If you’re a data analyst, I want to ask you a question.
How are you?
This year has been filled with unrelenting change, the fundamental kind where everything is shifting, from the way we work to the technology that makes it all possible.
As difficult as this change has been, I believe there’s also unprecedented opportunity ahead, especially for those who know the business and data best, especially for data analysts.
We recently commissioned a survey to better understand how the role of the analyst is changing. What emerged from the research was clear: in the age of AI, the role of the data analyst is still indispensable. Here’s why.
AI is making analysts more effective
In 2023, right at the inflection point of gen AI adoption, half of workers in business support and logistics roles were concerned AI would replace them. Today, many analysts still share that concern.
However, our survey painted a much different picture. A more hopeful one.
Almost every analyst we spoke with (96%) said they’re using AI to streamline tasks, while others (47%) said AI is helping reduce their workloads.
- 76% said AI is helping them be more effective and efficient
- 55% said AI is increasing their productivity
- 87% said AI has increased their job satisfaction
AI has become a force multiplier. Eight out of ten analysts (83%) are now influencing mission-critical decisions. They’re steering their organizations with the help of AI. These results are both positive and promising.
However, AI hasn’t made analysts solely more effective at their roles. It’s also changed them. Enterprise AI adoption has necessitated their involvement in nearly every business process and department that wants to operationalize AI.
The reason is twofold. The first is the missing logic layer.
The missing piece for enterprise AI: The logic layer
AI models can reason incredibly well with data, but they can’t infer the rules, calculations, and institutional knowledge that define how a specific business or department operates.
If a business leader asked you what the Q1 revenue in California was, you probably wouldn’t give them an answer right away. Instead, you would start asking questions.
Would you like the revenue net of sales costs? Do you want it the way we report it to Wall Street or the way we deliver it to store managers? Should we include partner revenue?
These kinds of questions matter. If an employee asked an LLM this same question, the technology would first need clean and accurate data, but it would also need to intimately understand your business.
Business logic is the hardest thing to capture in an AI workflow and the most difficult to scale because it lives in spreadsheets and undocumented processes, inside the heads of the business experts who have been honing their craft for years.
Business analysts are among the few employees who truly understand both the processes and data behind things like revenue recognition, commission structures, and customer segmentation. For that reason, organizations are leaning on analysts to build and maintain the logic layer for enterprise AI systems.
Making AI trustworthy
Then there’s the reliability gap. Trust from AI systems has to be earned. Its outputs must be verifiable and reproducible. Decision-makers need to understand how an AI system came to an answer and trust that its answer is correct to act with confidence.
In our research, the top three barriers to succeeding with AI were:
- Difficulty explaining or interpreting AI outputs to stakeholders (55%)
- Limited analytic skills, among business users (54%)
- Poor data quality (50%)
Bear in mind that almost half (47%) of AI projects fail because of poor data quality or governance.
While analysts have always been instrumental in cleaning and validating data — the “input” for AI workflows — analysts have gained a new role as “AI overseers.”
In our research, analysts told us they were spending nearly four hours in their new AI oversight capacity, validating or correcting AI-generated outputs, with nearly a fifth of analysts (16%) spending more than six hours a week on these tasks.
Their involvement is not an accident. AI workflows need transparency and accountability at every stage. Someone will always be needed to explain what’s happening to the data and verify that AI has arrived at the right answer. Analysts around the world are stepping into this role.
63% of analysts say validating AI outputs has become a more important skill
Analysts are powering enterprise intelligence
If you’re an analyst who’s been worried and uncertain about the future, I understand. However, I hope our latest research gives you hope. An AI-powered organization is one that deeply requires analysts’ involvement.
First, for cleaning and preparing data and building the foundation for trustworthy AI. Next, for applying business logic to calculations and AI workflows. And finally, for validation. Analysts are needed at every step: beginning, middle, and end.
Businesses tomorrow will not be run by “vibe coders.” Rather, the most valuable employees will be those who know the data and the business inside and out. These are the business analysts, the rev ops professionals, the finance team members, and the supply chain experts, the ones who can ensure AI is working the way it was meant to.
Read the full research report “State of the Data Analyst: The Rise of Business Logic” to learn more about the future role of the analyst.