Generative AI for Market Research and Business IdeaValidation: How Leaders Can Test Markets, Competitorsand Product-Market Fit Faster
- May 29
- 16 min read
Executive Takeaway
Generative AI does not validate a market, prove product-market fit or determine whether customers will pay. It can, however, help leaders test assumptions faster and more rigorously: mapping markets, identifying competitors, preparing customer discovery, challenging business models and defining the next cheapest experiment. Used with verified sources and real customer evidence, AI can help organizations learn before they spend.
Generative AI for Market Research and Business Idea Validation
How Leaders Can Test Markets, Competitors and Product-Market Fit Faster
The Most Expensive Innovation Mistake Is Building Around Untested Assumptions
Many new ventures, products and strategic initiatives do not fail because a team lacks the ability to build. They fail because the organization invests in a solution before it has tested whether the underlying assumptions are true.
The assumed market may be smaller or harder to reach than expected. The customer problem may be inconvenient rather than urgent. The apparent innovation may already be addressed by direct competitors, substitutes or internal customer workarounds. The buyer may differ from the user. The economics may depend on acquisition costs or pricing that prove unrealistic. An MVP may test features while missing the riskiest commercial assumption entirely.
These mistakes are costly because they consume more than capital. They absorb leadership attention, product capacity, investor confidence and time that could have been directed toward a stronger opportunity.
The most expensive mistake in innovation is not building too slowly. It is building confidently around untested assumptions.
Generative AI creates a meaningful opportunity to reduce this risk—not by replacing market evidence, but by accelerating the early work required to identify what must be proven, what may be false and what decision should come next.
Why Market Validation Needs to Become Faster—Without Becoming Less Rigorous
Sound innovation has always required market research and competitive analysis. The U.S. Small Business Administration describes market research as a way to find customers and reduce risk, while competitive analysis helps define an advantage in the market. It also distinguishes between existing-source research, which can support general and quantifiable questions, and direct research with consumers, which is necessary to understand specific audience reactions and needs.
That distinction remains essential in the AI era.
Traditional early-stage validation can be slow and fragmented:
Market sources are researched manually and summarized inconsistently.
Competitor analysis focuses on familiar companies while missing substitutes.
Customer interviews are conducted but findings are not synthesized systematically.
Business-model assumptions remain implicit rather than documented and challenged.
Product teams rush into building before defining what the MVP is meant to prove.
Executive decisions are based on enthusiasm, fragmented evidence or delayed reports.
At the same time, competitive cycles are compressing. New technologies reshape customer expectations quickly. AI-enabled organizations can explore alternatives, prepare experiments and iterate at a pace that challenges slower decision processes.
Leaders therefore need a faster way to move from idea to informed hypothesis—but they must not confuse speed with certainty.
Generative AI can compress research and synthesis. It cannot remove the need for verified sources, customer conversations, commercial testing or executive judgment.
What Generative AI Changes in the Validation Process
Early market validation is fundamentally a learning process. A leadership team begins with assumptions and seeks evidence strong enough to support a decision: proceed, pivot, narrow, pause or stop.
Generative AI can strengthen this process by expanding the range of questions considered and reducing the effort required for first-pass analysis.
AI Can Support Work Such As
Scanning an initial market landscape.
Organizing relevant market drivers and trends.
Identifying direct, indirect and substitute competitors.
Structuring customer-segment hypotheses.
Preparing customer-interview questions.
Synthesizing interview notes into themes for human review.
Challenging a proposed business model.
Comparing pricing or go-to-market assumptions.
Identifying the riskiest unproven assumptions.
Designing MVP experiments and success measures.
Drafting decision memos that distinguish evidence from inference.
The opportunity is significant because AI can help a team examine more dimensions of an idea earlier. It can increase the breadth of the first pass, improve the discipline of analysis and help leaders focus human research where uncertainty matters most.
The right principle is:
Use AI to accelerate the journey from idea to evidence—not to replace the evidence.
Use AI as a Critical Analyst, Not a Cheerleader
One of the greatest risks in business idea validation is confirmation bias: teams become attached to an idea and interpret research in ways that support the direction they already want to take.
Generative AI can amplify this risk if leaders ask it only to strengthen a favored argument, prepare a persuasive pitch or estimate a large market without demanding evidence and challenge.
