Generative AI is changing how organizations approach data analytics, but it doesn’t replace the need for strong data foundations.
It can accelerate how teams access data, uncover insights, and make decisions. But on its own, it doesn’t fix underlying issues like poor data quality, unclear use cases, or limited governance. Without those in place, generative AI can just as easily scale problems as solve them.
Early adoption entered businesses and organizations through accessible, creative use cases. Think image generation and content drafting. Teams could produce polished content in seconds and quickly test ideas with minimal effort. Now, a more advanced wave of generative AI tools is emerging. GenAI is moving into analytics, querying data, uncovering insights, and accelerating decisions that used to take days.
Still, generative AI isn’t plug-and-play. Strong analytics capabilities don’t come from the tool alone. They come from knowing where it fits, where it doesn’t, and how to apply it with intention.
Good Data Doesn’t Come Easy
For years, the analytics function has faced a familiar set of constraints. Data is available, but access is bottlenecked. Insights exist, but getting to them takes time. Decisions get made on incomplete information, not because the data doesn’t exist but because the process of surfacing it is too slow.
GenAI can change that, but it can also make existing problems worse if businesses aren’t careful. Before the tools can deliver, organizations need an honest look at what they’re up against:
| Poor data quality and consistency Incomplete, outdated, or conflicting data undermines any analytics tool. GenAI can do a lot, but it can’t fix bad data. Instead, it scales it. Flawed inputs will produce flawed outputs faster. | Unclear ownership Who is responsible for data quality? Who flags anomalies when something looks off? In many organizations, those questions don’t have clear answers. |
| Data silos When data is spread across various systems or departments, it can be difficult to access. Teams may have limited information and make unintentionally uninformed decisions. | Weak data lineage When organizations can’t trace where data comes from, how it has been transformed, and how it’s being used, it becomes difficult to validate results or troubleshoot issues. |
| Definitional misalignment Metrics that seem straightforward can be defined differently across teams. Not only can this lead to conflicting analyses, but it can also result in misaligned decisions, whether the analysis was completed by a human or a GenAI solution. | Workforce readiness Analytics tools have always required some level of literacy to use well, and generative AI data solutions are no different. Teams need to understand how to use the tool and how to evaluate what it produces. |
How to Use AI in Data Analytics
When it comes to data analytics, AI is expanding what’s possible. However, the organizations getting the most out of it are being selective about where and how they apply it.
Generative AI can add significant value when it comes to:
- Data collection
- Data cleaning
- Data analysis
- Data visualization
- Predictive analytics
- Data-based choices
Where Good Intentions Meet Bad Outcomes
Generative AI and data analytics can be the perfect pairing, but they can also amplify underlying weaknesses and introduce greater risks. Without strong foundations, organizations can quickly find themselves making decisions based on incomplete or inconsistent data, scaling flawed insights, and placing too much trust in outputs they don’t fully understand.
So, what exactly can go wrong?
Starting With Tools Instead of the Use Case
Jumping to tools before defining the use case is one of the most common mistakes. GenAI is easy to access, which makes it tempting to move fast. Without clarity on what decisions need to improve or what good looks like, teams often end up applying AI in ways that have nothing to do with business priorities.
Taking Outputs at Face Value
GenAI produces insights that are fast, well-structured, and confident in tone. If there aren’t any validation processes in place, teams might simply accept those insights at face value, even if they’re misleading, incomplete, or based on flawed data.
Operating Without Clear Governance
Then there’s governance — or the lack of it. Who owns data quality? Who approves access? Who’s accountable when an AI-driven decision goes wrong? Those questions rarely have clean answers.
At the same time, the ease of using generative AI increases the likelihood that sensitive, proprietary, or personal data will be introduced without proper safeguards, raising compliance and security concerns that many organizations aren’t fully prepared to manage.
Scaling Poor Data Quality
Data quality issues don’t automatically disappear with better tools. After all, generative AI is only as reliable as the data it draws from, whether that data is public or internal. Fragmented systems, inconsistent definitions, and incomplete metadata can all lead to outputs that appear credible but are ultimately inaccurate. Without strong governance and data management practices, those issues are brought to the surface and scaled.
Underestimating the Cost to Scale
Cost is another variable that businesses underestimate. While the barrier to entry is low, running AI at scale isn’t. Querying large datasets, maintaining infrastructure, and supporting ongoing usage add up quickly, especially if there’s no clear plan for managing them.
Falling Behind On AI Literacy
Finally, there’s the literacy gap. GenAI tools are evolving fast, which means teams are constantly trying to catch up. Even with the same inputs, outputs can vary, creating inconsistency in results and making it more difficult to rely on AI in repeatable, production-level workflows. Without the skills to interpret, question, and validate what’s being generated, teams may struggle to distinguish between useful insights and misleading ones.
Best Practices for Data and Analytics Strategy Development That Holds Up
Whether an organization is building a data strategy from scratch or revisiting one that predates the current GenAI wave, a few principles apply. Here are some best practices for data and analytics strategy development:
1. Keep Humans at the Center
Generative AI accelerates analysis and surfaces insights quickly, but human judgment remains essential. Teams must validate outputs, apply context, and critically assess results before taking action.
Human involvement ensures training data remains relevant and unbiased, providing essential oversight that enhances accuracy and trust.
2. Start With Business Outcomes, Not Tools
Before adopting new technology, organizations should define the decisions they aim to improve and establish clear success criteria.
Without that foundation, even the most capable tools risk being applied to the wrong problems—producing insights that don’t move the business forward.
3. Strengthen Data Foundations: Quality, Definitions, and Lineage
Generative AI is only as reliable as its underlying data. Organizations must prioritize clean, complete, and relevant data.
Consistent definitions and clear data lineage are also required, enabling teams to understand data origins, transformations, and interpretation. Without this, outputs may seem credible but lead to incorrect conclusions.
4. Establish Governance, Access, and Accountability
As AI accelerates data usage and scale, effective governance becomes increasingly important.
Organizations need clear policies for data access, usage, and ownership, as well as defined accountability for AI-driven decisions. This includes safeguarding sensitive data and ensuring compliance with privacy and security standards.
5. Enable Collaboration Across Teams
Effective analytics requires collaboration. Business, technical, and governance teams must work together to align data, tools, and decision-making. When collaboration fails, insights lose both relevance and impact.
6. Measure, Monitor, and Refine Over Time
Deploying generative AI is an ongoing process. Organizations should define success metrics beyond usage, focusing on accuracy, efficiency, and business impact. Continuous monitoring, evaluation, and refinement are essential to maintain system performance and adapt to evolving needs.
Get the Basics Right. Then Get to Work.
Generative AI is changing what’s possible in data analytics, and the organizations that approach it thoughtfully are the ones that will turn that potential into consistent, compounding value.
Successful AI integration isn’t about the tools. It’s about the strategy behind them. That’s where the right partner can make the difference.
At SEI, we help businesses build the data and analytics foundations so generative AI can be applied in ways that actually deliver value. From aligning strategy and governance to improving data quality and workflows, we ensure organizations are set up to scale AI responsibly. The result is not just faster insights, but stronger, more reliable decision-making.
