# More Data, Less Understanding 2026-09-09 · Somerset County, New Jersey · Reported Feature Zappi's latest industry study finds companies connecting more consumer data while growing less satisfied with insights — a paradox about judgment that AI speed cannot solve by itself. Businesses have spent years building the machinery to know more about people. Transactions can be connected to loyalty programs, browsing behavior can be paired with campaign performance, survey results can be… --- A new study from consumer-insights platform Zappi suggests the reality is more complicated. In its third annual Connected Insights Imperative, the company reports that 48 percent of surveyed organizations now describe their consumer insights and data as connected, up from 38 percent in 2025. At the same time, overall satisfaction with the insights function fell from 60 percent to 48 percent. The result creates an uncomfortable business paradox: organizations are connecting more information, using more technology and apparently becoming less satisfied with what all of that machinery is helping them understand. The research, released September 2 and based on a survey of more than 250 marketing and insights professionals, should be read as industry research rather than neutral public polling. Zappi sells a consumer-insights platform and has an obvious commercial interest in the argument that businesses need more mature systems for connecting research and decision-making. Even so, the findings are useful because they illuminate a tension that extends well beyond one vendor's framework. The ability to gather, store and process information has expanded dramatically, while the harder organizational task — deciding what matters, preserving learning over time and applying it at the moment a decision is made — has not automatically kept pace. Zappi's numbers do not support the simplistic conclusion that connected data is failing. In fact, its own maturity framework shows the opposite. Satisfaction rises from just 9 percent among organizations it classifies as disconnected to 32 percent among fragmented organizations, 62 percent among connected organizations and 75 percent among the small group Zappi calls AI-accelerated. The decline in overall satisfaction therefore appears alongside evidence that more mature systems are associated with substantially higher satisfaction. The gap is not between data and no data. It is between possessing connected information and building an organization that can repeatedly convert that information into useful judgment. Artificial intelligence makes that distinction especially visible. Zappi reports that 93 percent of surveyed organizations are using AI in some form, yet only 8 percent meet its definition of AI-accelerated. Most of the reported uses remain procedural: 39 percent cited data analysis and interpretation, 36 percent cited productivity and efficiency, and 33 percent cited generating new ideas and concepts. Those are meaningful applications, but they mostly make existing work faster. They do not necessarily change how knowledge flows through an organization or whether a decision-maker receives the right piece of consumer evidence at the moment it can still alter the decision. That difference is easy to lose in the current AI conversation because speed is so visible. A model that can summarize a hundred research documents in seconds creates an obvious demonstration of capability. The harder question is whether the summary changes what the company does. If the underlying studies were inconsistent, if the context around the data was poorly documented, if teams disagree about which metrics matter, or if previous research remains buried inside separate functions, artificial intelligence can make the retrieval process faster without resolving the organizational confusion underneath it. A faster route through fragmented knowledge is still a route through fragmented knowledge. Zappi's own explanation of connected insights points directly at this problem. The company argues that consumer knowledge becomes more valuable when individual research projects stop living as isolated answers and instead accumulate into a body of evidence that can inform future advertising, innovation and brand decisions. In that model, the important unit is not the dashboard, the report or even the individual study. It is institutional memory. One campaign can teach a company something about a message, a character, a product benefit or a particular audience, but the advantage appears only if the next team can find that learning, understand its context and know when it is relevant again. This is where the modern obsession with data volume can become misleading. A company can measure more things without becoming better at asking questions. It can connect systems without connecting interpretations. It can centralize information without creating agreement about what the information means. The technical achievement of putting signals in the same place is not the same as the intellectual achievement of explaining human behavior, because consumers do not experience themselves as rows in a unified database. They carry motives, contradictions, habits, identities, financial constraints and changing contexts that often produce behavior no single metric can explain. One of the more revealing findings in the Zappi report concerns the balance between behavioral information and motivational insight. Nearly half of connected organizations, 47 percent, report balancing behavioral or performance data with consumer insight and motivational data, compared with just 14 percent of disconnected organizations. That distinction is fundamental. Behavioral data can tell a company what people clicked, purchased, abandoned or repeated. It is considerably less reliable at answering why. A business that mistakes behavioral visibility for psychological understanding can become extremely precise