A more valuable use is to instruct AI to challenge the idea from several perspectives:
Critical AI Role | What It Should Challenge |
Skeptical Market Analyst | Whether the market is attractive, reachable and sufficiently differentiated. |
Competitor Intelligence Researcher | Whether alternatives already solve the customer problem. |
Customer Research Planner | Whether the pain is urgent, frequent and connected to a buying decision. |
VC-Style Reviewer | Whether the model is commercially viable, scalable and defensible. |
Product Strategy Critic | Whether the MVP tests the riskiest assumption or merely builds preferred features. |
Go-to-Market Reviewer | Whether the buyer, channel, pricing and acquisition assumptions are realistic. |
The best use of AI in early innovation is not to make an idea sound compelling. It is to identify what could cause the idea to fail before the market delivers that lesson at greater cost.
A Five-Part AI-Assisted Validation Framework
Leaders can structure AI-supported business validation around five questions. Each question connects to a decision and produces an evidence agenda—not a final answer generated by AI alone.
1. Market Landscape: Is There a Meaningful and Reachable Opportunity?
An attractive idea must address a real market opportunity. Generative AI can accelerate the initial assessment of the market category, relevant trends, possible demand drivers, customer segments and uncertainties requiring verification.
AI-Supported Questions
What market category does this idea actually compete within?
What customer segments may experience the problem most acutely?
Which economic, regulatory or technology trends affect demand?
What adjacent markets or use cases may be relevant?
Is the category emerging, mature, fragmented or consolidated?
What evidence would support or contradict the market hypothesis?
Potential Outputs
Market landscape summary.
Initial customer-segment map.
Market-driver and risk analysis.
Preliminary TAM/SAM/SOM hypothesis.
List of external sources and data requiring verification.
Uncertainty register for leadership review.
Executive Caution
AI-generated market size estimates should never be accepted as validated numbers simply because they are presented clearly. Market sizing requires transparent assumptions, verified data sources and a clear explanation of what is included or excluded.
Leadership decision supported: Is this opportunity significant enough to justify deeper evidence gathering?
2. Competitor Intelligence: Who Already Solves the Problem?
Teams frequently define competition too narrowly. They consider firms with a similar product, while overlooking substitutes, manual workarounds, existing vendors, internal processes or large platforms that customers may prefer instead.
Generative AI can help expand the competitive lens and organize findings across categories.
Competitor Categories to Explore
Competitor Type | Executive Question |
Direct competitors | Who offers a closely comparable solution to the same segment? |
Indirect competitors | Who solves part of the problem through a different approach? |
Substitute solutions | What do customers use instead of purchasing a dedicated solution? |
Legacy or manual alternatives | Is the true competitor an existing process, spreadsheet, consultant or internal workaround? |
Platform threats | Could a major existing provider introduce this capability as a feature? |
Emerging entrants | Which new firms or technologies could shift the landscape? |
Potential Outputs
Competitor and substitute map.
Feature or service comparison.
Publicly observable pricing comparison, where available and verified.
Positioning analysis.
Differentiation gaps and white-space hypotheses.
Likely competitive-response questions.
Executive Insight
The most dangerous competitors are often not the firms that look most like the proposed business. They are the alternatives customers already trust, already pay for or already use at no additional cost.
Leadership decision supported: Is there a defensible reason customers would change their behavior?
3. Customer Pain Validation: Is the Problem Urgent Enough to Change Behavior?
A problem can be real without being commercially compelling. Customers may acknowledge inconvenience yet remain unwilling to change tools, allocate budget, adopt a new process or take on implementation risk.
AI can help teams prepare for customer discovery by structuring hypotheses and designing sharper questions. It cannot replace actual customer engagement.
AI-Supported Questions
Who experiences the problem, and who controls the budget?
How frequently does the problem occur?
What is the operational, financial or emotional cost of the problem?
What existing workaround does the customer use?
What would cause the customer to switch?
What adoption barriers, risks or procurement requirements may arise?
What evidence would disprove the assumption that this pain is urgent?
Potential Outputs
Target customer and buyer hypotheses.
Jobs-to-be-done summary.
Pain severity and frequency hypotheses.
Customer-interview discussion guide.
Assumption-to-evidence map.
Risk ranking for customer discovery.
