Research Briefs: Nearly Thirty Years in the Making
In 1997, I joined Intraware, where we were experimenting with a different approach to technology research and advisory services. A shoutout to Peter Jackson, who gave my boss, James Brentano, the sandbox to build what James envisioned as a better Gartner, with smarter people and a different approach to answering questions. And to James, for hiring me and letting me be part of that experiment.
The personal computer revolution of the 1980s and early 1990s had already changed how organizations worked, and the internet was beginning to change how they connected, operated, and conducted business. In 1997, 53.5% of Americans used computers, but only 22.2% used the internet. Enterprise adoption was accelerating, although the business implications were still emerging.
Our idea was simple: organizations kept asking similar questions, so why not develop thoughtful, reusable answers they could access when needed? The questions were largely technology-centric. What is distributed computing? How does a particular architecture work? Where does an emerging technology fit? They required research and technical understanding but usually had reasonably definable answers.
Nearly thirty years later, AI is creating another major transition in enterprise technology. The parallels with the early internet are imperfect but useful. Both introduced capabilities that organizations initially evaluated as technologies before understanding their broader business implications. Both challenged established assumptions about architecture, processes, economics, and competitive advantage.
The gap between adoption and results is equally familiar. McKinsey's November 2025 research found that 62% of surveyed organizations were experimenting with AI agents, while only 39% reported any enterprise-level EBIT impact from AI. Its August 2026 research found that 40% of respondents from enterprises with at least $1 billion in revenue reported scaling AI agents, compared with 27% the previous year. Adoption and experimentation are advancing, while the organizational and economic implications remain unresolved.
Technology has become inseparable from business strategy, operating models, economics, talent, governance, and competitive positioning. Executives need to understand the decisions technology creates, the available alternatives, and the consequences of their choices. That became the foundation for The Board.
Its monthly intelligence reports identify consequential developments, connect them to enterprise realities, and maintain continuity as the market evolves. They help executives understand what is changing, which conversations deserve attention, and, equally important for my work, which questions remain unresolved.
My first attempt to address those questions was a series of explainers. They assembled useful research and information but remained too technology-centric. There were too many data points, too much explanation of the technology, and insufficient interpretation of why the information mattered or how executives could use it. I was organizing research around subjects when the work needed to begin with the executive questions those subjects raised.
That experience led to Research Briefs.
A Research Brief begins with a consequential executive question and develops the evidence, reasoning, and perspective necessary to answer it.
- How should enterprises rationalize AI spending?
- How should organizations determine where AI can contribute to solving increasingly complex problems?
- How much decision authority should enterprises give AI agents?
These questions involve competing objectives, organizational constraints, uncertain outcomes, and decisions whose consequences differ across enterprises. They require substantial research, a willingness to challenge established assumptions, and a clear connection between the evidence and the decisions being considered.
We begin with published research, documented enterprise experience, academic work, and market developments. We identify recurring questions across independent sources and examine the prevailing interpretations. What does conventional analysis overlook? Which assumptions remain untested? What contradicts the accepted position? Where might a technically sound recommendation fail when applied to the economics, capabilities, or operating realities of an enterprise?
We develop an argument, challenge it, return to the evidence, examine competing interpretations, and refine the conclusions. Sometimes conventional wisdom survives scrutiny; sometimes a different explanation emerges, or the investigation reveals that the original question was inadequate. The process continues until the analysis provides a defensible position, identifies its limitations, and establishes what the findings mean for executive decisions.
Each brief stands independently. There is no prescribed sequence because enterprises encounter different questions at different times, and overlapping evidence may support different decisions.
Generative AI has also made this model practical in ways that were unavailable even a few years ago. Without it, producing this kind of research would take me months rather than days. Finding relevant material, exploring competing interpretations, identifying contradictions, and revisiting earlier work consumed considerable time before I could complete the substantive analysis.
My experience is that the greatest advantage comes from making it easier to find, organize, compare, and interrogate research. I can investigate more evidence, test competing explanations, and refine conclusions within a practical publishing cycle. The quality of the result still depends on identifying the right questions, understanding the limitations of the evidence, and exercising judgment about the conclusions. Knowing how to work with the technology, including when to challenge or disregard its output, is essential to realizing those benefits.
