If you research industry coverage, you need to find relevant reporting even when it uses different words from your own description of the market. A keyword alert finds coverage containing the words you registered. A meaning search finds coverage about the subject whatever words it used. When you are watching a market and not a name, that difference decides what you ever see.
The gap keyword alerts leave
Set an alert on your category's standard term and you collect the articles that use it. You do not collect the three-thousand-word piece about your market written by somebody who never once reached for that term.
Writers do not agree on vocabulary. An article about restaurants failing to fill shifts may never say "workforce management". A piece about the regulation that governs your industry may name the regulation and never the industry. A podcast about the problem you solve will describe it in the words of the people who have the problem.
Those are usually the best documents you will find, because they are about the problem and not about the product category.
How a meaning search finds them
Your query becomes a position. Stored documents are compared against it by position, not by matching words. A document about your subject sits close by whether or not it shares any of your vocabulary.
So you describe what you are after. You stop guessing which words the writer picked.
When each method is right
| You are looking for | Use | Because |
|---|---|---|
| Mentions of your brand | Keyword | The name is exact and near misses are noise |
| Mentions of a competitor | Keyword | Same reason |
| A named regulation, standard or product | Keyword | The term is the thing |
| Coverage of your market | Meaning | Writers use their own vocabulary |
| Articles about a problem you solve | Meaning | The problem gets described, not named |
| Publications to pitch | Meaning | You want subject fit, not phrase matching |
Brand monitoring is a keyword job. Market monitoring is a meaning job. Most teams need both and set up the first one only.
Reading a result set correctly
A similarity search reads up to 250 candidate documents and returns 25 rows once they are grouped. So two things are true of every result you see.
It is a shortlist. Documents outside the candidate set were never considered. Nothing found does not mean nothing exists.
Narrowing the query changes which candidates get read at all. A specific search is a different search, not a filtered version of the broad one. When something you expected is missing, describe it another way before you decide it is not there.
The coverage window
The News Index keeps two monthly tables, so it reaches this month and last month.
It tells you what is being published now. It is not an archive. For a market you follow week to week that window is usually fine, because coverage from two years ago rarely changes this quarter's plan. When something does matter, save it while it is still in range.
Writing a query that works
Describe the subject the way somebody writing about it would. Not the way you name it internally.
Weak: workforce management restaurants. A category label finds category-label articles.
Better: restaurants struggling to fill shifts at short notice and the effect on service quality. That describes the situation an article would cover.
Longer and more specific wins here. Keyword search trained you to do the opposite.
What to do with what you find
Coverage of your market is research and prospect discovery in the same pass. Ask three things of every relevant article. Would you want to appear in this publication? Does this writer cover the subject regularly? Does the piece raise a question your own content should be answering?
The third one pays best and gets skipped most. An article describing a problem your buyers have, written by somebody with no stake in your product, tells you exactly what your market is thinking about.
When researching a subject, choose the search method for the question you are asking. Open the results to confirm their relevance before using them in a brief or report.
See the News Index, researching industry news, or what each search method measures.