The Reuters Institute's Digital News Report 2025 delivers a figure few Moroccan communications departments have truly internalized: 78% of Moroccans get their news online, a growing share of it through social networks and messaging apps. In its 2026 edition, the same institute measures trust in news at 28% — a low level that pushes audiences toward alternative sources and accelerates the dispersion of the reputation narrative. On this terrain, Western media-monitoring tools are not calibrated.
The illusion of global coverage
The dominant platforms — Cision, Onclusive, Talkwalker, Meltwater — built their corpora on Anglo-Saxon and European aggregates. Their source grids cover the major French-language titles (Le Monde, Les Échos) and pan-African newswires, but treat the Moroccan ecosystem as a supplement. Hespress, the country's leading online news site, is often only a secondary, poorly tagged source; Medias24, TelQuel, Le360, Lakome2, H24Info are inconsistently covered; and the comments — where the real reputation conversation actually lives — are ignored.
Three blind spots of generic tools
- Real multilingualism: an average Moroccan tweet mixes 2.3 languages (Darija, French, MSA, sometimes English). Per-document classifiers get the dominant language wrong and miss the code-switching.
- The source hierarchy: in Morocco, a Hespress or Le360 comment sometimes weighs more than a Medias24 article in shaping public sentiment. Generic tools invert that hierarchy.
- Tempo: Moroccan virality builds on WhatsApp and Facebook before it reaches the press. Tools that wait for press coverage are already 24 to 48 hours behind.
What makes the Moroccan market specific
The Moroccan information landscape has three structural features no generic platform models correctly. First, extreme concentration: according to a study cited by Hespress, the site captures the largest share of online audience, followed by Le360 (17% weekly usage) and Medi 1 Radio (10%). The critical mass plays out across fewer than ten players. Second, a linguistic mix where institutional French coexists with a Darija that carries the emotional load of reputation. Third, acceleration through private messaging — WhatsApp has become a full-fledged broadcast channel, beyond the reach of classic scrapers.
The AI-generated content challenge
The 2025-2026 edition introduces a new factor: AI-generated content targeting Morocco spread massively, notably during the qualification campaign for the 2026 World Cup, according to observations relayed by Hespress. This flow — deepfakes, automated articles, coordinated accounts — pollutes unfiltered corpora and skews sentiment scores. A platform dedicated to the Moroccan market must integrate detection of this type of content, which no generic tool does correctly today.
What a dedicated platform changes
A platform built for Morocco reverses three trade-offs. It prioritizes coverage of local sources over international exhaustiveness. It treats Darija as a first-class language, not a fallback dialect. And it integrates private messaging through aggregated-signal partnerships rather than ignoring it in the name of privacy. The result is not a broader dashboard — it is a dashboard that sees the signals the others miss.
The cost of miscalibration
The cost of a miscalibrated tool is not measured in subscription euros. It is measured in missed signals. When a communications department discovers a boycott 48 hours after it effectively started — because the tool was waiting for French-language coverage — the cost is that of a crisis that has already taken shape. When a dashboard shows positive sentiment while the Darija conversation flips, the cost is that of a decision made on a false reading.
“We were paying an American tool to tell us what Moroccans thought of us. It told us what the American tool thought Moroccans thought of us. Not the same thing.”— Head of Communications, Moroccan agri-food group (anonymised)
The decision criterion
The criterion for choosing a reputation-intelligence platform in Morocco is not the size of the global corpus. It is the depth of the Moroccan corpus, the quality of Darija processing, and the ability to catch weak signals before they become headlines. On these three axes, generic tools do a disservice: they give the illusion of control while leaving invisible the majority of the signal that matters.
Harch Atelier is built on these three axes. Coverage of 30+ Moroccan and pan-African sources, a trilingual sentiment engine (French, MSA, native Darija), and early detection on aggregated messaging signals. The platform was designed for the Moroccan market — not adapted to it.