Anatomy of a Spam Report: From the Red Button to the Blacklist

You tap "Report as spam" and, a few hours later, a business number is blacklisted. Between those two moments, a complex technical process unfolds—combining collaborative databases, machine learning algorithms, and human decisions. Understanding this mechanism is essential for any call center that wants to protect its numbers.

Step 1: The user's action

It all starts with a simple gesture. A person receives an unsolicited call. They have several options to report it:

• The smartphone's native button: on Android ("Block and report as spam") and iOS ("Report as junk"), operating systems now include a report button directly in the call log.
• Anti-spam apps: Truecaller (450 million users worldwide), Hiya, Orange Téléphone, and Tellows collect reports from their communities.
• The 33700 platform: France's official platform managed by AF2M (the French Multi-operator Multimedia Association). Users forward the SMS or report the number by texting 33700.
• "J'alerte l'Arcep": the telecom regulator's platform recorded 23,383 alerts for unsolicited calls and messages in 2025, a 113% increase over 2024.

Step 2: Database aggregation

Each report feeds into one or more databases. The path diverges depending on the channel used:

Crowdsourced databases

Truecaller and Hiya operate on a participatory model. When a user reports a number, the information is added to a central database. According to Truecaller's official documentation, spam lists are created by the user community that chooses to report unsolicited calls and SMS messages. Users can also suggest a name for the caller, enriching the collaborative directory.

An important point: Truecaller also incorporates positive feedback. If users indicate that a number is not spam, the algorithm takes this feedback into account to avoid false positives.

Carrier databases

Orange, SFR, Free, and Bouygues each maintain their own databases of reported numbers. Reports via the 33700 platform are forwarded to the relevant carriers by AF2M, which centralizes and redistributes alerts.

The regulator's platform

ARCEP, through "J'alerte l'Arcep," collects reports for regulatory purposes. The platform has accumulated nearly 380,000 reports since its launch in 2017. This data helps identify spikes, weak signals, and systemic issues—directly feeding regulatory decisions.

Step 3: The algorithm kicks in

Raw reports alone don't blacklist a number. Platforms apply sophisticated algorithms to decide whether a number deserves the "spam" label.

Multi-criteria scoring

Contrary to popular belief, there is no fixed threshold of reports that automatically triggers a spam flag. Truecaller confirms this in its technical documentation: "There is no fixed number of reports that automatically marks a number as spam. Instead, our system analyses overall patterns and behavior over time."

Typical factors considered include:

• Call volume: a number making 200+ calls/day with less than 10% answer rate presents a suspicious profile.
• Average call duration: calls systematically under 5 seconds (immediate hangup) are a strong signal.
• Report-to-call ratio: more than the absolute number, it's the proportion of recipients who report that matters.
• Report velocity: 50 reports in 1 hour weigh more than 50 reports in 1 month.
• Geographic diversity: reports from varied regions reinforce the credibility of the spam pattern.
• Number history: a recently activated number with no positive history is more vulnerable.

Machine learning

Machine learning algorithms analyze these factors in real time. They compare a number's behavior against millions of known patterns—both legitimate and fraudulent. This approach can detect nuisance before reports even start rolling in.

Step 4: The verdict and its consequences

Once a number exceeds the algorithmic threshold, several levels of sanctions apply:

Level 1: The "Likely Spam" label

The number appears with a warning on the recipient's screen. It can still be answered, but the response rate drops by 40-60% on average.

Level 2: Automatic blocking

Some apps and carriers block calls outright. The caller isn't even aware of the block—the call appears to ring normally on their end but never reaches the recipient.

Level 3: Network-level blacklist

In the most severe cases (spoofing, confirmed scams), carriers can block the number at the network level. This is the last resort, applied notably under the STIR/SHAKEN protocol and the French MAN.

Step 5: The feedback loop

The process doesn't stop at flagging. Systems incorporate a continuous feedback loop:

• Unflagging is possible: on Truecaller, businesses can contest a spam label through the Verified Business program. Hiya offers a similar process via Hiya Connect.
• Carrier appeal: businesses can contact carriers to request spam label removal, provided they can prove compliance.
• Gradual rehabilitation: a number that stops generating reports will see its score improve gradually—but the process takes weeks or even months.

Number spoofing: the wrench in the works

The reporting system suffers from a major flaw: caller ID spoofing. In 2025, spoofing reports surged by 123% on the "J'alerte l'Arcep" platform, with 10,643 additional reports in one year. According to the 2026 Customer Satisfaction Observatory, 43% of consumers reported being spoofing victims at least once in three months.

In practice, a fraudster can spoof your business number to make malicious calls. Recipients then report your number—and it's your reputation that suffers. This is precisely why ARCEP strengthened its measures:

• Mandatory masking of unauthenticated French mobile numbers originating from abroad (November 2025).
• Opening of an administrative investigation targeting carriers (January 2026).

Competitor abuse of the reporting system

Another system flaw: abusive reports orchestrated by competitors. Some businesses organize mass reporting campaigns to blacklist their rivals' numbers. Modern algorithms try to detect these coordinated patterns, but the phenomenon remains a reality for many call centers.

How to protect your numbers: 5 concrete actions

Understanding the report journey allows you to act proactively:

1. Monitor your reputation continuously — Use instant verification tools to detect any flagging before it spreads.
2. Respect volume and velocity limits — A gradual warm-up protocol is essential for new numbers.
3. Control your report rate — Target better, identify yourself clearly in the first seconds, and understand how reputation scores are calculated.
4. Register for business programs — Truecaller Verified, Hiya Connect, and carrier programs let you display your company name and reduce false positives.
5. Prepare a crisis plan — Have a protocol ready for the first 24 hours if blacklisted.