Most innovation programs fail before a prototype is built. They fail in the silence between strategy and execution, in the moment when a team decides what problem to solve.
Research published in the Harvard Business Review, surveying over 100 top level executives across 17 countries, found that 85 percent agreed their organizations were fundamentally bad at diagnosing problems, with 87 percent acknowledging that this weakness carried serious financial costs[1].
The root cause is cognitive. The human brain is wired to leap from an observed situation directly to a proposed solution, a shortcut called solution bias. In business-to-business (B2B) strategic innovation, where sales cycles are long, client relationships are complex, and the cost of a misaligned product reaching late-stage development can run into tens of millions, this shortcut is catastrophic.
Because of this bias, B2B product development often treats innovation as a purely creative exercise. When companies focus on the creative output rather than the commercial outcome, they often fall into the trap of developing technically brilliant solutions that nobody needs. Organizations frequently suffer from the “Problem Definition Trap,” gravitating toward technically interesting projects that completely lack commercial relevance. Developing a product that solves an irrelevant problem consumes R&D budgets, ties up engineering talent, and delays the launch of solutions that could actually drive revenue.
This is why identifying and defining problems worth solving is one of the key activities of front-end innovation. The foundational premise is straightforward: the “Explore” phase must chronologically and philosophically precede any attempt to ideate, and it must be funded, staffed, and given the time required to produce genuine insight, rather than just internally-generated assumptions.
The essence of innovation is commercialization. Therefore, the competitive advantage in modern business-to-business innovation belongs to the organization with the empirical patience and methodological rigor to discover the right problem to solve before committing significant capital to solve it.
The essence of innovation is commercialization. Therefore, the competitive advantage in modern B2B innovation belongs to the organization with the empirical patience and methodological rigor to discover the right problem to solve before committing significant capital to solve it. Today, the most powerful, underutilized tool for doing this at speed and depth is a well-prompted large language model performing structured trend analysis.
The Explore–Ideate–Shape Framework and Where AI Trend Analysis Fits
Bringing systemic order to early-stage innovation requires a structured framework. The Explore-Ideate-Shape engine comprises three sequential, interdependent phases:
- Explore: Identify genuine problems worth solving, contextualize them based on predetermined constraints, and synthesize them into actionable innovation themes anchored in both market evidence and the strategic reality of the organization.
- Ideate: Generate a high volume of diverse concepts grounded strictly in the opportunity areas uncovered during “Explore.” Teams that ideate without validated problems recycle internal assumptions and produce concepts that are creative but directionless.
- Shape: Stress test promising concepts using the “killer questions” methodology against stringent commercial, technical, and strategic criteria before significant capitalized development begins.
While the sequence is non-negotiable, an iterative approach that allows teams to loop back and forth enables efficient opportunity shaping. Bypassing the “Explore” stage corrupts the “Ideate” stage entirely. The framework designates three to six months per exploration campaign, because meaningful insight requires iterative hypothesis testing, not a one-day workshop. Traditionally, a trend analysis of the scale shown in the example below takes two to three months of effort.
“Explore” encompasses several complementary methods: customer and expert interviews, jobs-to-be-done research, technology and competitive landscape mapping, value chain analysis, market intelligence, and trend analysis. This article focuses on trend analysis specifically: a structured methodology for identifying macro-level directional shifts that are creating or intensifying high-priority problems across industries.
Two Dimensions of Problem Identification: Within the “Explore” phase, problem identification operates on two levels:
- Identifying problems worth solving (Market-Level): What significant, underserved problems are emerging from structural shifts in technology, regulation, economics, or behavior? This is the output of a rigorous trend analysis report.
- Contextualizing problems worth solving (Company-Level): Of the validated market problems, which can our organization build, defend, and commercially sustain a solution to? This requires filtering the market problem through predetermined constraints such as type of business, strategic fit criteria, strict maximum R&D budgets, required ROI timelines, or explicit boundaries of where the company will not play.
The AI-driven trend analysis approach described here directly addresses the first dimension, rapidly mapping market level problems while leaving the second dimension to the strategic contextual judgment of the human innovation team.
The Anatomy of Structured Trend Analysis
Structured trend analysis is a systematic, evidence-driven inquiry organized around a defined set of analytical components, each of which builds on the last. When executed rigorously, it produces not a list of trends, but a map of the problems those trends are creating, prioritized, quantified, and grounded in the structural forces that make them urgent. The four essential components of structured trend analysis are not optional modules; they are a logical sequence, and skipping any one of them produces an output that is directionally interesting but strategically incomplete.
