The United Nations 2030 Agenda set out 17 Sustainable Development Goals, 169 targets, and 231 indicators to measure global progress toward a more equitable, sustainable world. It is the most ambitious measurement framework in history. And it is being undermined by a problem that rarely makes headlines: a chronic, structural shortage of reliable ground-level data.
The 2025 UN SDG Progress Report makes the scale of the problem plain. Of the 169 targets, only 139 could even be assessed using available trend data. Of those, just 35% show adequate progress — 18% on track, 17% making moderate headway. The remaining 65% are stalling, regressing, or simply invisible because the data needed to measure them does not exist in sufficient quantity, quality, or timeliness to support meaningful conclusions.
This is not primarily a political failure. It is a data infrastructure failure.
The Specific Goals Where Data Fails
The OECD’s September 2025 report Mind the SDG Data Gaps identifies the patterns precisely. Over 40% of SDG indicators in OECD countries rely on data that is more than three years old — particularly for environmental and agricultural goals. Over 30% of targets in the Planet dimension lack sufficient data for either distance-to-target or trend analysis. Four goals — Gender Equality (SDG 5), Sustainable Cities (SDG 11), Climate Action (SDG 13), and Peace and Justice (SDG 16) — have trend data coverage below 30%.
These are not obscure indicators. They are the goals around which entire development strategies are built. And the reason so many of them lack adequate data has a common root: they require observation of real-world human behavior, environmental conditions, and social dynamics in physical spaces — in communities, markets, urban environments, and rural regions that national statistical systems do not have the capacity to reach at the frequency and granularity that meaningful measurement requires.
The UN itself has acknowledged this directly. The ESCAP Asia-Pacific SDG Progress Report 2025 noted that data gaps “leave some of the most vulnerable populations invisible in official statistics, limiting policymakers’ ability to address their needs effectively.” Visibility requires observation. Observation requires presence. And presence at the scale that global development monitoring demands has never been systematically organised.
Why Satellite Data and AI Cannot Close the Gap Alone
The SDG measurement community has invested heavily in remote sensing, satellite imagery, and machine learning-based estimation to fill data gaps without deploying people. These tools have genuine value for environmental monitoring — forest cover, water body changes, agricultural land use — where the signal is visible from orbit and interpretable by trained models.
But the goals with the deepest data gaps are not primarily environmental. They are social, behavioral, and institutional. SDG 5 on gender equality requires data on lived experience, economic participation, and intersecting inequalities that satellites cannot see and algorithms cannot reliably infer. SDG 11 on sustainable cities requires understanding how urban populations actually use infrastructure, access services, and navigate public space. SDG 16 on peace, justice, and institutions requires evidence of community-level governance, safety, and access to legal recourse that does not surface in digital exhaust.
These dimensions of human experience can only be documented by humans who are present, trained, and operating under structured observation protocols. The OECD’s own analysis notes that “the role of surveys and polls in capturing subjective indicators like trust and life satisfaction, offering timely and more disaggregated insights, has been increasingly discussed and tested for selected official SDG indicators.” The direction of travel is clear: filling the remaining data gaps requires human observation at scale.
The Data Quality Problem Underneath the Coverage Problem
Even where SDG data exists, its quality is increasingly under scrutiny. The 2025 SDG Index covers 167 of 193 UN member states — 26 countries are excluded entirely because more than 20% of their indicator data is missing. Of those included, the data underpinning many indicators is old, inconsistently collected, and produced under methodologies that vary enough between countries to make comparison unreliable.
For enterprise organisations operating across multiple jurisdictions — whether in research, infrastructure, logistics, retail, or development finance — this is a familiar problem. You cannot make reliable decisions against benchmarks you cannot trust. You cannot report progress to regulators, investors, or development partners against indicators whose measurement methodology is contested or incomplete. The SDG data quality problem is not just a UN problem. It is an enterprise data problem.
What Structured Observation Contributes
The Sentinel Watch methodology — deploying vetted human observers to document specific events and conditions under defined taxonomies, producing structured, timestamped, geolocated, quality-reviewed datasets — is directly applicable to the categories of SDG indicators that remote sensing and statistical surveys cannot adequately reach.
Programs can be scoped against specific indicator requirements: documenting access to services in underserved urban areas for SDG 11; recording environmental compliance behavior for SDG 12 and SDG 15; observing economic participation patterns disaggregated by gender and location for SDG 5 and SDG 10. The taxonomy is defined by the client — whether that is an enterprise sustainability team, a development finance institution, or a research organisation. The output is structured data that meets the provenance, consistency, and quality standards that serious SDG measurement demands.
This is the point at which Sentinel Watch’s commercial methodology and the UN’s development measurement agenda converge. Sentinel Watch’s .org operations work directly within that development mission. But the underlying observation infrastructure — the vetted observer network, the structured capture protocols, the quality pipeline, the MCP-accessible data delivery — is the same. Enterprise clients with sustainability reporting obligations, ESG commitments, or development-adjacent programs operate in a world where SDG alignment is increasingly expected by investors, regulators, and counterparties. Observation data that is traceable, consistent, and structured to enterprise standards is the foundation that makes that alignment credible rather than performative.
The data gaps are real. The 2030 deadline is four years away. And the tools to begin closing them — at least for the goals that require human presence to measure — already exist.
References
- UN Statistics Division. (2025). The Sustainable Development Goals Report 2025. unstats.un.org
- OECD. (2025). Mind the SDG Data Gaps: Insights from the OECD SDGs Hub. OECD Policy Insights on Well-being, Inclusion and Equal Opportunity, No. 20. oecd.org
- UNESCAP. (2025). Asia and the Pacific SDG Progress Report 2025. unescap.org
- Sustainable Development Report. (2025). SDG Index and Dashboards 2025. sdgindex.org
