A practical method to identify, scope, and document IT assets for later carbon calculation. The objective is to build a reliable inventory that is complete, auditable, and repeatable.
- Establish what is in scope, where data will come from, and who is accountable.
- Produce an inventory that supports transparent calculation, assurance, and reporting (with an eye to CSRD evidence needs).
Define boundaries before listing assets:
- Organisational: legal entities, business units, and shared services included.
- Operational: owned, leased, and cloud-hosted IT; third-party services where you control configuration or usage.
- Temporal: reporting period (e.g., calendar year) and inventory freeze dates.
- Geographical: sites, data centres, colocation, and remote/home offices as applicable.
Tip: Document exclusions (and rationale) to aid assurance and avoid scope creep.
Group assets so later calculation rules are consistent:
- Compute: servers (physical/virtual), HPC, edge/IoT gateways.
- End-user: laptops, desktops, thin clients, mobiles, monitors, peripherals with notable load (e.g., docks, large screens).
- Network: routers, switches, Wi-Fi, firewalls, WAN/SD-WAN.
- Storage: arrays, tape libraries, object storage.
- Printing & Imaging: MFDs, scanners with dedicated power.
- Data Centre & Facilities (IT share): racks, PDUs, UPS attributed to IT, cooling allocation rules.
- Cloud & SaaS: IaaS/PaaS resources, managed databases, serverless, major SaaS with material usage.
- Development & Test: CI/CD runners, build agents, shared labs.
- Other Material Categories: AV/meeting tech, point-of-sale, ATMs, specialised medical/industrial IT where relevant.
Focus effort where it matters:
- Energy materiality: estimated kWh or spend (Pareto 80/20).
- Fleet size & refresh rate: large fleets (laptops, mobiles) are lifecycle-significant.
- Data availability: prefer categories with reliable metering or usage logs.
- Change velocity: rapidly scaling cloud or projects merit early inclusion.
- Regulatory relevance: items tied to CSRD disclosures (energy use, GHG, estimates, and methods).
Identify primary systems and evidence for each category:
- CMDB / Asset register (hardware attributes, purchase dates, serials).
- Procurement/ERP (models, quantities, invoices for embodied emissions proxies).
- Power & Environment (smart meters, BMS, DCiM, PUE/DCiE, UPS logs).
- Platform telemetry (hypervisor, Kubernetes, cloud billing/usage reports, API exports).
- Network & Print management (port statistics, page counts).
- SaaS admin (licences, active users, storage and compute metrics).
- HR & Workspace (headcount/location mapping for allocation keys).
Record the source, query, timestamp, and owner for audit.
- Inventory Owner: accountable for completeness and quality.
- Category Stewards (e.g., End-User, Cloud, Network): collect and validate data.
- Facilities: provide energy and PUE; agree allocation methods.
- Finance/Procurement: verify quantities and spend; supply lifecycle inputs.
- Sustainability: define thresholds, methods, and review evidence.
Use a simple RACI to prevent gaps.
- Confirm boundaries and exclusions.
- Map categories to data sources and owners.
- Extract raw lists (CMDB export, cloud inventory, meter IDs).
- De-duplicate & normalise (model names, asset tags, regions).
- Assess completeness using cross-checks (e.g., purchase vs CMDB vs AD).
- Classify each asset: category, subcategory, lifecycle stage metadata (purchase date, warranty, refresh policy).
- Attach evidence (file links, query IDs) for assurance.
- Score data quality and note assumptions (see below).
- Create allocation keys (e.g., by business unit, region, product).
- Freeze the inventory with versioning for the calculation phase.
- Fleet devices: sample by model cohort; extrapolate using purchase or MDM counts.
- Remote offices: sample representative sites; scale by headcount or installed load.
- SaaS: use active user metrics or billed consumption as a proxy.
Document methods and error bounds.
Score each field (0–3) to guide improvements and support CSRD disclosures:
- 3 – Metered/Measured (device or circuit meter; authoritative API).
- 2 – Verified Logs/Billing (platform usage, invoices).
- 1 – Engineering Estimate (nameplate power, vendor TDP, cohort averages).
- 0 – Assumed (insufficient data; flagged for remediation).
Record DQ score, assumption text, and responsible owner.
- Version-controlled inventory (timestamped exports).
- Change log for inclusions/exclusions.
- Quarterly mini-reconcilation with procurement and cloud billing.
- Evidence repository with retention aligned to audit needs.
- Keep method descriptions, data sources, estimation techniques, and limitations alongside the inventory.
- Maintain consistency year-on-year; when methods change, record the rationale and restate prior figures where material.
- Ensure internal control evidence: approvals, RACI, and review checkpoints.
- IT Asset Inventory (CSV/SQL/Sheet) with fields below.
- Boundary & Assumptions Log.
- Data Quality Register.
- Allocation Keys by business unit, geography, or product.
| Field |
Description |
| Asset ID / Source ID |
Primary key and originating system |
| Category / Subcategory |
From the taxonomy above |
| Make / Model |
Normalised naming |
| Quantity |
Count or capacity unit |
| Location / Region |
Site, country, or cloud region |
| Ownership |
Owned/leased/third-party |
| Purchase/Commission Date |
For lifecycle timing |
| Power/Meter Ref or Usage Metric |
kWh, hours, or platform usage field |
| Evidence Link |
Query, file, or meter reference |
| Allocation Key |
BU/region/product assignment |
| Data Quality Score (0–3) |
As defined above |
| Notes / Assumptions |
Free text for context |
Outcome: a clean, evidenced list of relevant IT assets that is ready for calculation, review, and reporting.