Understanding the Trend: Topic and Drivers. The foundation of any trend analysis is a precise definition of the trend topic itself, not a broad domain label, but a focused, directional shift in technology, behavior, economics, or regulation that is reshaping one or more industries over a defined time horizon. A well-scoped trend topic is specific enough to yield a concentrated set of four to eight high-priority problems, broad enough to affect multiple actors and carry strategic relevance, and anchored to a primary driver rather than pulled simultaneously by many unrelated forces. Electrification, for instance, is a domain, not a trend topic. Electrification of Commercial Vehicle Fleets in North America and Europe over a five-to-ten-year horizon is a trend topic — specific, directional, and analytically tractable.
The drivers of a trend are the underlying forces that originate, accelerate, or sustain it. Structured trend analysis uses a STEEP framework (Societal, Technological, Economic, Environmental, and Political/Regulatory) to systematically identify and assess each driver against three attributes: its velocity (how fast it is changing), its certainty (how predictable it is), and its impact radius (how broadly it affects industries and actors). This discipline prevents the common analytical failure of treating a trend as a single monolithic force when it is, in reality, a convergence of multiple independently moving drivers, each with its own timing, trajectory, and implication for the problems it creates.
Problem Landscape. Understanding a trend creates the analytical precondition for identifying the problems it generates. The Problem Landscape component translates trend intelligence into a structured map of friction points and unmet needs, not a list of pain points, but a rigorous inventory of problems that meet a defined threshold: they are real, they are underserved, and they are large enough to justify investment in solving them. Each problem must be defined across six components:
- The specific pain point experienced
- The affected actor or stakeholder segment
- The root cause that makes the problem structural rather than incidental
- The magnitude in terms of the number of actors affected and at what severity
- The urgency driven by the velocity of the trend
- The solvability, in terms of technical, commercial, and regulatory tractability at this point in time.
The six definitional components are embedded within eight analytical sub-sections in the deep-dive format. Structured problem identification also requires prioritization. Not all problems are equally worthy of investment, and a rigorous trend analysis must produce an explicit ranking with defensible scoring criteria, not a flat enumeration of everything that is difficult.
Opportunity Analysis. Once the Problem Landscape is defined and prioritized, the third component quantifies the value at stake and maps the competitive terrain. Opportunity analysis answers three questions that problem identification alone cannot:
- How large is the economic prize for solving this problem?
- Who is currently attempting to solve it, and where are the genuine white spaces?
- What would a first mover need to do to build a defensible position before the window closes?
This component requires named competitive players, real investment and M&A signals, and recent proof points, not generic statements about market potential. Its output is not a market-sizing exercise, but a strategic landscape: an account of where value is being created, where it is being left uncaptured, and what organizational posture is required to capture it.
Strategic Implications. The final component translates analysis into foresight. No trend follows a single trajectory, and no problem landscape is static. Strategic implications analysis identifies the inflection points at which the trend will accelerate, decelerate, or bifurcate, and the leading indicators that signal which path is being taken. It develops multiple scenario narratives, not as abstract planning exercises but as operational frameworks: if the slow case materializes, what should an investor, operator, or technology provider do differently than in the fast case? The monitoring dashboard that accompanies this component gives practitioners a structured set of signals to watch, transforming a static report into a living analytical instrument. Without this component, a trend analysis report is a snapshot; with it, the report becomes a guide for sustained strategic navigation.
These four components — Understanding the Trend, Problem Landscape, Opportunity Analysis, and Strategic Implications — constitute the logical architecture of structured trend analysis. Practiced rigorously, they produce the kind of output that front-end innovation requires: not a catalogue of what is changing in the world, but a prioritized, evidence backed answer to the question of which problems that change is making worth solving. The challenge, historically, has been the cost and time required to execute all four components at the depth they demand. That constraint is precisely what the AI-assisted methodology described in the next section addresses.
How to Generate a Trend Analysis Report Using AI
The methodology described here is open-access. The AI prompt provided in the appendix can be used by any team, in any organization, with any major large language model (such as Claude, Gemini, or ChatGPT).
What the Prompt Produces: The prompt instructs an AI model to generate a complete, structured trend analysis report. To ensure high-impact usability, the prompt dictates the following structural inclusions:
- Part I (Understanding the Trend): Definition and scope, historical milestones, convergence of forces driving urgency, stakeholder map, and maturity assessment.
- Part II (Problem Landscape): A rigorous problem identification framework followed by deep dives into each identified problem, each analyzed across eight dimensions: problem statement, root cause mapping, causal driver linkage, affected actors, current vs. desired state, magnitude and market size, urgency and inflection points, and barriers to solving.