Executive Caution
AI-generated personas and pain-point summaries are hypotheses—not customer evidence. Real validation requires direct conversations, observed behavior, pilots, commitments or other credible signals that customers will act.
Leadership decision supported: Is the pain significant enough for a real buyer to adopt and pay for a solution?
4. Business Model Stress Test: Can the Opportunity Become a Viable Business?
Even when a customer problem is meaningful, the proposed business may still fail commercially. The product may be expensive to deliver, difficult to sell, hard to retain or easy for competitors to replicate.
Generative AI can act as a structured critic to help leaders expose assumptions that require testing.
AI-Supported Questions
What revenue model best fits the buyer and the problem?
What pricing logic is assumed, and how might customers challenge it?
Which acquisition channels are likely, and what might they cost?
How complex is the sales cycle or implementation burden?
What affects margin, retention or expansion potential?
What makes the offering defensible rather than merely useful?
What would a skeptical investor, board member or buyer challenge first?
Potential Outputs
Business-model critique.
Alternative monetization scenarios.
Pricing hypothesis comparison.
Customer-acquisition and sales-cycle risk map.
Commercial assumptions register.
Investor or executive challenge questions.
Executive Insight
A painful problem is not automatically a good business. The organization must be able to serve the need profitably, reach the buyer efficiently and sustain differentiation over time.
Leadership decision supported: Does the model justify an MVP investment, or does the commercial design need to change first?
5. MVP and Product-Market Fit Roadmap: What Must Be Tested First?
A common innovation mistake is building a simplified product while failing to test the assumption that matters most. A strong MVP is not merely a reduced version of the planned product. It is an efficient mechanism for learning whether the opportunity is worth pursuing.
Generative AI can help organize MVP options, define experiments and identify the link between a feature, a hypothesis and a measurable learning outcome.
AI-Supported Questions
Which assumption creates the greatest risk if it proves false?
What is the smallest test that could generate meaningful evidence?
What must customers see or experience before their response is credible?
What success measures distinguish interest from adoption or willingness to pay?
What should be intentionally excluded from the initial MVP?
What would trigger a decision to proceed, pivot, pause or stop?
Potential Outputs
Riskiest-assumption map.
MVP scope options.
Feature-prioritization matrix.
Pilot or experiment design.
Success and failure thresholds.
Learning agenda and decision gates.
Executive Message
A good MVP is not built to show how much the team can create. It is built to learn whether the most important business assumption is true.
Leadership decision supported: What is the next cheapest credible test before major investment?
Turn AI Research into Executive Decisions, Not More Information
AI can produce large amounts of analysis quickly. That creates a new risk: teams may generate reports, comparisons and ideas without translating them into decisions.
A disciplined validation process should end each research stage with a decision-ready artifact that separates evidence from assumptions.
A Practical Decision Memo Structure
Decision Memo Section | What It Should Contain |
Opportunity Statement | The proposed product, service or venture and the target customer. |
What We Know | Verified evidence supported by identified sources or direct customer input. |
What We Assume | Hypotheses that remain unproven. |
Contradicting Evidence | Signals that weaken or challenge the opportunity. |
Key Risks | Market, customer, commercial, execution and regulatory uncertainties. |
Decision Options | Proceed, narrow, pivot, pause, stop or investigate further. |
Next Cheapest Test | The lowest-cost credible action to reduce the most important uncertainty. |
Decision Owner and Criteria | Who decides and what evidence will guide the decision. |
Decisions AI-Assisted Validation Can Support
Proceed with deeper customer discovery.
Narrow the customer segment.
Reposition the offering around a more urgent problem.
Redesign the MVP to test a different assumption.
Revisit pricing or go-to-market assumptions.
Explore a partnership rather than building independently.
Pause investment until critical evidence is available.
Stop a weak idea before it consumes additional resources.
The value of AI-assisted validation is not the volume of generated research. It is the quality and speed of the next decision.
Build a Source-Grounded Evidence Base for High-Stakes Validation
General AI tools can help structure questions, generate hypotheses and conduct preliminary scans. However, where leaders are making material investment decisions, preparing investor or board materials, or evaluating sensitive markets, validation should be grounded in verifiable evidence.
A stronger workflow combines AI-supported exploration with curated sources and human review:
Use Generative AI to structure the problem, identify questions and conduct an initial scan.
Gather relevant market reports, public competitor information, customer evidence, regulatory materials and internal research.
