[Skip to content](https://www.innguma.com/en/resources/faq/relevance/#contenido)

Frequently asked questions

# Who decides what is relevant in watching?

**Short answer** 

Your organisation does. The relevance of a signal is measured against the information needs defined by whoever takes decisions, not by its popularity or volume. At Innguma, the AI filters and classifies signals according to the criteria configured, and people decide what is watched, what is published and who receives it.

## What makes a signal relevant?

Palop and Vicente defined watching as an ‘organised, selective and permanent’ way of capturing information from outside[\[1\]](https://www.innguma.com/en/resources/faq/relevance/#ref-1): selecting is part of the definition. UNE 166006:2018 directs the system towards decisions based on relevant, reliable information to which value has been added[\[2\]](https://www.innguma.com/en/resources/faq/relevance/#ref-2). In practice, a signal is relevant when it meets these criteria at the same time:

-   **Pertinence:** it answers a defined information need, not just a topic.
-   **Novelty:** it adds something that was not already in the history.
-   **Reliability:** it comes from a source that can be identified and checked.
-   **Timeliness:** it arrives before the decision is taken or the deadline expires.
-   **Impact:** it can change a decision about a product, a customer, an investment or a compliance obligation.

A widely shared news item may meet none of these; a line in an official gazette may meet all of them.

## How are information needs defined?

The competitive intelligence literature, building on Jan Herring’s work on Key Intelligence Topics (1999), groups needs into three families: information for strategic decisions, early-warning topics, and players that influence the organisation[\[3\]](https://www.innguma.com/en/resources/faq/relevance/#ref-3). To make them concrete:

1.  **Start from decisions, not topics.** ‘Batteries’ is a topic; ‘knowing whether a competitor is preparing a change of cell chemistry before the next tender’ is a need.
2.  **Ask whoever decides** what they need to know, how far in advance, and what they would do differently if they knew.
3.  **Prioritise.** Not all needs weigh the same; critical ones justify more sources and more frequent review.
4.  **Translate each need** into specific entities, technologies, terms and source types.
5.  **Set when it is reviewed**: needs change with strategy.

## Who decides each part?

| Who | What they decide | Example |
| --- | --- | --- |
| Executive management and area heads | Which information needs are priorities | Following three competitors and a draft regulation |
| Watch manager | Which watches, sources and filters cover them | Patents in two classifications, the BOE and competitors’ websites |
| The platform’s AI | Nothing: it processes according to those criteria | Groups duplicates and classifies by technology, entity and type of event |
| Analyst | Which signals are validated, commented and published | Publishes the patent application with a comment on its scope |
| Readers | Whether what they receive serves them, and say so | Ask for one topic to be removed and another added |

## What role does AI play?

Innguma’s AI filters, groups duplicates and classifies each signal by category, technology, entity, type of event, place and tone. That reduces what a person has to read, and allows the content to be browsed along those dimensions in AI Feed. But the AI does not set priorities or publish on its own: the analyst reviews and validates in Factory, the space where Innguma processes and structures signals, and decides what passes to the Library. AI processes; your team decides. Knowledge accumulates.

## How do you know whether the relevance criteria need adjusting?

-   **Many signals come in that nobody publishes:** the filter is too broad. Narrow the terms or entities.
-   **A lot is published and little is read:** the recipient or the format is failing, not the capture.
-   **An area learns of something through another channel:** a source is missing, or a need has not been defined.
-   **The same signals appear in several watches:** the needs overlap; they should be reorganised.
-   **What is published changes no decision for months:** the need is no longer a priority.

UNE 166006 follows the high-level structure of management systems, with performance evaluation and improvement[\[2\]](https://www.innguma.com/en/resources/faq/relevance/#ref-2). Applied to relevance, this means reviewing these symptoms periodically, not only when someone complains.

## Sources and references

1.  \[1\] [Palop, F. and Vicente, J. M. (eds.). Technology watch and competitive intelligence. Its potential for Spanish business. Fundación Cotec, 1999](https://www.oocities.org/es/pedrobonillo/oyc/estudio15.pdf)
2.  \[2\] [Utrilla, F. Watch and intelligence systems in R&D+i management. UNE Standardisation Review (La Revista de la Normalización Española), No. 3, 2018](https://revista.une.org/3/sistemas-de-vigilancia-e-inteligencia-en-la-gestion-de-la-id.html)
3.  \[3\] [Grèzes, V. The Definition of Competitive Intelligence Needs through a Synthesis Model. Journal of Intelligence Studies in Business, 5(1), 2015 (reviewing, among others, Herring, J. Key Intelligence Topics, Competitive Intelligence Review, 1999)](https://doi.org/10.37380/jisib.v5i1.111)

## Related

- [What happens if the AI makes a classification error?](https://www.innguma.com/en/resources/faq/ai-errors/)
- [Does AI replace analysts?](https://www.innguma.com/en/resources/faq/ai-and-analysts/)
- [Watch filters: what they are, how they work and why they decide the quality of a watch](https://www.innguma.com/en/resources/software/filters/)
- [Academy](https://academy.innguma.com/en/)

- [Request a demo](https://www.innguma.com/en/demo/)
- [Back to the FAQ](https://www.innguma.com/en/resources/faq/#faq-vigilancia)