Why AI Is Forcing a Rethink of Analytics, Value, and Platform Strategy

Hi there,

A lot is changing in analytics right now, but not all change matters equally.

This week’s edition focuses on three shifts worth paying attention to: why AI is expanding analytics beyond structured data, how self-service can turn reporting into expansion revenue, and what to watch for when evaluating AI analytics platforms in a market that is moving faster than most roadmaps can keep up with.

Video published: March 17, 2026

A lot of enterprise analytics still rests on one assumption: the most important business data lives in structured systems. NVIDIA’s keynote challenged that.

At GTC 2026, Jensen Huang pointed to a divide that matters more than most teams admit. On one side, there is the structured data ecosystem, already massive and well established, serving as the ground truth for enterprises and AI systems. On the other, there is a much larger and faster-growing universe of unstructured data: PDFs, voice, video, images, and all the information companies create every day without a clean way to query it.

That is where AI starts to change the equation.

Structured data is still useful. It is still necessary. But working only with structured data will become less valuable over time, because more of the business lives outside rows, columns, and dashboards than most systems were built to handle.

And user expectations are about to move with it.

People will not want answers from one dashboard, one warehouse, or one reporting layer. They will expect answers from the whole knowledge base, with structured and unstructured data working together in one response.

That is where traditional BI starts to look limited.

Dashboards can show what was modeled. Reports can summarize what was already captured. But AI can start to make sense of what was previously hard to search, hard to connect, and hard to use. That is a much bigger shift than a better analytics interface.

This is why analytics is becoming AI Analytics.

The real change is not just faster insight. It is a new expectation that systems should help users understand the full context of the business, not just the part that was easy to structure.

Traditional BI will not disappear overnight. But over time, tools built mainly for structured reporting will look increasingly incomplete next to systems that can reason across everything the business knows.

Key Takeaways

For SaaS companies, expansion revenue is often the most efficient path to growth. While new customers matter, deepening value within existing accounts is faster and less costly. Dresner Advisory Services underscores this dynamic, identifying direct revenue generation from embedded analytics as a top external business outcome for technology companies.

The opportunity exists because analytics has clear value tiers.

Nearly every SaaS product offers some version of basic reporting (standard dashboards, fixed KPIs, static views, etc.) and customers expect it to be included. But these “best practice” views describe an average business, not their business.

The real monetization leverage comes from advanced capabilities—especially self-service—that let customers define metrics, dashboards, and reports around how they actually operate. That’s where analytics shifts from generic reporting to something customers are willing to pay more for.

Embedded analytics creates expansion revenue when it moves beyond averages. In practice, that means:

This shift changes the value conversation. Analytics is no longer a generic feature—it becomes a way for customers to embed their competitive advantage into the product. When customers can model their own business inside your product, willingness to pay increases naturally, and premium tiers justify themselves.

Key takeaways

A lot of AI analytics platforms look compelling in demos. The harder question is whether they still make sense once the AI landscape shifts, the roadmap changes, and your team has to build on top of them.

That is the real challenge right now.

Some vendors are selling future potential more than current reality. Features may look polished but still be experimental. Roadmaps are often vague. And because AI itself is changing so fast, it is hard to know which models, interfaces, or workflows will actually matter a year from now.

That is why teams need to separate the platform from the features.

Features will change fast. Chat interfaces, copilots, forecasting, and automation layers will keep evolving. The platform underneath matters more. Its architecture, extensibility, integration model, and deployment flexibility will determine whether it can adapt as AI changes.

The safest approach is to evaluate the core first.

Can it support multi-tenancy, permissions, and real product complexity? Can it integrate cleanly with your stack? Can it support future AI capabilities without forcing rework?

Then look at the AI layer.

Not just whether it looks impressive today, but whether it is modular, flexible, and aligned with your product direction.

Key Takeaways

Analytics is no longer just a reporting layer. It is becoming a product, revenue, and infrastructure decision all at once. As AI changes what users expect and what platforms can do, the real advantage will go to teams that build for where the category is heading, not where it has been.

See you in the next edition,

Arman Eshraghi
The Embedded Intelligence Brief
Connect on LinkedIn