The Approach That Makes Your AI Initiative 3.4X More Likely To Succeed

A conversation with Jennifer Jackson, CMO of Actian, on the Building for Others podcast

Data Matters More Than You Think

There's a number from a research report that Actian published earlier this year that should stop any executive in their tracks. Companies that have implemented data products and data contracts are 3.4 times more likely to have three or more AI projects running in production at scale, actually delivering results.

Actian partnered with BARC, a Germany-based industry analyst firm, and surveyed 300 data professionals across multiple industries and geographies (the Americas, EMEA, and Asia-Pacific) to test a hypothesis: that data maturity and AI maturity are the same track. The research validated it. Companies that have done the hard, unglamorous work of treating their data as a managed asset, packaging it at a defined level of quality, reliability, and AI-readiness, are the ones actually shipping AI at scale.

Jennifer Jackson, CMO of Actian, walked us through how this thesis cuts against the grain of how most organizations have approached the AI moment. The instinct has been to reach for pace at all costs: just sign up for a model, teach everyone prompt engineering, and throw spaghetti at the wall. What the data shows is that the companies winning are the ones who had already built a foundation underneath their model/app layer. As JJ put it on the podcast, "I really believe that the practice of understanding your data, taking care of it, governing over, and really knowing your data will ultimately win the race."

For a lot of organizations, that's an uncomfortable conclusion: closing the AI gap is a data discipline problem, and that’s a larger project than just signing up for Cowork seats.

This Means Data Teams Are Leading Workforce Transformation. Did They Sign Up For That?

At elvex, we're seeing the same dynamic play out with our own customers. Data teams have  become the owners of AI transformation inside many organizations, which isn’t always something that data scientists and analysts pictured themselves doing when they started their careers. They're the ones who understand what the data actually looks like under the hood, how it will influence AI answers, and all the plumbing required to set up good agentic workflows. AI workflows are, ultimately, getting the right answer into the right place, which means data teams are the ones getting pulled into every AI initiative whether or not that was in their job description.

That's both exciting and, for some, uncomfortable. AI transformation initiatives are frequently about re-imagining how people approach their work, helping them be more productive. Generally that’s either an HR initiative or inside the teams themselves. Most data scientists and data engineers didn't sign up to lead workforce transformation or to be in charge of changing how hundreds of colleagues do their jobs every day. Yet that's increasingly the role they're being asked to play. Leadership should recognize this burden and give data teams the mandate, resources, and credit that this ask requires.

How Actian Is Building the Stack: Three Acquisitions, One Vision

Actian is the data and AI division of HCLSoftware, which has made three strategic acquisitions that have reshaped what Actian can do.

It started in September 2024 when HCLSoftware completed the acquisition of Zeenea, a Paris-based data catalog and governance company, bringing cloud-native metadata management, lineage tracking, and an adaptive knowledge graph to the Actian Data Intelligence Platform.

Then in December 2025, HCLSoftware acquired Wobby, an early-stage Belgian startup building AI Data Analyst agents for data warehouses. Wobby's core capability is a natural language interface backed by a proprietary semantic layer and agentic architecture. Actian relaunched it as Actian AI Analyst, adding a "Steward AI Agent" that automatically builds and maintains the semantic layer: the piece most natural-language analytics tools skip, which is why they tend to produce answers you can't trust.

That same month, HCLSoftware announced the acquisition of Jaspersoft from Cloud Software Group. The embeddable reporting and analytics platform was closing just as JJ sat down for this podcast. The through-line across all three: Zeenea ensures the data is discoverable and governed. Wobby ensures anyone can query it. Jaspersoft ensures it's reportable.

Democratizing Data: When Anyone Can Ask, Everything Changes

For decades, there's been a wall between business users and their own data. It was called SQL. If you couldn't write a query, you were dependent on someone who could. You submitted a ticket, waited, got a spreadsheet back, and by the time you had your answer, the question had changed. The democratization of data has been a promise the industry has been making for twenty years.

Actian AI Analyst is making good on it. The platform lets non-technical business users query enterprise data in plain English, with no SQL required. The Steward AI Agent maintains the semantic layer underneath (reminder, this is what separates a trustworthy answer from a hallucinated one).

JJ described what this actually feels like from the inside: "I can have the conversation that I need to have with the data in my words. I can pressure test that data by asking it to validate things and prove things to me, and then I walk away with an understanding that, months ago, would have taken weeks for me to get."

Drinking Your Own Champagne

She wasn't speaking hypothetically. The morning of this podcast recording, JJ had a company all-hands coming up. She needed to understand the quarter: leads generated, opportunities created, the headline story she could bring back to her team. So she queried Actian's own AI Analyst to get it. She asked her data, had a back-and-forth dialogue with it, and walked into the all-hands with the answers she needed.

She called it "drinking our own champagne": A CMO using her own product to query her own pipeline data the morning of a big meeting, with no data analyst in the loop, no ticket queue, no waiting. That's what democratization of data looks like when it works: someone getting the answer they needed and moving on.

Zero-Click World: JJ's Take On What Marketers Should Do When Google No Longer Sends You Traffic

This is the part of the conversation that probably hits closest to home for anyone running a marketing team right now. JJ has been thinking hard about what she calls the zero-click world: the shift happening as AI overviews, ChatGPT, Perplexity, and similar tools give users the answers they're looking for without ever sending them to a website. In a world where the AI gives you the answer and lists the sources but most people never click through, the entire model of digital marketing starts to wobble.

JJ was candid that when zero-click first became a serious topic twelve to eighteen months ago, panic was the initial reaction. Her answer: build brand, because brand has never mattered more. "I also believe brand has never ever been more important than it is today. Your reach as a brand is probably as important as it was thirty years ago when we didn't have all the digital."

When the AI synthesizes an answer and surfaces sources, the question becomes which brands it knows, trusts, and cites. That's a function of authority, presence, and structured visibility: the same things that made brand matter before the digital era, now operating through a different mechanism.

What Actian Is Actually Doing About It

Actian is already working the tactical side of this:

• Measuring GEO results in Profound: They use this tool to manage and track their visibility in AI-generated results (what JJ called the GEO or AEO world: generative engine optimization and answer engine optimization).
• Tracking referrals from ChatGPT and Perplexity: Monitoring where traffic is actually coming from in the AI era, not just traditional search.
• Writing for machines as well as humans: "We're writing for machines as well as humans now." That means thinking less about keywords and more about how people actually ask questions: natural language, full prompts, the way someone would phrase something to an AI rather than type it into a search bar.
• Two versions of the website: JJ has been turning over the idea of one version optimized for machine consumption and one for humans. The content that serves a human reader and the content that serves an AI crawler are genuinely different things, and the gap between them is only going to widen.

Listen to the Full Episode

JJ covers a lot more ground than we could fit here, including her take on the cultural side of AI adoption, the tension between moving fast and doing the data work right, and how to approach running a modern marketing team.

Listen to the full episode if you're a data leader, a CMO, or anyone trying to figure out what enterprise AI actually requires beneath the surface, this one is worth your time.