Intelligence Reports · Guide
Where market signals actually appear
Most competitive intelligence is gathered from one kind of source: English-language trade press. It is the easiest to reach and the last place a signal shows up. The things worth knowing early appear somewhere else first.
Written from running weekly market monitoring for organizations from lean teams to a Fortune Global 500 enterprise.
Three kinds of source, and what each is for
Practical market monitoring draws on three source classes. They behave completely differently, and a program built on only one inherits that class’s blind spot wholesale.
Good for: confirmed developments, deals that have closed, executive moves, funding rounds. Written to be understood, usually accurate, easy to verify.
Blind spot: it is a lagging indicator. By the time a trade title covers a competitor’s expansion, the lease is signed and the hires are made. And its attention follows its readership, which is rarely the same as your market.
Good for: the earliest signal on anything physical or regional. A plant, a depot, a permit, a regional hire, a municipal dispute. The local paper covers it because it is local news, weeks before any English trade title decides it is sector news.
Blind spot: it is hard to reach, and machine translation at headline level gives you a lead rather than a confirmed fact. Treat it as a prompt to look, not as something to act on unverified.
Good for: tenders and RFPs, with the one thing news never carries — a deadline, a buyer and often a value. Structured, official, and actionable on the day it appears.
Blind spot: enormous volume, and titles written for legal precision rather than clarity. Most of what matches your keywords will not be biddable by you, which makes filtering the whole job.
The three are complementary rather than redundant. Procurement tells you what is being bought. Local press tells you what is being built. Trade press tells you what has already happened, and confirms the other two.
What monitoring in English only costs you
This one is measurable, so we measured it on our own publication. Across the sources we put in front of readers, roughly a quarter are not in English, spread across some forty languages — and most stories carry at least one. An English-only program does not lose a quarter of the detail evenly; it loses whole perspectives, and disproportionately the ones nearest the event.
That figure is a floor rather than a ceiling. It counts only sources we could positively identify as non-English, and it is drawn from international news rather than from any single industry — but the direction holds anywhere the thing you are watching happens outside your own market.
There is a second, subtler cost. Reaching more sources does not automatically mean knowing more. In our study of where the world’s news copy actually comes from, most outlets covering an international story were not reporting it themselves. Five hits on a keyword can be one account, republished five times. Counting sources is not corroboration.
Reach is not evenly distributed
It is tempting to assume that a wide net catches the world evenly. It does not. Our study of which national presses actually get heard found coverage concentrating hard on a handful of countries, with much of the world barely represented — and that is from a process actively trying for breadth.
For market monitoring the implication is concrete: if a market matters to you, it has to be named. Left unstated, it will be crowded out by wherever the news happens to be loudest that month, which is usually wherever the current conflict is.
Why keyword alerts underperform
Keyword alerting is the default entry point into competitive intelligence, and it fails for a structural reason rather than a quality one. A keyword matches strings; relevance is about the business. Those are different problems.
- Your company name catches recruitment posts and passing mentions, not developments.
- A competitor’s name catches their marketing, which is what they want you reading.
- A product category catches everything in the category, most of it in markets you do not serve.
- Nothing catches the item that matters when it is phrased in another language, or as a tender reference number.
The fix is not better keywords. It is judging each item against a description of the business — what it sells, where, to whom, and what it wins on — and discarding everything that merely matches. That is a slower and more expensive step, and it is the only one that turns matches into intelligence.
A practical starting point
If you are setting up market monitoring from scratch, in order of return on effort:
- Write the description of your business first. Longer than feels necessary. Capabilities, sectors, geography, what is in scope and explicitly what is not. Everything downstream is bounded by this.
- Add procurement portals for the markets you actually bid in. Highest ratio of actionable to interesting, and the only source class with deadlines.
- Add trade press for confirmation, not for discovery.
- Add local-language sources for the regions that matter, and accept that these are leads to check rather than facts to act on.
- Decide who receives what before the first report, not after.
Most programs do these in reverse: start with alerts, add sources when something is missed, and never write the description at all. That produces volume quickly and intelligence never.
Related guides
Report designWhat a competitor monitoring report should actually containThe structure, a worked example, and why the hard part is deciding what to leave out.Read the guide →
CadenceWhat to monitor, and how often each one movesThe eight categories worth tracking, and why only two of them reward watching daily.Read the guide →
We read all three, in over a hundred languages.
Trade press, local-language sources and public tender portals — judged against your priorities and delivered as a short weekly report, with every item linked to where it came from.