- Part III (Opportunity Analysis): Economic value at stake, societal and environmental impact, first-mover advantage dynamics, competitive landscape with named players, investment and M&A signals, white spaces, and recent proof points.
- Part IV (Strategic Implications): Inflection point analysis, a leading-indicator monitoring dashboard, and three fully-developed scenario narratives (slow, base, and fast case) each with explicit strategic implications.
Every section has mandatory content requirements. A self-check rubric of over twenty checkpoints (included with the sample prompts below) ensures the output meets standards for completeness, specificity, definition compliance, and actionability.
A Worked Example: Electrification of Commercial Vehicle Fleets
To validate the approach, this methodology was applied to the trend topic Electrification of Commercial Vehicle Fleets using a Claude Sonnet 4.6 (Thinking) model. The resulting report (available for download below) identified five high-priority problems worth solving:
- Depot charging infrastructure — the highest-ranked problem, representing a $47 billion market opportunity by 2030, where fleet operators take delivery of electric vehicles they cannot charge because utility grid interconnections and transformer installations are delayed by 18 to 36 months, a structural misalignment no incumbent is solving end-to-end.
- Fleet energy management and grid integration — a $9.1 billion software market where unmanaged depot charging triggers peak demand charges that instantly eliminate the fuel savings of electrification; the operator who owns this platform owns the operational brain of the modern fleet.
- Public and corridor megawatt charging (MCS) reliability — a critical infrastructure gap estimated at $20 billion or more, locking Class 8 long-haul trucks out of electrification until 1-megawatt-or-greater highway charging networks become commercially viable at scale.
- High-voltage technician shortage — a structural deficit of 50,000–80,000 certified technicians across North America, creating compounding downtime costs as fleets scale and OEM service networks remain constrained by proprietary diagnostics.
- TCO financing and insurance models – a $15 billion financing market where residual value uncertainty on battery assets prevents lenders from underwriting commercial EV procurement, leaving small-to-medium fleet operators without access to viable financing terms despite available IRA subsidies of up to $40,000 per vehicle.
Each problem was analyzed with market sizing, root cause mapping, urgency timelines, named competitive players, and scenario-based strategic implications. The full report, produced in under 30 minutes of active generation time, represents work that would typically require a team of analysts several weeks.
How to Use It
Step 1 (Copy the prompt): The full AI prompt is provided in Appendix A. It is ready to use once the five customization fields at the bottom are completed.
Step 2 (Fill in five customization fields at the bottom of the prompt before sending):
- The specific trend topic (well-scoped: specific enough for 4–8 problems, broad enough for strategic relevance)
- Optional problem hypotheses to stress-test
- Your primary audience (e.g., innovation team, investors, operators, regulators)
- Geographic focus
- Current date.
Step 3 (Run with an extended-thinking AI model): The prompt works with any capable large language model. Extended-thinking or reasoning-capable models produce significantly deeper root cause analysis and more credible market estimates.
Customization is encouraged. The prompt’s content requirements, scoring criteria, and section structure are all adjustable. Teams may add industry-specific frameworks, company-specific strategic filters, or domain glossaries directly into the prompt before running it.
What’s the Impact of This Approach?
The shift from traditional to AI-assisted trend analysis is structural, not incremental:
Speed: A rigorous trend analysis report that traditionally requires two to three months of analyst research and synthesis can be produced in a matter of hours. This does not eliminate the need for human judgment; however, it significantly reduces the time cost of information assembly, leaving the team’s cognitive capacity available for interpretation and contextualization.
Depth and consistency: The structured prompt enforces a level of analytical rigor that is difficult to maintain consistently across a manual research process. Every problem is analyzed across the same eight dimensions. Every driver is assessed on three attributes. Every section meets a defined quality threshold before the AI finalizes the output. This consistency makes outputs from different trend topics directly comparable enabling portfolio-level prioritization across multiple innovation themes simultaneously.
Accessibility: Senior strategists and innovation leaders can commission and interpret trend analysis reports without requiring a dedicated research team. Small organizations, academic institutions, and innovation consultancies that cannot staff full analyst functions gain access to a quality of insight previously available only to well-resourced enterprises.
Customization without complexity: The prompt is modular. An organization can add its own strategic context such as type of business, capability constraints, revenue thresholds, geographic focus, existing IP as additional filtering criteria, and the AI will incorporate these into the problem contextualization layer. The framework is a starting point, not a ceiling.
