Key takeaways
- Fromagerie Ermitage, a French cheese company, receives a weekly press and social media report produced by a 93-node n8n workflow, with no manual step between collection and delivery.
- Every collected item goes through keyword scoring across 19 indicators plus a time filter; only relevant, recent content reaches the final report.
- The media monitoring market spans €0 to €32,000 per year for the same core function, according to Genpress's 2026 benchmark; the median Meltwater contract sits at $25,800 per year based on Vendr's purchase data.
- A custom workflow makes sense when the monitoring scope is precise and stable; for a heavily exposed brand or a one-off need, an off-the-shelf platform or free alerts remain the right call.
The problem: too much information, not enough signal
Collecting information has never been easier. Google Alerts is free, RSS feeds have existed for twenty years, and every social network offers keyword search. The problem for a company like Fromagerie Ermitage is not getting articles; it is getting far too many of them.
Manual press and social monitoring always follows the same slope. Week one, someone reads everything. Week three, they skim. Week ten, the tab stays closed and the watch only exists on the org chart. In between, the company has paid for hours of reading to perform a sorting job nobody ever wrote down: what deserves to be escalated, to whom, and why.
That sorting, not the collection, is where the value sits. And that sorting is exactly what can be turned into code.
What was built: collection, scoring, weekly report
The workflow built for Ermitage runs on n8n, an open-source automation tool that can be hosted on the company's own infrastructure. The chain, documented on my projects page, runs in three steps:
- Collection. The workflow watches online press and social media across the scope defined with the client, at regular intervals.
- Scoring and filtering. Every item is scored by keywords across 19 indicators, then filtered by date to discard old articles that keep resurfacing in search results.
- Delivery. A weekly report is generated automatically and sent to the teams: formatted, ranked, readable in a few minutes.
The whole chain adds up to 93 n8n nodes. That number does not measure the complexity of the need; it measures the share of reliability work: handling sources that go down, deduplication, logging every run. Readers of this journal have met this machinery before: it is the same one that produces IA Brew, my automated newsletter. The subject changes, the skeleton stays.
Why score every item against 19 indicators?
Because a single keyword tells you nothing. The word "cheese" shows up in thousands of pieces of content every day; most are of no interest to a company in the sector. A weighted indicator grid answers a different question: does this content touch a subject that matters to this particular company, right now?
The detail of the 19 indicators belongs to the client. The principle, though, transfers to any SMB: you turn the company's priorities into weighted keyword families, every item gets a score, and a threshold decides what makes the report. The output is not a pile of articles; it is a hierarchy: what demands a full read, what deserves a glance, what gets discarded.
The time filter completes the grid. Search engines regularly resurface content that is months old; without a date guardrail, a monitoring report fills up with stale news that looks fresh. It is an invisible detail in a demo, and one of the first reasons teams lose trust in an automated watch.
How much does media monitoring cost in 2026?
The question deserves numbers, because the spread is startling. According to the benchmark published by Genpress in 2026 (source in French), the same core function, watching what is being said, costs anywhere from €0 to €32,000 per year depending on the tool. On the international platform side, purchase data compiled by Vendr puts the median Meltwater contract at $25,800 per year, across 126 analyzed purchases. The table below sums up the orders of magnitude.
| Option | Indicative cost | Best for | Main limit |
|---|---|---|---|
| Free alerts (Google Alerts, Talkwalker Alerts) | €0 | One-off need, zero budget | Raw collection, no sorting, no social media |
| Light tools (Mention, Brand24) | €25 to €249 per month | Small brand, simple social listening | Generic scoring, poor fit for a specific trade |
| Platforms (Cision, Meltwater) | €400 to €4,000 per month | Heavily exposed brand, PR team | High annual cost, features largely unused by SMBs |
| Custom n8n workflow | One-time build, quoted (Sprint from €400) | SMB with a precise, stable monitoring scope | No TV, radio or proprietary media databases |
A platform is rented; a workflow is owned. The honest comparison is therefore not the first month, it is three years of subscription against a build paid for once, tuning included.
When a custom workflow is the wrong choice
Three situations where I advise against ordering a custom watch:
- Your brand is heavily exposed. If you appear on TV or radio and journalists call you, a platform like Cision or Meltwater brings media databases, broadcast transcripts and press contacts a workflow cannot replicate. The price is justified.
- Your watch takes less than 30 minutes a week. Well-tuned free alerts are enough. Automating a process that costs almost nothing is the surest way to lose money.
- Your scope changes every month. Custom scoring is tuned for a stable scope. If your topics keep shifting, a light tool at €25 per month like Mention will give you more flexibility.
Custom wins in one precise case: a stable business scope, a content volume that overwhelms manual sorting, and the will to own the tool rather than rent a platform of which you would use only a fraction of the features.
What you can transpose to your own business
The Ermitage machinery goes beyond press monitoring. Tender announcements, customer reviews, competitor prices, competitors' job postings: whenever public information arrives in volumes too large to read, the same chain applies. Collect, score against your own criteria, filter by date, deliver in a format someone actually reads.
Three questions to test whether this concerns you: how many hours per month does your team spend reading in order to sort; which recent decision would you have made earlier with the right information; and who, today, decides what deserves to be escalated? If the third answer is "nobody", the problem is not the tool.
If you want to see what an automated watch would produce on your own scope, the simplest step is to talk it through: book a 30-minute call, and see how I build this kind of chain on the n8n automation page. At minimum you will leave with an honest opinion on the best option for your case, including when that option is not me.
Frequently asked questions
How much does an automated media watch with n8n cost?
The build is billed once, quoted based on the number of sources and the depth of the scoring; my engagements start with a Sprint from €400. After that, running costs come down to hosting the workflow, whereas a media monitoring platform charges €400 to €4,000 per month, every month.
How is this different from Google Alerts?
Google Alerts collects but does not sort: no scoring, no social media coverage, no formatted report. A custom workflow adds the part that carries the value: scoring every item against your own criteria and delivering a ranked report the team can read in minutes.
Does a custom workflow replace a media monitoring platform?
Not always. Platforms such as Cision or Meltwater cover TV, radio and proprietary media contact databases that a workflow cannot replicate. Custom wins when the scope is precise and stable and the real need is sorting web and social content, not press relations.
Worth remembering. The value of a media watch is not in the collection; it is in the sorting, and sorting can be coded as soon as the criteria are explicit. Before paying for a subscription or ordering a workflow, write down in plain words what deserves to be escalated in your company. If nobody can write it, no tool will know it for you.