Organize important sources in a controlled workspace or source-grounded AI environment appropriate to the information sensitivity.
Use AI to compare documents, synthesize findings and draft structured outputs from that source collection.
Verify key statements, assumptions and numbers before using them in decisions or external materials.
Convert the analysis into a decision memo and a real-world validation plan.
Practical source-grounded workspaces, including tools such as NotebookLM for suitable use cases, can help teams analyze selected sources and inspect supporting references. More tailored retrieval-based systems may be appropriate where organizations require stronger integration, governance, access management or scale.
Why This Matters
Market claims can be verified rather than simply accepted.
Contradictory evidence becomes more visible.
Research can be reused across investor, product and strategy decisions.
Teams reduce the risk of relying on confident but unsupported AI-generated conclusions.
Leadership can distinguish sourced evidence from interpretation and inference.
NIST identifies “confabulation”—often referred to as hallucination—as a risk where Generative AI produces confidently presented but false or erroneous content. In innovation decisions, that risk reinforces a simple principle: AI can support the research process, but material claims require source verification and human judgment.
What AI Can and Cannot Validate
A credible AI-enabled innovation strategy must be explicit about limits. Generative AI is powerful in preparation, synthesis and challenge. It is not a substitute for market behavior.
AI Can Help Leaders Assess | AI Cannot Finally Prove |
Whether a market hypothesis appears plausible and what data is needed. | Whether a sufficient number of customers will actually buy. |
Which competitors and substitutes may be relevant. | Whether a customer will switch from an existing alternative. |
What customer questions should be asked and what pain assumptions matter. | Whether the pain is urgent enough to support willingness to pay. |
Which business-model assumptions require stress testing. | Whether real acquisition costs, retention or margins will perform as assumed. |
What MVP scope could test the riskiest assumption. | Whether the actual product experience will create adoption and loyalty. |
What evidence is missing and what decision options exist. | Whether execution, timing and team capability will be sufficient in practice. |
AI accelerates validation. Customers, markets and real-world results still make the decision.
Common Mistakes in AI-Assisted Business Validation
Leaders can misuse AI in validation when they treat fluency as proof or speed as certainty.
Common Mistake | Why It Is Dangerous | Better Practice |
Asking AI to justify a preferred idea | Reinforces confirmation bias. | Assign AI a skeptical critic role and request contrary evidence. |
Accepting market-size estimates without verification | Produces false confidence in opportunity size. | Require source citation, assumptions and sensitivity ranges. |
Mapping only direct competitors | Misses substitutes and existing customer behavior. | Analyze direct, indirect, substitute, legacy and platform alternatives. |
Treating AI-generated personas as evidence | Confuses hypothetical customers with actual buyers. | Use personas to prepare interviews, then validate with real customers. |
Skipping customer discovery because AI is faster | Removes the strongest evidence of urgency and willingness to pay. | Use AI to improve discovery design and synthesis, not avoid it. |
Building a large MVP too early | Consumes capital before key assumptions are tested. | Design the smallest credible experiment around the riskiest assumption. |
Producing reports without stage-gate decisions | Generates information without improving allocation choices. | End validation cycles with a proceed, pivot, pause or stop decision. |
The leadership discipline is to keep asking: What is supported by evidence? What is still an assumption? What is the next test that most efficiently reduces uncertainty?
Five Practical Prompt Structures for Executive Validation
The examples below are not a complete validation system. They are starting structures that should be used with trusted source review, human judgment and real market testing.
1. Market Landscape and Evidence Prompt
Act as a skeptical senior market research analyst. I am evaluating [business idea] for [target customer segment]. Identify the likely market category, demand drivers, customer segments, relevant trends, market risks and major uncertainties. Separate: (a) claims requiring external verification, (b) reasoned hypotheses, and (c) open questions. Recommend the most credible source types to verify each material claim.
2. Competitor and Substitute Mapping Prompt
Act as a competitor-intelligence researcher. For [business idea] serving [customer segment], identify categories of direct competitors, indirect competitors, substitute solutions, legacy workarounds and potential platform threats. Propose a comparison structure covering positioning, target customer, business model, publicly observable pricing evidence, strengths, limitations and differentiation gaps. Flag findings that require source confirmation.
