Artificial Intelligence's potential in enhancing marketing is limited until your data quality is up to par.
In the rapidly evolving world of artificial intelligence (AI), a recent industry report titled "State of Generative AI in the Enterprise (2025)" by Stack AI sheds light on a concerning discrepancy: while over 73% of enterprises have adopted AI technologies, only around 52% are realizing the results they had anticipated from their AI investments.
The primary culprit for this disconnect is a combination of technical, organizational, and operational obstacles. Issues with data quality, data silos, legacy system integration, talent gaps, governance challenges, and difficulties scaling AI projects from pilot phases to broader enterprise adoption are the main factors hindering enterprises from fully leveraging AI's potential and achieving expected returns on investment (ROI).
The report highlights that over 70% of global enterprises use AI in at least one business function, but the time to scale AI and achieve strong ROI often spans 6 to 12 months. Companies investing more than 5% of their IT budget in AI tend to report higher positive ROI (70–75%) compared to those investing less (50–55%).
Poor data quality and siloed data, legacy IT systems that are difficult to integrate with AI solutions, and a shortage of skilled AI personnel create barriers in converting AI pilot projects into full-scale, high-performing deployments. Governance and organizational readiness also impact success in realizing anticipated AI benefits.
One of the most overlooked aspects of modern marketing is identity resolution. Marketing teams must understand how data moves, where it breaks, and how to resolve inconsistencies at scale. Flawed data reduces accuracy, introduces bias, accelerates drift, and undermines customer trust. Data silos are a leading obstacle for many organizations, making it difficult to link behaviors across touchpoints.
To address these challenges, a focus on data hygiene is increasingly strategic for AI success. Most brands are focused on building a central data lake with a common identity across all of their data. AI is a system that learns by example and relies on the structure and reliability of the data it receives. Without orchestration, more data simply compounds the problem for AI.
The pace of AI investment is expected to double over the next two years. However, the main issue in AI performance is not model design or computing power; it's data quality. A persistent customer identity allows models to associate behavior with individuals, and this understanding is crucial for brands aiming to engage with consumers effectively.
In a world where 72% of consumers are more likely to engage with brands that understand their full identity, the importance of data quality for AI success cannot be overstated. 73% of enterprises have adopted AI technologies, and 92% of leading marketers consider first-party data essential to growth. As AI continues to permeate businesses, the focus on data quality will only grow more critical.
References: 1. Stack AI, "State of Generative AI in the Enterprise (2025)" (2025), Link to the report 2. EY, "Global AI Adoption Index 2020" (2020), Link to the report 3. Deloitte, "State of AI in the Enterprise, 2020" (2020), Link to the report
- To realize the anticipated returns on investment from AI, companies must address technical, organizational, and operational obstacles, with poor data quality and siloed data, legacy IT systems, and talent gaps being the main factors hindering AI's potential.
- The report also emphasizes the importance of data quality in the success of AI, stating that a persistent customer identity allows models to associate behavior with individuals, which is crucial for brands aiming to engage with consumers effectively.
- As AI continues to permeate businesses, the focus on data quality will only grow more critical, especially as 92% of leading marketers consider first-party data essential to growth. This underscores the importance of personal-finance, business, and technology sectors investing in data-and-cloud-computing and artificial-intelligence for their operations.
