I still remember the moment she blurted “I love the dashboards but…” In that split second, I knew I had to ignore everything before the “but” and brace myself for the truth coming after that! I put on a brave face but the truth hit me hard, as it so often did.
“I love the dashboards, Andy, but can’t we plug in something like a ChatGPT on top of them, so I don’t have to interpret the data? I could simply chat and get my answers.”
It wasn’t meant as an insult to my years in data engineering and dashboarding. Yet, seeing the CMO of a fashion apparel brand not keen on visual exploration bothered me deeply.
Cut to a different scene, where I found myself facing a young manufacturing CEO. Lauren* had recently taken the reins from her father and was remarkably fluent in the language of IT. It was a pleasant surprise, considering most manufacturers I knew were brilliant domain experts but easily overwhelmed by tech. As Lauren scrutinized the custom-made enterprise data platform, I quickly disclaimed that a chat-like AI assistant is already on our roadmap.
She nodded, but was already digging into the dashboards. She drilled up and down like a seasoned data analyst, slicing through different charts and even setting up a custom view that mapped the operator productivity across supervisors.
That was my moment of reckoning. I couldn’t help but contrast the two leaders. The marketing leader was a storyteller at heart, for whom flashy bar charts or hover tooltips were just cognitive overload. She needed a natural language AI assistant. On the other hand, the manufacturing CEO was a woman of few words. The dashboards perfectly served her analytical mind.
Pondering over these two starkly different personalities, I can’t help but wonder how data engineering services firms spend weeks crafting modern data platform architectures – encompassing robust data pipelines, governed data lakehouses, and real-time dashboard insights. Yet enterprise data platforms often fail not because of engineering constraints or tech resistance. It’s because the design overlooks the user persona.
Let me illustrate a third persona we encountered during a Tableau to Power BI migration for Luminotrix* – a laser tool manufacturer. The company had used an on-premise Tableau version since its inception but when they grew, the per-user Tableau costs began to rise. An underutilized M365 license sparked the Power BI migration, projecting a healthy savings of more than 30%. However, certain users were accustomed to Tableau’s inherent visual sophistication. To bridge the gap, Trigent replicated 14 Tableau features in Power BI including floating layouts, customizable tooltips and the live-query experience.
During the post migration user feedback, the CIO caught us completely off-guard. “You got the data platform layer right. But I think some of our users don’t have to use Power BI at all.” On the surface, his observation pointed to Power BI’s native integration with Excel, where users can directly analyze data while using the governed data model from Power BI.
But the CIO was actually referring to the fact that the busy plant managers had zero time to probe numbers or talk to AI assistants. They might instead need autonomous agents that monitor anomalies, alert when exceptions arise, and recommend the best course of action. Don’t be surprised if this persona is not logging into your dashboards. They don’t want to explore data – they just watch out for alerts and precisely follow the agent instructions.
The below table sums up the priorities of these user personas
Narrative Seekers | Data Lovers | Action takers | |
Expectations on data platform | Precise contextual answers | Metrics and visualizations that helps discover insights | Agents to instruct the next course of action |
Mindset | Let’s chat and figure this out | Let me find my way out through drill-ins and drill-outs | Just tell me what to do |
Relationship with data | No, I can’t interpret it | I love numbers | I don’t want to see data |
Preferred interaction | Natural language | Interactive dashboards | Recommendations and alerts |
Success metric | Fast answers. | Rich visualizations | Better decisions |
You may have heard experts and novices alike proclaim that dashboards are dead, but those conclusions were often drawn with only one user persona in mind: the marketing leader who would rather converse with an AI assistant than explore charts. But what about the young manufacturing CEO who loves granular insight? Lauren* believes that a single centralized dashboard ensures that no critical issue slips through the cracks. She even confessed that she opens his dashboard every morning before reading her emails. Then there is the third persona: the plant manager who might not be glued to your dashboard, but relies on agentic alerts to navigate daily operational decisions.
The success of enterprise data platforms invariably rests in the hands of these distinct user personas and once their unique needs are catered to, there are no ifs and buts. Your dashboards are truly loved.
FAQs
1 Are dashboards becoming obsolete in enterprise data platforms?
No. Dashboards are evolving rather than disappearing. While some business users prefer interacting with data through conversational AI assistants, others still rely on interactive dashboards to explore trends, drill into metrics, and uncover insights. Increasingly, enterprise data platforms must support multiple modes of interaction—including dashboards, natural language interfaces, and autonomous agents—to serve different user personas.
2 How do user personas influence the success of an enterprise data platform?
The success of an enterprise data platform depends as much on user adoption as on technical architecture. Some users seek instant answers through natural language, others prefer visual exploration with dashboards, while action-oriented users value intelligent alerts and recommendations over data analysis. Designing the platform around these distinct personas significantly improves adoption, decision-making, and business outcomes.
3 How should organizations balance dashboards, AI assistants, and autonomous agents?
Organizations should avoid treating these capabilities as competing alternatives. Dashboards remain ideal for analytical users, AI assistants simplify data access for conversational users, and autonomous agents support operational users who need timely recommendations rather than visualizations. A modern enterprise data platform should seamlessly combine all three experiences so every user can interact with data in the way that best suits their role and decision-making style.
