As we have just concluded the first month of 2026, it’s a good opportunity to consider last year, what people are talking about for the new year, and the state of the insights industry. It’s been great talking with clients about what they are seeing and hearing and experiencing, and their plans for 2026, and I had the good fortune to connect recently with agency leaders at the CEO Summit in south Florida organized by the Insights Association. Here are some key themes…
There are clearly applications of AI in insights that are driving productivity improvements today. However, there are also many failures. MIT published a study in the back half of 2025 that reported 95% of corporate GenAI pilots are failing. That’s high but also somewhat natural for any new tech adoption, and Gartner refers to this period as the “trough of disillusionment.” But, what’s important is the expectation for AI successes is much higher than reality today.
The elevated expectation for AI impact is entrenched in the financial markets and adopted by corporations. The C-suites are then pushing this down in their organization, and it’s putting insights groups and leaders under tremendous pressure. They are pushed with AI “can’t you just…” as if everything is automatically and immediately easier. Insights leaders are still expected to deliver high quality, and so the pressure is intense.
At the CEO Summit, an insights leader at one of the very largest corporations told the story of their CEO recently declaring that everything they do should take half the time. That’s a lofty challenge. Insights accepted the challenge and analyzed all their processes. It inevitably led to a conversation with their internal business partners about what is “good enough” for results, and how “certainty” can no longer be provided. Insights leaders at other companies are having similar discussions as to whether 70-80% accuracy is “good enough”.
Data quality is the elephant, and this remains a systemic challenge for insights. Ironically AI is a big part of the problem as well as part of the solution here. Many would argue this is the number one issue, even over AI. The task of analyzing collected data and applying tools of all types has never been more involved. 20% is the absolute floor of data that is getting thrown out, and replaced, today. Experts agree that human review is essential which potentially makes things less efficient and take longer, but is needed to ensure data quality.
Productivity improvements are expected, and often today’s AI tools are not enough to deliver them, at least in early 2026. It doesn’t matter. The expectations will drive reality. Therefore, what is required is a review of every insights process and function to drive out wasted time and money. AI is the catalyst for this review and change. Agency leaders are all going down this path. There has never been a better time to ask questions of client and data partners as to what is needed and how to mutually drive efficiency.
What can get lost is the human factor. It’s not just a belief in human led AI, or that advances with AI favor the most experienced and talented people. It’s that we simply have a need for humanness. Past technology advances produced a need for humanness in other areas. It’s been argued that quant research moving to the internet 25 years ago produced more interest in in-person qualitative. Marketers all crave a need to connect with their actual customers and rebel in some way against having them seem more distant.
All of this had led to a concern about the potential dilution of quality in insights. The pressure to drive dramatic improvements through AI, as well as reduced headcount and smaller spending budgets, has created stress on client insights leaders. Their leading role in helping their organizations take confident action has not changed. This is the challenge for agency leaders and their call to action. Listen and adapt, and continue to drive high quality insights. We are excited for the challenge.