This changes the economics of research. Organizations repeatedly spend time and money investigating substantially similar questions, gathering evidence, consulting specialists, evaluating alternatives, and developing conclusions. Some of that work is enterprise-specific, but much of the underlying investigation can be shared.
Research Briefs apply that principle to executive decision intelligence. The research, competing interpretations, evidence assessment, and conclusions can serve multiple organizations, while their application depends on each enterprise's priorities, constraints, capabilities, and alternatives. The substantial cost of developing the intellectual foundation can be distributed across many users, much as software development separates reusable capabilities from individual implementation requirements.
The Board's taxonomy and research methodology provide the organizing foundation. Monthly intelligence identifies developments and emerging questions, while Research Briefs investigate those questions in depth. Together, they create a growing body of knowledge that can support executive deliberation and more focused advisory engagements when an organization's circumstances require additional investigation.
The differentiation comes from the research process and the experience informing it.Consulting firms, analysts, universities, vendors, and increasingly AI systems can assemble evidence, identify trends, and generate recommendations. Determining which questions deserve attention, recognizing incomplete interpretations, and understanding what happens when recommendations encounter enterprise realities requires a combination of research discipline and practical judgment.
My career has provided several perspectives on those problems. I have worked as an analyst, with technology vendors, with executives making and funding decisions, and with organizations trying to implement change. I have seen the differences between what technology promises, what vendors can deliver, what executives expect, and what organizations can accomplish. That experience helps me recognize assumptions worth challenging, connect findings across disciplines, and examine consequences that may be missing from conventional technology analysis.
Experience does not replace evidence or guarantee correct conclusions. It informs how the evidence is examined and interpreted. Combining that perspective with independent research, adversarial analysis, and a disciplined process of refinement is what makes this approach distinctive.
The first Research Briefs have already been used in executive work, and the experience suggests that the model is substantially closer to what I have been trying to build. I will continue refining it, with an initial commitment to developing at least one new Research Brief each month and additional briefs when the emergence and preponderance of consequential questions in published research and documented case studies justify them.
In 1997, we were helping organizations understand a technological transition whose business implications were still emerging. In 2026, we face another such transition, with substantially better tools for investigating the questions and a greater need to connect technological capabilities with executive decisions.
That was an idea worth pursuing in 1997. I think it is an even better one today.
Reading Resources
- AI, Human Cognition and Knowledge Collapse — Daron Acemoglu, Dingwen Kong, and Asuman Ozdaglar (2026). Examines how AI-generated recommendations may improve current decisions while weakening the incentives for human learning and the development of collective knowledge.
- Designing Human-AI Collaboration: A Sufficient-Statistic Approach — Nikhil Agarwal, Alex Moehring, and Alexander Wolitzky (2025; revised 2026). Explores the allocation of decisions between people and AI, including the effects of overconfidence and reduced human effort.
- Growth in AI Knowledge — Joshua Gans (2025). Investigates the economics of AI-assisted knowledge creation and the circumstances in which easier access to existing knowledge may encourage or discourage new research.
- Genius on Demand: The Value of Transformative Artificial Intelligence — Ajay Agrawal, Joshua Gans, and Avi Goldfarb (2025). A theoretical examination of how AI could change the allocation and economic value of scarce human expertise.
- Canaries in the Gold Mine: Early Productivity Gains from Artificial Intelligence Creating Organization Capital — Tania Babina, Alex X. He, and Renhao Jiang (2026). Examines how investment in AI-related capabilities may create durable organizational knowledge and contribute to productivity.
- Shifting Work Patterns with Generative AI — Eleanor Dillon and colleagues (2025). A field experiment showing that individual time savings from AI do not necessarily produce broader changes in work organization.
- The Productivity J-Curve: How Intangibles Complement General Purpose Technologies— Erik Brynjolfsson, Daniel Rock, and Chad Syverson (2021). Explains why general-purpose technologies require complementary organizational and knowledge investments before their full economic value becomes visible.
- AI as Strategist — Joshua Gans (2025). Examines the relationship between expertise, organizational coordination, credibility, and authority when AI contributes to strategic decisions.