ORRJO Intelligence, the research platform
We use AI to read your market, and we built ORRJO Intelligence to check what it finds.
AI tools give us the speed. Nothing an AI produces is treated as a fact. It's treated as a lead, and an ORRJO analyst traces it back to the original source before it's written down.
30 minutes. No pitch.
One recent engagement: 220 verified sources, and 2 factual errors caught and fixed before client delivery.
Updated
Why is AI research hard to trust?
AI research tools are fast and cheap. They also ship claims nobody has verified, paraphrase buyers as if quoting them, and cite someone else's summary as if it were the original.
If you're about to commit budget to a market, you need to know who buys, what competitors charge and what AI tools say about you. You also need to know which of those answers you can rely on.
Why did ORRJO build its own research system?
Traditional research firms are rigorous, but their delivery takes months, and AI research tools are fast but unchecked. We wanted research that was quick and checked, so we built it ourselves. That meant building the stages, the rules that stop the work when evidence is missing, the label on every key claim, and the private portal your team reads it in.
It's called ORRJO Intelligence, the research platform. What comes out is four documents in one private portal: a GTM Intelligence Report, a GTM Playbook, a Messaging Pack and a Leadership Deck.
It's available on its own, and on a full ORRJO engagement it's already included. Most clients take it inside a wider programme, where they invest £4,000 to £50,000 a month. It's separate from the ORRJO Platform, the live dashboard and CRM environment that comes with every retainer.
What does ORRJO Intelligence do?
It reads your market. On a recent engagement the sources ran from the client's own website, social channels and listings to industry communities, competitor sites, reviews and pricing, and government and industry data.
It starts with what you already know. The work begins in working sessions with your team, and with the call transcripts, CRM data and previous research you choose to share. The data you share with us isn't used to train AI models. What your team tells us is labelled as your account, which means it's still to be verified and isn't repeated as market fact.
It labels every key claim. Each one has a numbered reference that names the source, the date it was read and a confidence rating: High, Moderate or Lower. Where a number can't be verified, the report says so.
It stops instead of guessing. If the original source for a claim can't be found, the work halts. A competitor figure more than 18 months old is flagged, and a buyer quote that can't be attributed to a named person is held. An ORRJO analyst then decides whether to ship the finding flagged with a lower confidence rating or hold it until it's verified.
It's checked by a reviewer who didn't write it. The research is checked against itself for contradictions, and where sources conflict, the conflict is resolved first. Then the reviewer checks the claims against live sources and the wording against the evidence.
It can keep checking after delivery. Optional add ons flag when a competitor's site changes, track whether AI tools are starting to recommend you, and warn you when a source the report cites has changed.
How does ORRJO Intelligence work?
ORRJO Intelligence works in five stages, the same on every engagement.
1. Intake. We set the question we're answering, what's in scope and what you already know.
2. Research modules. Specialised modules each take a different angle: how big the market is, who the competitors are and what they charge, what buyers say in reviews and forums, and what sends someone looking for a solution.
3. Verification. Every claim is opened at its original source and graded before it goes into the report.
4. Quality gates. The research is checked for contradictions, then by a reviewer who didn't write it.
5. Delivery. First findings by day 15. The documents go live in a private portal your whole team can use.
What does AI do, and what do people do?
In the research, AI handles source retrieval, checks on competitor sites, and format and date checks. It also helps shape the skeleton and first draft of a report.
An ORRJO analyst directs the queries, verifies every claim, rates each finding, shapes the narrative and signs off every report. The AI never decides to ship a finding. An ORRJO analyst decides, every time.
Where else does ORRJO use AI?
Across the rest of an ORRJO engagement, AI also handles enrichment, signal detection, conversation summaries and data checks. Every call on your campaign is recorded, transcribed and scored out of 100 across 17 criteria by ORRJO.ai, our sales call coaching product.
Our SDRs write every sequence and make every call.
What do the numbers look like?
One recent engagement used 220 verified sources. The pre delivery accuracy audit caught 2 factual errors, and both were fixed before client delivery.
A recent engagement drew on around 380 sources and carried 76 numbered references, with every key claim labelled with the evidence behind it.
Before anything goes out, we challenge it. In the 7 weeks to 17 November 2026, we raised more than 2,200 challenges on 59 documents before they went out. Circa 3 in 4 led to an edit.
Does ORRJO use AI?
Yes, ORRJO uses AI, and we publish where. AI does the high volume work on every ORRJO engagement, and an ORRJO analyst checks and signs off every report.
Most agencies either pretend AI is not in their stack, or pretend it is the whole stack. We do neither.
The five mechanisms behind ORRJO Intelligence are set out on our methodology page.
How does ORRJO check AI research?
ORRJO treats everything an AI produces as a lead to be checked. An analyst opens the original source before a claim is written down, and a reviewer who didn't write the report checks it before delivery.
Every key claim then has a numbered reference with its source, the date it was read and a confidence rating.
Which AI tools does ORRJO use?
ORRJO runs its research on the most capable AI models available, including Claude, alongside Perplexity for live search.
We also use OpenAI's tools: Codex on website build work, and ChatGPT as one of the engines we test when we measure what AI tools say about you.
Who can see the research ORRJO does for us?
ORRJO's client portals are private by default. Access needs a code plus each person's name and work email, pages are kept out of search engines, and every download is stamped with the name of the person who downloaded it.
Did ORRJO build the AI itself?
No. ORRJO built the research system: the stages, the checks, the evidence labels and the portal. The AI models it runs on are made by other companies.
See what it would cover in your market.
It takes 30 minutes, and we'll tell you what ORRJO Intelligence would cover for your market. It runs as a monthly retainer on 6 or 12 month contracts, with scope and price agreed on the call.