Why a Methodology Matters
Every intelligence product carries an implicit methodology. Most never state it. The data sources are listed, the features are described, the outputs are shown, but the reasoning that connects data to conclusion is treated as proprietary machinery—closed, opaque, and beyond examination. The user is asked to trust the output without being able to evaluate the process that produced it.
We do not believe that is acceptable. Intelligence that informs high-stakes decisions must be transparent about how it works, what it assumes, and where its limits are. This is not just an ethical position. It is a practical one. A methodology you cannot examine is one you cannot challenge, and one you cannot improve. We publish our methodology because we believe our clients are entitled to understand what they are relying on—and because we believe a methodology that withstands scrutiny is stronger than one that hides from it.
The Five Principles
The Forreast Intelligence Methodology is built on five principles. Each is a design choice that shapes how we collect data, build models, and deliver analysis.
1. Structure over volume. The dominant philosophy in business intelligence is that more data is better. We disagree. Data without structure is noise. A million press releases processed as text is less valuable than ten thousand processed as edges in a graph that models the relationships between the entities they describe. We prioritise the quality of the model over the quantity of the data. Our WorldGraph exists because of this principle: the graph structure is the intelligence, and the data serves it.
2. Continuous over periodic. The world does not operate on quarterly cycles. A supply chain that was resilient in March may be fragile in April because a single supplier was acquired or a single route was disrupted. Our monitoring runs continuously, and our alerts fire when the structure of the system changes, not when a calendar says it is time to reassess. This means our clients receive intelligence when it matters, not when a report is due.
3. Contextual over absolute. A risk score of 7 out of 10 means nothing without context. Is this score high for this entity's peer group? Has it been rising or falling? What structural factors drive it? We deliver intelligence that is always contextualised—against the entity's history, against its peers, against its network position. An absolute number is data. A contextualised number is intelligence.
4. Actionable over interesting. We have a simple test for every piece of intelligence we deliver: does it change a decision? If it does not—if it is merely interesting, if it confirms what was already known, if it has no decision attached—then it is not intelligence. It is content. We do not deliver content. Every signal in our system is connected to a recommended action, a structural explanation, and a measure of confidence. Our clients pay for intelligence, not for entertainment.
5. Honest about uncertainty. No intelligence is certain. Models have limits, data has gaps, and the future is not determined. We state our confidence level on every output, we identify the assumptions our analysis depends on, and we tell our clients when the evidence is insufficient to support a conclusion. This is not weakness. It is the only honest basis for decision-making. An intelligence product that never says "we don't know" is not more confident—it is less trustworthy.
How the Methodology Works in Practice
These principles are not aspirational. They are operational. They manifest in every part of our pipeline:
The WorldGraph embodies Principle 1. We model entities and their relationships, not just entity attributes. When new data arrives, it is integrated as edges, not just as updated fields. This means the intelligence grows structurally richer as the graph grows, not just larger.
Our monitoring layer embodies Principle 2. We continuously ingest data from hundreds of sources—filings, trade registries, sanctions lists, news, vessel movements—and update the graph in near real-time. Alerts are triggered by structural changes, not by schedules.
Our scoring engine embodies Principle 3. Every score is computed relative to a peer group, a network neighbourhood, and a historical baseline. No score is delivered in isolation.
Our signal pipeline embodies Principle 4. Every signal is connected to a recommended action, a structural explanation, and a confidence measure. Signals that do not change a decision are filtered out before they reach the client.
Our analysis reports embody Principle 5. Every conclusion states its confidence level, its supporting evidence, and the assumptions it depends on. When the evidence is ambiguous, we say so.
Why This Is a New Standard
The intelligence industry has operated for decades on a model that is increasingly inadequate. That model treats intelligence as a product—a report, a score, a dashboard—delivered on a schedule, with the methodology hidden behind the output. It optimises for appearance of certainty and completeness, not for the quality of the reasoning.
The Forreast Intelligence Methodology replaces that model with one that treats intelligence as a continuous, transparent, structurally grounded process. It is built for a world that is too complex and too fast-moving for the old approach. It does not promise to eliminate uncertainty. It promises to navigate it honestly, with tools that match the complexity of the environment.
This is not a feature set. It is a standard. We believe it is the standard that intelligence should have been held to all along, and we believe it is the standard that the next decade will demand. The organisations that adopt it will have a clearer view of the systems they operate in, a better basis for the decisions they make, and a stronger position in a world where the cost of being wrong is rising.
That is what we built Forreast to provide. Not a dashboard. Not a report. A methodology—and the infrastructure to execute it.