The Evidence Engine
How organisations capture and connect signals from across the business into reusable evidence
Most organisations have more knowledge than they realise.
They conduct research. They talk to customers. They run webinars and events. They have sales conversations. Their product teams accumulate expertise. Their support teams hear recurring problems. Their executives hold positions that nobody has ever written down.
That knowledge exists. What is missing is the infrastructure to capture it, connect it and make it reusable.
The Evidence Engine is that infrastructure.
What’s different about Editorial Intelligence? We build an engine that continuously turns organisational evidence into reusable knowledge.
Where this sits
Evidence only compounds when it is connected.
What the Evidence Engine does
Signals are gathered from across the organisation — sales conversations, customer success, product usage, support tickets, advisory boards, webinars and events, research programmes and executive thinking. Most of this signal is currently lost. The Evidence Engine creates the infrastructure to catch it.
Individual signals become evidence when they are connected to each other and to the narratives they support. A single customer interview is a data point. Three interviews pointing at the same problem, connected to research from six months ago, is evidence of a pattern.
Connected signals need to be interpreted — the pattern identified, the argument extracted, the strategic significance understood. This is where editorial judgement operates. The Evidence Engine creates the space for that judgement to happen systematically rather than accidentally.
Not all evidence deserves activation. Some signals refine an existing narrative. Some open a new one. Some are interesting but not yet strategically significant. The Evidence Engine creates a way of deciding what to act on now, what to hold, and what to file for future use.
Evidence ages. Research becomes dated. Customer stories need updating as businesses grow. Market positions shift as industries change. The Evidence Engine creates a system for keeping evidence current — not as a filing exercise, but as an ongoing editorial discipline.
Where evidence comes from
Most organisations draw from only a fraction of these sources. The Evidence Engine makes the full range accessible — and connected.
Where customer intelligence fits
Customer intelligence and Editorial Intelligence solve different parts of the same problem.
Customer intelligence helps an organisation understand customers: what they experience, what they need, the language they use and the patterns emerging across research, interviews, support, sales and behaviour.
Editorial Intelligence works on what happens next. The Evidence Engine connects that customer understanding with other organisational evidence and keeps it available for reuse. Editorial judgement decides what matters, what remains uncertain and what deserves development. Narrative Architecture shapes selected evidence into coherent positions. Editorial Activation turns those positions into useful expression and experiences.
In mature organisations, the opportunity is not always to produce more research. It may be to get more value from customer intelligence the organisation has already paid to create — by connecting it, reusing it and allowing new evidence to strengthen or challenge what is already known.
Editorial Intelligence does not replace customer research, insights or UX research. It gives their outputs somewhere to compound.
Your meetings are already producing evidence
AI notetakers such as Granola and Otter make it possible to capture far more of the knowledge generated in everyday conversations. A customer interview may contain distinctive customer language. A sales call may reveal recurring objections or buying signals. An internal meeting may contain SME or executive expertise that has never been formally documented.
But recording and summarising a meeting is not the same as building organisational intelligence.
The transcript remains the source of record. AI summaries can help people navigate a conversation and extract promising signals, but they are not automatically trusted evidence. Important claims, quotes and interpretations should be checked against the original transcript.
Observations become more useful when they are connected across multiple conversations and other evidence sources. Editorial judgement is still required to decide what matters, what remains uncertain and what should be connected or developed.
The goal is not to accumulate meeting summaries. It is to turn conversation archives into reusable organisational evidence.
Context is capital
Preserving context used to have an awkward economic problem. Organisations could store enormous amounts of material — interview notes, research, transcripts, presentations, customer stories and internal documents — without having a practical way to revisit and compare all of it. Storage was cheap. Reuse was expensive.
AI changes that equation. Retrieval, comparison, clustering and first-pass synthesis can now happen across volumes of material that would be unrealistic for a person to review manually every time a new question appears.
That makes accumulated context more valuable — but only when it is usable.
A pile of documents is not capital. Context becomes strategically valuable when it is maintained, retrievable, connected to its provenance and available to improve future work.
The important change is not that AI eliminates the need to think deeply. It increases the reuse value of thinking that has already happened. A customer interview can inform more than one case study. A research project can keep generating new questions after its launch campaign ends. An observation from six months ago can become significant when new evidence arrives today.
This is why the Evidence Engine is not simply an archive. Its purpose is to preserve enough structure around useful material that future people and AI systems can interrogate it without starting from zero.
AI compresses the cost of retrieval. Editorial judgement determines what deserves to compound.
The difference between knowledge and evidence
Knowledge is what the organisation knows.
Evidence is what the organisation can demonstrate.
A company may know that its customers struggle with a particular operational problem. That is knowledge. When it has research data, customer stories, sales conversation patterns and product usage data all pointing to the same conclusion, it has evidence.
The distinction matters for three reasons.
For narrative credibility. Knowledge can be asserted. Evidence can be demonstrated. In markets where everyone claims authority, evidence creates the difference between a position that holds up under scrutiny and one that doesn’t.
For AI visibility. AI systems infer authority from patterns of connected evidence across multiple sources. An organisation that can only point to its own assertions creates weak signals. An organisation with connected evidence from research, customer stories, industry bodies and earned media creates strong ones.
For compounding. Knowledge stays inside the organisation. Evidence, once captured and connected, can be activated, distributed and built upon. Each new piece of evidence either strengthens an existing position or reveals a new one.
The Evidence Engine in practice
Hidden Hours shows what a working Evidence Engine looks like over time.
It began with a research question about the operational burden on accounting professionals. The research generated initial evidence. That evidence revealed a pattern — structural pressure on the profession, not just isolated inefficiencies.
From there, the Evidence Engine expanded the evidence base:
- Customer interviews added practitioner experience
- Case studies demonstrated the pattern in specific organisations
- Event conversations tested the narrative with live audiences and refined it
- Sales conversations revealed which elements resonated most
- Regulatory changes added new evidence layers
- Product usage data connected operational evidence to specific solutions
Each new source added to the same evidence base. Each new piece of evidence either strengthened the position or refined it. The narrative is substantially more authoritative than it was at launch — not because more content was produced, but because more evidence was connected.
That is the compounding effect the Evidence Engine creates.
What the Evidence Engine is not
It is not a content archive. Archives store finished work. The Evidence Engine stores signals, insights and evidence — the material that future work is built from.
It is not a CRM. Customer relationship management systems track interactions. The Evidence Engine extracts the strategic intelligence from those interactions.
It is not a research repository. Research repositories store reports. The Evidence Engine connects research findings to the narratives they support and the other evidence that reinforces them.
It is the layer that sits underneath all of those systems — connecting what they separately hold into something the organisation can actually use.
The five flagship frameworks
The Evidence Engine is the second of five interconnected frameworks that together describe the Editorial Intelligence operating model:
- Editorial Intelligence — the discipline
- The Evidence Engine — the infrastructure that feeds everything else
- Narrative Architecture — how evidence becomes connected strategic position
- Editorial Activation — how narratives reach buyers and AI systems
- AI Visibility & Buyability — how reputation influences discovery and purchase decisions
Without the Evidence Engine, the other four frameworks have nothing to work with. It is the foundation of everything that compounds.
Connected frameworks: Editorial Intelligence · Narrative Architecture · Editorial Activation · AI Visibility & Buyability · Daily OS · EI for product teams
This framework forms part of Editorial Intelligence OS — the practical operating system built to make the discipline repeatable.