[Skip to content](https://academy.innguma.com/en/intelligence-in-the-organisation/continuous-improvement-turning-every-monitoring-cycle-into-organisational-learning/#contenido)

Lesson 6.5 · Learning path 06

# Continuous improvement: turning every monitoring cycle into organisational learning

How do you make the system evolve when questions, sources or technologies change?

**You take away** a 90-day improvement charter.

≈ 30 min · Advanced

[Do the practice](https://academy.innguma.com/en/intelligence-in-the-organisation/continuous-improvement-turning-every-monitoring-cycle-into-organisational-learning/#hazlo-tu-tu-carta-de-mejora-a-90-dias)[Exercise](https://academy.innguma.com/en/intelligence-in-the-organisation/continuous-improvement-turning-every-monitoring-cycle-into-organisational-learning/#compruebalo)[Read the lesson](https://academy.innguma.com/en/intelligence-in-the-organisation/continuous-improvement-turning-every-monitoring-cycle-into-organisational-learning/#no-basta-con-instalar-un-proceso)

In 1 minute

Installing a process is not enough.

## Installing a process is not enough

A monitoring plan ages. Strategy changes, new sources appear, others stop publishing, recipients have different needs and AI tools change which tasks can be assisted. That is why monitoring must not end at implementation: it needs **scheduled reviews and event-triggered reviews**.

An effective review does not start by asking “which features do we add?”, but “which decisions were we meant to support, and which ones do we still support?”. It then examines coverage, quality, usefulness, risks and the capacity of the people involved. Sometimes improving means removing sources, simplifying a newsletter or retiring an indicator.

## A 90-day cycle

**Days 1–15 · Baseline.** Select a priority question, review the latest products and identify an observed friction point: broken sources, duplicates, low usage or conclusions without traceability.

**Days 16–30 · Improvement hypothesis.** State what you will change, why, and with what evidence you will assess it. Avoid “use more AI” as a goal; specify the task: “reduce classification time without increasing critical omissions”.

**Days 31–60 · Bounded trial.** Change one process for a controlled scope; keep the previous procedure as a reference when feasible. Identify quality controls and review owners.

**Days 61–75 · Assessment.** Check improvements, costs, errors and unintended effects. The assessment must include users and recipients, not just analysts.

**Days 76–90 · Decision and documentation.** Adopt, modify, reverse or repeat the trial. Record changes in methodology, the training required and the date of the next review.

The 90-day cycle is **an educational proposal**, not a UNE 166006 requirement or a cadence that is valid for every organisation.

Interactive · Explore step by step

### A 90-day cycle, stage by stage

Go through the five stages with the example of the general newsletter nobody uses.

The 90-day improvement cycle in five stages.

## AI as an object of improvement, not as a judge of itself

Assistants can help propose filters, structure findings or prepare drafts. Every automation must come with an **assessment using known cases, difficult samples and monitoring for omissions**. A faster result can hide bias or information that was wrongly discarded. The system must allow human review, log errors and stop automations when the risks outweigh their benefits.

The key questions evolve: if Company A decides to run an industrial trial, general competitor monitoring is no longer enough. It needs information on technical integration, potential partners, operational requirements and purchasing criteria. The plan changes because **the decision has changed**, not because a new tool has appeared.

Company A case · simulation

Fictional case

## Company A case · The system that continues

The fictional management of Company A receives an intelligence note. A limited assessment is agreed, an owner is assigned, sources and hypotheses are logged, and signals that will trigger a review are set. After several weeks, the pilot results may reinforce, weaken or reformulate the initial interpretations. Intelligence is valuable because it lets you learn from change, not because it “gets the future right”.

Educational simulation: fictional information and organisations.

Do it yourself

## Do it yourself · Your 90-day improvement charter

Choose **one** friction point: filter quality, source scope, too many newsletters, no feedback or little versioning. Describe the baseline, intervention, risk, owner, metric and final decision criterion. Also write the assumption that, if it fails, would force you to abandon the experiment.

**Final deliverable of the learning path:** a **minimum viable intelligence system** that includes a product for a recipient, a memory sheet, roles, three indicators and a 90-day plan. It can be implemented with conventional tools and without buying software.

R14 · Intelligence product 

### Intelligence product specification and briefing

Define the need and deliver actionable conclusions.

[See the template →](https://academy.innguma.com/en/library/intelligence-product-specification-and-briefing/)

R06 · Evidence 

### Log of signals and evidence

Distinguish observation, source and interpretation.

