Organisations continue to invest in AI, but broad returns are not keeping pace. The latest AI Momentum Survey from Dun & Bradstreet shows that 65% plan to spend more on AI over the next twelve months. At the same time, 28% report broad or strong returns, the same proportion as in Q3.
The survey is conducted quarterly among 10,000 businesses in 32 countries. The Q4 results primarily show how data readiness is linked to AI returns. Organisations with fully AI-ready data report broad results far more often than organisations whose data is partially ready.
AI continues to scale, but returns have yet to follow
The move from pilot to production is becoming more common. Now, 37% of organisations are scaling AI, compared with 34% in Q3. Yet for many organisations, returns remain limited to individual applications: 44% report returns within specific AI projects. Only 9% say that most pilots have been scaled up with clear results.
A successful application within one team does not automatically mean that the same value will emerge in other processes. As its use expands, AI must be able to reliably combine information from different systems. What is still checked or supplemented manually during a pilot requires a structured approach when scaling up.
Interesting read: The next phase of AI: from experimentation to ROI
Data readiness is linked to broader AI returns
The difference by data readiness is substantial. Of the organisations with fully AI-ready data 81% report broad or strong returns. Among organisations with partially ready data, that figure is 12%. The latter group also reports returns limited to individual applications far more often: 57%, compared with 16% of organisations with fully ready data.
The survey does not prove a causal relationship, but it does show a clear association. For organisations looking to scale further, this is a reason to also assess the underlying information. Can AI reliably combine data? Is it clear which organisation the data refers to? And does that information remain up to date when multiple teams work with it?
Interesting read: AI experiment shows why reliable business data makes the difference
The data foundation is only growing to a limited extent
As organisations expand their use of AI, fully ready data remains rare. The Q4 results show the following breakdown:
- 8% describe their own data as fully ready for AI, compared with 6% in Q3.
- 33% describe their data as largely ready, compared with 36% in Q3.
- 44% describe their data as partially ready, compared with 47% in Q3.
The improvement at the top does not mean that the entire market is progressing at the same pace. Making business data available is one step; consistently identifying, connecting and maintaining that information requires more. Duplicate customer records, outdated data and missing corporate group relationships can affect multiple processes and decisions simultaneously when scaling up.
Interesting read: From AI-ready to agent-ready: Is your data ready for Agentic AI?
From AI investment to lasting value
The quarterly figures show that further investment does not automatically go hand in hand with broader returns. As AI takes on a greater role, reliable business context becomes more important. An application must be able to determine which data belongs to the same legal entity, how companies are connected and whether the information is still up to date.
Master Data Management helps capture and maintain those connections across systems. At Altares Dun & Bradstreet, we support organisations with up-to-date, verified business information and insight into business relationships. This creates a data foundation that can grow alongside the role of AI within the organisation.
Would you like to discuss your organization’s AI readiness and what is needed to scale AI further? Plan een gesprek met onze experts.