3. Customer Pain Discovery Prompt
Act as a skeptical customer researcher. Evaluate the assumed problem: [pain point] experienced by [customer segment]. Identify what would make this pain urgent, frequent, costly and purchase-worthy. Develop interview questions that test the assumption without leading the customer. Specify what evidence would confirm, weaken or disprove the opportunity.
4. Business Model Stress-Test Prompt
Act as a critical investment-committee reviewer. Challenge the business model for [idea]. Examine revenue model, pricing assumptions, buyer and user distinction, customer acquisition, sales cycle, delivery cost, retention, defensibility, regulatory exposure and scalability. Separate major risks from assumptions and recommend the next tests required before investment.
5. MVP Learning Roadmap Prompt
Act as a product and business validation strategist. For [idea], identify the riskiest unproven assumptions and design an MVP or experiment roadmap that tests those assumptions in the lowest-cost credible sequence. Include target users, test activities, success criteria, failure criteria, evidence required and the decisions each result should trigger.
Practical Use Guidance
Prompts are most useful when they require the AI to:
Distinguish fact, inference and assumption.
Identify missing evidence.
Challenge the preferred conclusion.
Recommend source verification.
Connect findings to a decision and next test.
From One-Off AI Research to a Repeatable Validation System
Organizations that repeatedly evaluate new products, services, ventures or market adjacencies can convert AI-assisted validation from a one-time exercise into a reusable innovation capability.
A repeatable validation system may include:
System Component | Purpose |
Standard validation framework | Ensures every opportunity addresses market, competitors, customer pain, business model and MVP evidence. |
Reusable prompt library | Supports consistent research and critical-analysis methods. |
Source-grounded research workspace | Preserves verified research, competitor evidence and decision material. |
Competitor and substitute tracker | Maintains current visibility of relevant alternatives. |
Customer discovery templates | Improves consistency of interviews and insight synthesis. |
Assumption and evidence register | Separates what is known from what requires testing. |
MVP experiment format | Links product activity to learning goals and success criteria. |
Decision memo template | Converts analysis into stage-gate decisions. |
Human review and governance standards | Protects decision quality and sensitive information. |
This capability is relevant not only to startups. It can help corporate innovation teams assess new products, established companies explore adjacent services, professional-services firms evaluate market offerings and investors or venture studios review opportunity pipelines with greater discipline.
Applications Across Different Organization Types
Startups and Founders
Generative AI can help founders pressure-test their initial idea, identify customer assumptions, prepare discovery, map competitors, develop an MVP learning plan and strengthen investor-readiness materials—while retaining direct customer evidence as the basis for product-market decisions.
Corporate Innovation Teams
AI-assisted validation can help innovation leaders compare market adjacencies, evaluate strategic fit, identify partnership or build options, draft business cases and make clearer portfolio decisions before significant development spend is approved.
Established Businesses
Organizations considering a new service, geographic expansion, customer segment or technology-enabled offering can use AI to accelerate market intelligence, analyze competitive alternatives and define pilot experiments linked to strategic decisions.
Investors, Advisors and Venture Studios
AI can support faster initial opportunity review, competitor scans, diligence question generation and risk mapping—provided the findings are treated as preparation for proper diligence rather than its replacement.
Measure the Value of AI-Assisted Validation
The value of AI in validation is not simply that a research report can be drafted more quickly. It is that leadership can allocate capital, product effort and attention more intelligently.
Value Area | Illustrative Measures |
Research speed | Time to first market landscape scan; turnaround time for competitor briefing. |
Evidence discipline | Number of material assumptions documented; percentage supported by verified sources or customer evidence. |
Customer discovery quality | Speed of interview preparation; quality of learning questions; insight synthesis turnaround. |
Product learning | Time to MVP definition; number of riskiest assumptions linked to experiments; test-cycle speed. |
Decision quality | Stage-gate decisions supported by explicit evidence; weak ideas paused or stopped before substantial investment. |
Commercial readiness | Improved clarity of buyer, pricing and go-to-market assumptions; stronger investor or executive briefs. |
Risk management | Unsupported AI claims identified; contradictory evidence captured; review compliance for material outputs. |
The ROI of AI-assisted validation is not only faster research. It is better-informed allocation of scarce capital, product capacity and leadership attention.
A Two-Week AI-Assisted Validation Sprint
A focused two-week sprint can help leaders determine whether an idea merits deeper investment, requires a pivot or should be stopped before major resources are committed.