A critical boundary to maintain: AI generated trend analysis is an accelerant for the Explore phase. Customer interviews, expert conversations, and field observation surface tacit knowledge that no language model can access. The highest-value use of this methodology is as a first-pass foundation: a rigorous, well-structured baseline that a human team can interrogate, challenge, enrich with primary evidence, and ultimately validate or refute through direct market engagement.
Conclusion and Next Steps
The “Explore” phase of front-end innovation is where the highest leverage and most consistently underfunded work in B2B innovation happens. Trend analysis, the systematic identification of macro level directional shifts and the problems they create, is one of the most powerful tools within Explore. The methodology described here makes that tool available to any organization willing to invest the time to apply it well.
AI [doesn’t replace] the analytical judgment of an experienced innovation strategist…it dramatically lowers the cost of assembling the raw material on which that judgment operates.
What has been demonstrated here is a repeatable, open-source approach to dimension one of problem identification: generating a credible, evidence backed map of problems worth solving in any given trend domain. This is a genuine contribution to the practice of front-end innovation, we believe, not because AI replaces the analytical judgment of an experienced innovation strategist, but because it dramatically lowers the cost of assembling the raw material on which that judgment operates.
Next steps:
- Dimension two contextualization prompt: The logical next step is to develop a second AI prompt that takes the output of the trend analysis report and filters it through a specific organization’s strategic context: its capabilities, market position, competitive landscape, revenue thresholds, and time horizon constraints. This would address dimension two (contextualizing problems for a specific company) with the same structured rigor applied to dimension one in this work.
- Validation methodology: A structured process for validating AI-identified problems through primary research (customer interviews, expert panels, or Jobs-to-be-Done analysis) would close the loop between AI-generated hypotheses and field-grounded evidence. This could be packaged as a companion prompt that generates an interview guide and hypothesis validation framework calibrated to the specific problems identified in the trend report.
- Multi-topic portfolio synthesis: Running the prompt across five to ten trend topics simultaneously, then applying a cross-topic prioritization framework, would enable an organization to construct a diversified innovation portfolio with explicit risk, timing, and capability alignment across themes, a significant advance over the current practice of evaluating innovation opportunities one at a time.
- Integration with the “Ideate” phase: The problem statements produced in Part II of the trend analysis report are, by design, structured to serve as the direct inputs to an ideation session. Developing a structured AI prompt that converts the “Desired Future State” and “Barriers” sections of each problem into ideation challenges and creative constraints would create a seamless Explore-to-Ideate handoff.
The B2B organizations that master problem definition today, at speed, with rigor, and with the discipline to contextualize market insight against strategic reality, will build the solution portfolios that define their industries tomorrow. The tools to do this are available now. The only requirement is the methodological commitment to use them.
Appendix A: The AI Prompt
The complete AI prompt for generating a trend analysis report is provided as a companion document: AI Instructions, available in Word or .txt form below.
The prompt includes:
- Role and objective instructions for the AI
- Canonical definitions of Trend Topic, Problem (six components), and Trend Drivers (STEEP framework)
- Section-by-section mandatory content requirements for all eight report sections, Executive Summary, and five Appendices
- Over twenty-checkpoint quality self-check rubric
- Professional document formatting standards
- A five-field topic customization guide.
The prompt is designed to work with any major large language model. Extended-thinking or reasoning-capable models are recommended for maximum depth.
Appendix B: Example Output
A full example of the trend analysis report generated using this prompt — on the topic Electrification of Commercial Vehicle Fleets — is provided as a companion document: Report_Commercial_Vehicle_Electrification.html, also available below. (Download and open in your web browser.)
The example report demonstrates the full output of the methodology, including five deeply analyzed problems, ten formatted exhibits, a competitive landscape with named players, three scenario narratives, and five substantive appendices.
[1] Wedell-Wedellsborg, T. (2017). “Are You Solving the Right Problems?” Harvard Business Review, January–February, pp. 76–83
Ranjan Dash is a Strategic Advisor at Early Charm Ventures. He is an Innovation Executive specializing in the commercialization of high-impact opportunities and external innovation. Ranjan received his Ph.D. and MBA from Drexel University, and he has over 20 years of industrial experience spanning corporate innovation, manufacturing, and entrepreneurial ventures.
Suresh Chandran is a Clinical Professor of Management at Drexel University’s LeBow College of Business. His interests are in the areas of innovation and strategy. Suresh received his Ph.D. from Vanderbilt University, and he has over 25 years of experience providing consulting advice to national and international companies.
Featured image by Javier Allegue Barros on Unsplash.
Document Download
Click filename(s) to download and view.
You must be logged in to post a comment.