems Approach
| Status | Data Quality Principle | Note |
|---|---|---|
| Met | Boundaries of Reporting | Set up at the start. Verified every month. |
| Met | Complete Data | Set up at the start. Updated and verified each month. |
| Met | Accurate Data | Test results verify 100% accuracy. |
| Met | Consistent Data | Audit-ready review and approval each month. |
| Met | Relevant Data | Customized output file automatically loaded each month. Reviewed and approved. |
| Met | Reliable Data | Would another person, using the same method, arrive at the same results? Yes! |
| Met | Sufficient Time to Build Insights and a Strong Narrative | Whew! I'm a manager again! |
