Gen AI Study: Practical Insights For Financial Desktop Product Teams

Gen AI features are a part of every conversation around financial desktop platforms. Everyone is talking about smarter search, automated summaries, and AI-assisted analysis.

Inside organisations, the story is a bit mixed. Product teams are not sure what really moves the needle. Procurement teams are looking for better proof of value. People are experimenting, using a few tools, and dropping others that don’t fit into their actual work patterns either.

This gap between capability and usefulness was the starting point for a behavioral market research study, focused on understanding how AI and GenAI features are really being used within financial desktop platforms.

Behavioral Market Research Objectives And Scope

The starting point was practical. AI had become part of platform positioning, but decision-makers still lacked clarity on how much value it was creating once the novelty wore off.

So instead of asking “Does the platform have GenAI?”, we looked at a more uncomfortable but useful set of questions:

• Are users even noticing these tools?
• Are they using them more than once?
• Do they trust the output enough to rely on it?

Because in this space, the real battle is not feature launch. It’s adoption. If a tool gets tried once and then ignored, it does not matter how good the demo looked.

The scope included users who rely on financial desktop platforms for serious work, analysts, bankers, traders, and product roles. We also included procurement and vendor management professionals, because they sit at a different point in the chain. Users talk about usefulness. Procurement talks about value and renewal decisions. Both matter, and they do not always agree.

Benchmark Framework And Evaluation Areas

We avoided the classic trap of benchmarking platforms like a product brochure.

Instead of listing features and scoring them, we evaluated how capabilities showed up inside workflows. That difference matters. A feature can exist and still be irrelevant if it does not appear at the right moment, or if it forces users to change habits they’ve built over the years.

The benchmarking areas included:

• Discovery and search improvements
• Summarisation and synthesis
• AI-supported analysis and signal spotting
• Workflow acceleration and reduction of manual effort

And we kept asking one underlying question: Does this make the job easier, or does it just add another step?

Many AI tools do not fail because they are bad. They fail because they interrupt the flow.

Key Insights From GenAI Benchmarking

Adoption Was Real, But Not Even

Most respondents knew AI features existed and had tried them. But usage levels varied a lot depending on role.

People who spend large parts of their day scanning huge amounts of information were more likely to adopt summarisation and discovery tools. In other roles, GenAI sat more in the “nice to have” category, used occasionally but not embedded into routine.

Awareness Did Not Guarantee Habitual Use

This one came up repeatedly.

Many tools got attention early. People clicked around, tested them, and stopped.

The biggest reason was friction. If a tool saves time quickly, users come back. If it requires extra checking or produces outputs that feel slightly off, it becomes something people avoid because they cannot afford to double-check everything all day.

Efficiency Gains Were Narrow But Meaningful

The behavioral market research did not show that GenAI magically speeds up everything. That would be a suspicious conclusion anyway.

Where it genuinely helped was in tasks like:

• Quickly scanning and filtering information
• Summarising long updates and documents
• Giving a starting point before deeper work

Trust Shaped Long-Term Behaviour

Trust is what turns AI from “cool feature” into “daily tool”.

When outputs were consistent, explainable, and felt anchored to the platform’s data, users grew more comfortable. When outputs felt unpredictable or overly confident, users disengaged quickly.

And you could sense the underlying fear: If this gets it wrong, it’s my name on the decision, not the system’s.

Procurement Expectations Were Moving Faster Than Usage

From a procurement perspective, AI is increasingly becoming an expectation. Not having it can look like being behind.

But procurement respondents were also cautious about paying more for “AI” without clear use cases. AI presence alone did not equal value. What mattered was:

• Adoption inside teams
• Visible efficiency gains
• Clarity on where AI actually supports workflows

In short, the conversation is shifting from “Do you have AI?” to “Is your AI used enough to justify its cost?”

Business Impact And Decision Outcomes

For the client, the main value of the study was not a big headline. It was clarity. The kind that prevents circular arguments.

Rather than discuss AI as an abstract concept, the client got a more nuanced idea of:

• What capabilities were actually reused
• Where adoption was reduced after early experimentation
• How usage differed by role type
• Which tools provided the most consistent efficiency benefits

This helped inform our product and roadmap decisions. Some capabilities clearly needed refinement. Some needed better placement and integration. And a few needed more honest positioning, because if a tool is rarely used, it should not be treated like a core differentiator.

Most importantly, decisions could be grounded in observed behaviour instead of assumptions and internal bias. That alone is a big win in product conversations.

Bottom Line: Why This Case Study Matters

GenAI is now part of the financial desktop landscape. That part is not changing. What will matter more over time is what actually sticks.

This behavioral market research reinforced one simple idea: usefulness beats novelty. If a tool saves time in a way that fits real workflows, people use it. If it creates doubt, friction, or extra work, it disappears quietly.