Days 1–2: Define the Opportunity and Assumptions
Activities:
Define the business idea, intended customer and expected value proposition.
Identify market, customer, commercial and execution assumptions.
Clarify what decision the sprint must inform.
Define preliminary proceed, pivot, pause or stop criteria.
Deliverable: Assumption map and validation decision brief.
Days 3–5: Conduct Market and Competitor Scan
Activities:
Use AI-assisted research to structure the market landscape.
Map direct, indirect, substitute and legacy alternatives.
Gather and verify key external sources.
Identify uncertainty, differentiation and evidence gaps.
Deliverable: Source-grounded market and competitor brief.
Days 6–8: Prepare Customer Discovery and Stress-Test the Model
Activities:
Define customer and buyer hypotheses.
Prepare non-leading interview questions.
Review possible adoption barriers and willingness-to-pay assumptions.
Critique revenue model, pricing, acquisition and defensibility.
Deliverable: Customer discovery plan and business-model risk register.
Days 9–11: Define MVP and Validation Experiments
Activities:
Identify the riskiest unproven assumptions.
Develop MVP or experiment alternatives.
Define target users and evidence to collect.
Establish success, failure and learning criteria.
Deliverable: MVP validation roadmap and experiment plan.
Days 12–14: Prepare Executive Decision Memo
Activities:
Consolidate verified evidence, assumptions, contradictory signals and unresolved risks.
Determine the most appropriate decision option.
Specify the next cheapest credible test or implementation action.
Assign decision ownership and follow-up measures.
Deliverable: Decision-ready validation memo: proceed, narrow, pivot, pause or stop.
Executive Takeaway
The objective of a validation sprint is not to produce a positive case for every idea. It is to improve the speed and quality with which leadership identifies what deserves further investment.
Illustrative Example: Validating a New AI-Enabled Compliance Service
Consider a company evaluating an AI-enabled compliance-support service for mid-sized organizations in a regulated sector.
Initial Idea
Develop a broad AI platform that helps organizations manage compliance documentation, identify gaps and prepare for audits.
AI-Assisted Validation Questions
Which regulatory or documentation pressures create genuine customer urgency?
Who owns the problem and the budget: compliance leaders, operational managers, finance, legal or external advisors?
What tools, consultants or internal workarounds are already used?
Where would customers accept AI assistance, and where would they require qualified human oversight?
What liability, confidentiality and trust issues could prevent adoption?
Which narrow workflow would enable a credible MVP test?
Potential Learning Outcome
The initial broad platform concept may appear attractive, but the research and customer discovery plan may reveal that the most credible starting wedge is narrower: AI-assisted audit-preparation documentation for a particular type of buyer, with explicit human review and controlled source materials.
Strategic Lesson
AI-assisted validation helps leaders narrow an idea before overinvesting. It can reveal that the real opportunity is not the largest product concept first imagined, but the smallest trusted workflow customers are ready to adopt.
AI Helps Leaders Learn Before They Spend
Generative AI is changing the economics of early research and opportunity assessment. Teams can scan markets more broadly, organize competitor intelligence more quickly, challenge business-model assumptions more systematically and prepare stronger customer-discovery and MVP experiments with less delay.
But speed alone is not a strategy. A fluent AI-generated answer is not market proof. A customer persona is not a customer commitment. A market-size estimate is not an investment case until its assumptions and sources have been verified. A product concept is not product-market fit until real buyers act.
The strongest leaders will use AI neither as a cheerleader nor as a substitute for evidence. They will use it as a disciplined analytical capability that makes uncertainty visible earlier and helps the organization determine what to test next.
The best use of Generative AI in market validation is not to make every idea look promising. It is to help leaders discover which assumptions are strong, which are weak and what must be tested before capital is committed.
In markets where time matters, the organizations that learn before they spend will be better positioned to build what customers actually value.
Use AI to Accelerate Evidence-Based Innovation Decisions
MENTOR Global Consultants helps startups, executive teams and innovation leaders apply Generative AI to market research, competitor intelligence, business idea validation, customer-discovery preparation, MVP experiment design and decision-ready strategy development.
Our approach keeps verified evidence and human judgment at the center: AI helps leadership teams investigate assumptions faster, structure evidence more clearly and define the next practical test before major investment decisions are made.