[See the template →](https://academy.innguma.com/en/library/log-of-signals-and-evidence/)

Check it · self-assessment

## Check your improvement charter

This part needs professional judgement: there is no automatic marking. Complete your work in the practice, rate it with the lesson's rubric and compare it with the indicative solution.

What your answer should include

-   A single friction point with an observable baseline.
-   A bounded intervention with an owner, a metric and a risk.
-   An adoption or reversal criterion and a stop rule.
-   The assumption that, if it fails, forces you to abandon the experiment.

Self-assessment rubric (0–2 per criterion)

Educational assessment: it does not audit UNE conformity.

Observable problem

Bounded experiment

Metric and risk

Owner

Adoption or reversal criterion

**Indicative solution**

Problem: a generalist newsletter with no feedback. Baseline: interviews that reveal a lack of use (simulation). Intervention: a short version per decision for one area; a bounded trial, quality review, interviews and a decision to adopt it if it improves usefulness without losing coverage.

**Mistake to avoid:** "Installing AI" with no task, baseline or controls.

Illustrative solution for a fictional case: reasoned alternatives are valid.

## 90-day experiment · Fix a practice, don’t add technology

The experiment applies the 90-day cycle described above; it groups the improvement hypothesis and the bounded trial into a single stretch (days 16–60).

**Fictional hypothesis, Company A:** general briefings go unused because they do not identify a decision and they mix news with inferences. It is not assumed that introducing AI is the solution.

| Stretch | Action | Acceptance evidence | Proposed owner |
| --- | --- | --- | --- |
| Days 1–15 | Select one recipient, analyse three deliveries and gather a baseline | Log of what is understood, used and questioned | Product owner |
| Days 16–60 | Formulate the hypothesis and test a briefing per decision, with sources and doubts visible | Two versioned products and feedback | Analyst and reviewer |
| Days 61–75 | Compare the format with the previous practice and log incidents | Changes in usage, coverage and review time | Methodology lead |
| Days 76–90 | Decide on adoption, adjustment or reversal | Minutes with evidence, owner and next review | Process sponsor |

**Stop rule:** halt the trial if the format omits a material alert, exposes sensitive information or cannot be audited. **Adoption rule:** approve the change only if it improves understanding or usage and maintains coverage of critical risks. If the interviews do not allow a conclusion about improvement, document “inconclusive result”.

**Final challenge:** submit a file that connects R14 (product), R06 (evidence), R15 (roles and versions) and R16 (improvement). Ask an outside person to reconstruct a decision and to propose which piece of data would force a review of it.

Interactive · Simulation

### You decide at the end of the experiment

Simulation: decisions in the 90-day experiment.

## References and limits

-   UNE · Asociación Española de Normalización. *UNE 166006:2018. Gestión de la I+D+i: Sistema de vigilancia e inteligencia* (2018). [Document](https://www.une.org/encuentra-tu-norma/busca-tu-norma/norma?c=N0059973) Reference framework for watch and intelligence systems. It is cited only as context, never as a source of requirements. Reference consulted on 2026-10-18.

-   Company A is a fictional case: its data, companies and figures do not describe real markets or organisations.
-   The indicators and rubrics are educational instruments, not measured results of Innguma or its clients.

Library · R14

## Intelligence product specification and briefing

Putting it into practice? Download template R14 · Intelligence product specification and briefing and fill it in with your case.

- [Download R14 (Excel)](https://academy.innguma.com/en/library/descargas/R14-intelligence-product-specification-and-briefing.xlsx)
- [Back to the learning path](https://academy.innguma.com/en/intelligence-in-the-organisation/)

**Download** 

-   [R14 · Intelligence product specification and briefing](https://academy.innguma.com/en/library/intelligence-product-specification-and-briefing/)
-   [R16 · Maturity assessment and 90-day plan](https://academy.innguma.com/en/library/maturity-assessment-and-90-day-plan/)
-   [R15 · Governance, responsibilities and versions](https://academy.innguma.com/en/library/governance-responsibilities-and-versions/)
-   [R01 · Map of decisions and intelligence questions](https://academy.innguma.com/en/library/map-of-decisions-and-intelligence-questions/)
-   [R06 · Log of signals and evidence](https://academy.innguma.com/en/library/log-of-signals-and-evidence/)

Open reading, no registration.