Why AI in DAM Is not a question of tools – but of operational maturity

The study measures AI maturity on a five-point scale – from «Early Stage» to «Leading.» The finding is robust: maturity correlates more strongly with AI success than any other factor captured in the dataset. Maturity is not the only driver of success. But it is the only one an organization can control itself. Industry, size, and budget cannot be controlled. Foundations can.

This becomes evident precisely where AI works best in DAM today. Metadata tagging and enrichment, at 65%, are among the top use cases. Workflow automation follows at 63%, and search and discovery at 58%. These use cases have one thing in common: they require clean metadata, clear processes, and integrated systems – in other words, the fundamentals. Where these are missing, AI remains stuck in pilot status. The leap to productive operation fails.

Maturity is not defined abstractly in the study. It is measured – as operational readiness across five stages, from «Not Ready Yet» to «Built for AI.» The correlation with success is linear and clear: 

• Level 1 – Not Ready Yet: 23% success
• Level 2 – Early Readiness: 35% success
• Level 3 – Building Readiness: 53% success
• Level 4 – Ready to Scale: 60% success
• Level 5 – Built for AI: 74% success

The majority of organizations are currently at Level 3 – «Building Readiness.» That’s enough to experiment. It’s not enough to scale. It is precisely this leap that determines whether AI investments have an impact – or remain in the pilot phase.

From our DAM projects, we observe what typically triggers this leap: structured metadata that goes beyond what AI itself can extract from an asset. AI recognizes content in an image – but it doesn’t provide copyright information, usage rights, or associations with products or campaigns. Yet it is precisely this information that is required for assets to be further processed, automatically deployed, and used across system boundaries. Those who maintain it properly can scale. Those who rely solely on AI auto-tags remain in pilot mode.

The study breaks down operational maturity into four dimensions. They must work together. In the context of DAM and content operations, they specifically mean:

• People: AI champions, trained asset operations teams, clear accountability. One in four organizations today has no defined ownership for AI initiatives. This shows one thing above all: The pain is widespread – and the basics don’t just happen on their own. An active decision is required.
• Process: Standardized workflows, defined approval paths, governance for AI outputs. 83% of respondents cite a lack of governance as the biggest pain point. This is the most common bottleneck in the entire study – and a clear signal of where processes are visibly causing pain today.
• Data: Consistent metadata, well-maintained taxonomies, clean asset structures. 66% cite data and metadata quality as a barrier. Data hasn’t just become important now. But AI exposes weaknesses that manual processes have compensated for over the years.

• Technology: Integration between DAM, PIM, CMS, and the workflow layer. Native AI usage in the content stack ranges from 23% to 59% across systems. This is a sign of fragmentation – not a sign of maturity.

The study identifies three disciplines that set successful organizations apart from the average. They are all operational in nature – not technological:

• Strategy that is operationalized: Only one in four organizations today has a formal AI strategy. What matters is not its existence. What matters is what it governs: which use cases are prioritized, who makes decisions regarding AI deployment and asset approvals, and which KPIs are used to measure success. According to the study, a strategy that addresses these three points is the most effective self-managed lever.
• Governance that is supported: Formal AI governance alone changes little. Impact only emerges when it is combined with active AI champion networks and training programs. This combination raises the success rate to 63%—compared to an average of 54%.

• Measuring against business results: Organizations that measure AI success against financial KPIs achieve a 66% success rate. Those that do not measure at all have a 40% success rate. This is no coincidence. It reflects an attitude: What cannot be measured cannot be controlled.

In our projects, we see a very clear difference between «testing AI» and «operating AI.» Those who seriously deploy AI in DAM monitor AI-tagged assets, verify the quality of automatic translations, and regularly review – for example, annually using the same test assets – how the models are evolving. This is not a question of tool selection, but a question of operational discipline. It is precisely this discipline that distinguishes Level 3 (Building Readiness) from Level 5 (Built for AI).

Across the entire dataset, operational maturity correlates most strongly with AI success in DAM. It is also the only factor an organization can actively build. It cannot be bought. It emerges across People, Process, Data, and Technology – in that order of their typical bottlenecks. Anyone who wants to make an impact with AI in digital asset management by 2026 should not start with tool selection. Instead, they should begin with an honest assessment of their current state.

In our view, the next six months will determine which DAM programs have built these pillars – and which will remain stuck in the pilot phase. Those who honestly answer the question «Are the pillars in place?» before making any further AI investments will avoid projects that cannot scale.