about what happened while remaining surprisingly uncertain about what caused it. The research also captures a growing uneasiness about letting AI stand in for people. Zappi reports that only 7 percent of organizations are currently using synthetic respondents within their insights functions. It also found a significant internal divide over whether synthetic data should be used for major decisions: 47 percent of insights professionals were unwilling to do so, compared with 18 percent of marketers. That disagreement is important because it shows the AI debate moving from whether companies will use the technology to where they are willing to trust it. There is a meaningful difference between using AI to locate patterns in existing evidence and using AI-generated people as a substitute for evidence from actual consumers. Zappi itself has warned about that distinction in its guidance on synthetic research, noting that fluent answers from large language models can create false confidence when the model, training data or validation approach is not appropriate for the decision at hand. The warning is unusually relevant to the broader problem raised by its new report. Modern tools are exceptionally good at producing outputs that look complete. Organizations still have to determine whether the output reflects reality, whether the underlying evidence is strong enough, and whether the result should change a business decision. The sophistication of the interface does not remove the burden of judgment. There is also a human organizational problem embedded in the satisfaction decline. Zappi reports that insights teams are increasingly seen as strategic partners or proactive advisors, which means expectations for those teams are rising at the same time that more tools and more data become available. Once a company connects its information, the existence of the connection stops feeling impressive. It becomes infrastructure. The question immediately shifts from whether the organization can access consumer knowledge to whether that knowledge is producing better creative work, stronger products, more confident decisions and fewer repeated mistakes. Success raises the standard by which the system is judged. That may be the most important interpretation of the falling satisfaction figure. The decline does not necessarily indicate that organizations know less than they did a year ago. It may indicate that they now expect much more from knowing. The technical promise of connected data and artificial intelligence has reset the baseline. If nearly every organization can deploy AI and almost half can describe their consumer information as connected, simply producing research or generating a summary no longer feels like an achievement. The expectation is increasingly that the system should remember, interpret, anticipate and make the next decision better. For business leaders, that is a much more demanding standard than installing another platform. It requires consistent definitions, disciplined research practices, shared access, clear ownership of learning, documentation of context and enough organizational trust for evidence to cross departmental boundaries. It also requires preserving the role of actual human interpretation, particularly when consumer behavior is ambiguous or contradictory. Technology can make patterns easier to detect, but deciding which pattern is meaningful remains a business and human judgment. The irony is that companies may finally be confronting the limits of the information strategy they spent years building. More data was supposed to reduce uncertainty. In many organizations, it has instead revealed how much uncertainty remains after the dashboards are connected. Artificial intelligence can compress the pile, search the pile and increasingly converse with the pile, but none of those capabilities guarantees that the organization has learned something true about the person on the other side of the transaction. Businesses do not ultimately need the maximum amount of consumer data. They need enough reliable evidence, connected with enough context, to make a better decision than they would have made without it. Zappi's latest research suggests that the industry is beginning to recognize the difference. The systems know more. The databases talk to each other. AI is nearly everywhere. The remaining challenge is the one technology has never made simple: understanding people well enough to know what to do next. Key findings Connected data48% of surveyed organizations described their consumer insights and data as connected, up from 38% in 2025Overall satisfactionSatisfaction with the insights function fell from 60% in 2025 to 48% in 2026AI use93% reported using AI in some form, while 8% were classified by Zappi as AI-acceleratedMaturity and satisfactionSatisfaction ranged from 9% among disconnected organizations to 75% among AI-accelerated organizationsHuman and behavioral evidence47% of connected organizations reported balancing behavioral/performance data with motivational consumer insight, versus 14% of disconnected organizationsSynthetic respondents7% reported currently using synthetic respondents; 47% of insights professionals said they were unwilling to use synthetic data for major decisions, compared with 18% of marketers SOURCE NOTES • Zappi / PR Newswire — Sept. 2, 2026 release, More Organizations Connect Consumer Data, but Satisfaction with Insights Falls • Zappi — Connected Insights Imperative 2026 overview • Zappi — Connected insights: What they are and how to use them • Zappi — How connected data powers better consumer insights • Zappi — Perspective on human, AI and synthetic research --- ProbleMattic is written and maintained by Matthew Kulcsar, a software engineer, project manager, technologist, platform builder, emergency-services-trained helper, grandfather, and lifelong collector of broken systems, odd behaviors, and useful nonsense.