AI Made Us Smarter. Why Is It Still So Hard to Make Decisions?

Organizations today operate in a constant stream of data: dashboards refresh, alerts trigger, AI summarizes in real-time. Individually, each insight might be useful. But together, they create a new problem: too many competing versions of the truth, with no clear way to validate or prioritize them. 

You’ve likely seen this happen in your own organization. Different teams show up with different data, pulled from different tools, interpreted through different lenses. Meetings that should drive direction turn into negotiations over sources. Strategy discussions drift toward reconciling perspectives instead of committing to action. 

In situations like these, acting quickly on fragmented intelligence feels risky; slowing down feels like the responsible choice. Then, by the time alignment is reached, the decision window has quietly closed.   

Speed dies in the gap between seeing and believing

In our CEO fireside chat, A-INSIGHTS co-founder Jeroen Lustig reflected on the discussions he’s been having in boardrooms over the past 16 months, where he’s been asking executives the same three questions: 

Are you making more decisions than a decade ago? More. 

With more or less certainty? Less. 

Faster or slower? Faster. 
 
For years, the default response to volatility has been to manage through it. The assumption wasn’t that uncertainty would completely disappear, but normalize. In the meantime, companies followed a (relatively) standard playbook: protect margins, preserve optionality, avoid irreversible moves until visibility improves. 

But this assumption has quietly collapsed under the realization that uncertainty is not a phase but the default operating environment for the foreseeable future. And this makes the old cadence — waiting for complete information before carefully building consensus — impossible to sustain. 

That’s the tension Stuart Reynish, Eetu Laaksonen, and Martijn Lustig work inside every day. Together, they lead Valona’s product, AI, data and tech. They’ve watched the same pattern show up across global enterprises: more data, faster analysis, and still a widening gap between “we see it” and “we move on it.” 

What follows are the patterns they’ve seen in organizations that broke this cycle: the three shifts in how intelligence needs to evolve to meet the moment.

1. If you need to ask a question, you’re already behind

We’re used to thinking about intelligence as something you go get: ask a question, run a query, pull a report. But this “intelligence-on-demand” model is built for a world where there’s time to notice something, formulate a question, and wait for the answer. That world doesn’t exist anymore. 

Martijn uses a recent example: the EU-India trade agreement. When the deal was announced, leadership teams faced immediate questions: Does this create opportunity? What’s the risk? 

India is an emerging producer. Products could now enter European markets at lower tariffs. Every manufacturing executive was seeing the same headlines. But the nuance of what it means varies dramatically depending on the business. 

If you produce in Europe and compete on price, it’s a threat. If you source from India, it’s an opportunity. If you operate in markets India can’t easily reach, it’s likely irrelevant. The signal is universal. The implications, specific.  

And “specific” calls for specific data: market size, trade volumes, tariff deltas, competitor capabilities, production efficiency comparisons. 

“As a human, it’s quite difficult,” Martijn explains. “You need time, you need access to all the data. Especially, you cannot bring it together in the timeframe that leaders need.” 

Without a system already in place, this kicks off weeks of back and forth; pulling information from different sources, reconciling contradictions, running multiple cycles of analysis and validation. By then, competitors will have already captured position.  

The fastest-moving competitors aren’t reacting to the headline at all. They were already monitoring India’s production capacity trends, European tariff exposure by product category, competitor expansion signals. When the deal was announced, they weren’t asking “what does this mean?” — they were confirming scenarios they’d already modeled and decided which moves to accelerate.  

Eetu puts it more bluntly: “You need global data and global data points. A market and competitive intelligence platform that can bring the global, the regional, do it in any language possible, to get to the concrete of what is happening. AI is the enabler to analyze this fast, get information to decision makers, and connect it to quantitative data points.” 

That’s the shift. The right system doesn’t wait for you to ask. It runs continuously in the background, monitoring the signals that could change your position, so that when something moves, the context is already assembled. Giving you that much more time to move on an open decision window.   

2. Intelligence only scales when reality is shared

Faster individual analysis created an unexpected problem: organizational fragmentation.  

Martijn describes what this looks like: “There’s a news article about a merger. Person A shows up with one set of underlying factors. Person B shows up with different ones. The discussion becomes about the data and the trustworthiness of the data, instead of the insight and the decision leaders need to make.” 

This is what happens when intelligence is generated ad hoc, one question at a time, by different people, using different sources, assumptions, and logic. The output may be fast, but the underlying reality isn’t shared. 

But speed doesn’t come from better individual insights; it comes when entire organizations reason from the same underlying reality; same data sources, same logic, same interpretation of what a signal means for the business’ specific exposure. 

When intelligence is anchored in that shared reality, strategy meetings shift. More time goes into interpreting what the signal means and deciding what to do next, and less time goes into reconciling whose version of reality to trust. 

3. Beyond the hype: Agentic AI as the ultimate dot connector

Faster analysis and shared data sources solve part of the problem. But they still require humans to connect the dots between isolated events. 

Say a merger rumor surfaces. Another trade deal gets announced. A key competitor hires a new CFO. Each might trigger event-driven analysis. But strategic intelligence isn’t so much about individual events as the patterns that connect them. 

This is where “agentic AI”—everyone’s new favorite buzzword—has a big role to play. 

Eetu is careful about the term. “Agentic doesn’t mean autonomous AI or handing decisions over to machines. It means workflows that run end-to-end—triggered by events, applying consistent reasoning, stopping at defined boundaries.” 

Back to the India-EU trade deal example. Its effect on tariffs is just the tip of the iceberg: leaders will also want to think through competitor expansion plans in those markets, their own supply chain dependencies, customer concentration risk, and production capacity shifts. Agentic workflows do so much more than simply analyzing the deal. They automatically examine what it means across all those connected dimensions — using the same organizational context, the same data sources, the same analytical logic. 

This is what closes the gap between “we see it” and “we move on it.” Not faster individual analysis. Connected intelligence that compresses the time between signal and decision—positioning organizations to move while windows are still open. 

Three questions worth asking

Before your next strategic planning cycle, it’s worth asking:

1. Decision lag: 
How many major signals did your organization see early but act on it too late? Not because the intelligence was wrong, but because alignment took way too long? 

2. Fragmented reality 
When was the last time two senior leaders walked into the same meeting with different data, different interpretations, or different levels of visibility into the same competitive shift? 

3. Reactive vs. ready:  
If the next India-EU trade agreement happened tomorrow, would your team already be monitoring the data layers needed to assess impact? Or would you be left scrambling to gather market size, trade flows, and competitor positioning from scratch? 

AI has lowered the cost of information, but it didn’t lower the cost of coordination. In a world of abundant insights, advantage belongs to those who align faster than they analyze.  

Stuart Reynish, Eetu Laaksonen, and Martijn Lustig lead product, AI, and technology at Valona Intelligence. Their work: competitive and market intelligence systems that close the gap between signal and action so leaders can take full advantage of decision windows.

Facing similar challenges? Let’s talk.