We Analysed 200 Spam Messages Sent to Nigerians: Here’s How Fraudsters Try to Make People Act

We Analysed 200 Scam Messages Sent to Nigerians

An unexpected text message promises free airtime. Another urges someone to reply to a short code. A third reports that a mobile-data subscription has expired.

Which of them is a scam?

The answer is more complicated than simply checking whether a message contains the words free, urgent or reply.

TechView Africa examined 200 SMS records labelled as spam in a Nigerian research dataset, checking repeated messages and observable language patterns to understand the difference between unwanted communication and evidence of fraud.

Our analysis found 182 distinct message texts among the 200 sampled records. Many contained promotional language, payment amounts or instructions to respond. But the sample also included ordinary account notifications and other messages that could not reasonably be declared fraudulent from their wording alone.

How TechView Africa conducted the analysis

We examined a reproducible 200-record sample from the spam-labelled portion of the publicly available ExAIS_SMS Nigerian dataset. The underlying research project, conducted at the Federal University of Agriculture, Abeokuta, collected 5,240 SMS records from 20 participants. Its authors categorised 2,350 as spam and 2,890 as non-spam.

For our 16 October 2026 verification, we analysed message repetition and screened the 200 selected texts for explicitly defined words relating to action requests, money, rewards, subscriptions and time.

We distinguished measurable text patterns from conclusions about criminal intent. The messages are historical, including examples from 2014, and should not be mistaken for a representative snapshot of scams circulating in Nigeria in 2026.

What we found in the 200 messages

Our first discovery was that the number of records is not the number of distinct messages. Of the 200 sampled SMS records, 182 had distinct message texts. The other 18 were repeated-text occurrences.

Repetition matters because it can amplify unwanted communication. However, identical texts do not establish how many senders were involved or whether any particular message formed part of a criminal campaign.

We then screened the sample for observable language patterns.

Source: TechView Africa's independent screening of 200 ExAIS_SMS records. The categories overlap and count word matches, not confirmed scams.

1. Requests to act are common—but not automatically suspicious

Eighty-three messages contained at least one of the terms reply, text, dial, call or send. That finding is interesting because requests for action appear throughout commercial SMS communication.

A telecommunications company may legitimately ask subscribers to dial a code to check their balance. A promotional service may ask customers to text a short code to register.

A scammer can use a similar instruction to direct someone towards a fraudulent website, payment or identity-verification process. The instruction alone does not settle the question. The surrounding claim and the destination of the requested action matter.

2. The word “free” is not evidence of fraud

We found the word free in 35 sampled messages. Some messages advertised free trials, promotional music services or other commercial offers. Those offers may have conditions, including automatic renewal or subsequent charges.

A free offer can be misleading without necessarily being a criminal scam. The Economic and Financial Crimes Commission's scam-awareness guidance explains a more serious pattern: criminals may announce unexpected lottery winnings and subsequently demand an advance payment. That is meaningfully different from an ordinary promotional text advertising a trial.

Our assessment: Look beyond the promised benefit. Ask who sent the offer, whether you participated and what you are being asked to surrender or pay.

3. Money appears frequently, but context changes its meaning

Sixty-two messages matched our search for Naira amounts, fees or similar monetary expressions. The sample included promotional pricing, subscription charges and service-account information.

A message mentioning money is not necessarily attempting to steal it. For example, a genuine service notification may explain a charge already incurred. By contrast, a message impersonating a financial institution and directing the recipient to disclose a PIN would raise a substantially different concern.

The Central Bank of Nigeria's fraud-awareness guidance identifies impersonation and attempts to obtain confidential information as important indicators of phishing and social engineering. Readers should therefore distinguish a notice about a transaction from an unexpected instruction to authorise another one.

4. Time-related language needs careful interpretation

Our automated screening detected time-related words in 41 records. But a word such as today can appear in an ordinary birthday notification, while expires may describe a legitimate mobile-data subscription.

Neither is automatically an attempt to manipulate the recipient. The concern becomes greater when the sender uses a deadline, threat or supposed emergency to prevent verification. Nigeria's Electronic Fraud Forum advises consumers to be cautious when unexpected messages demand immediate action. That distinction is central to understanding why simple keyword filtering cannot reliably determine criminal intent.

5. The Nigerian context is important

The ExAIS_SMS dataset is especially useful because it contains communications from Nigerian mobile-phone users rather than relying entirely on overseas datasets.

Its mixture of telecom promotions, account notices, informal messages and other content demonstrates that a spam label can cover very different kinds of communication. But the corpus also has limitations. It comes from a small university-community group, includes historical messages and does not establish national fraud prevalence.

For a separate perspective on genuinely documented scam language, Nigerian researcher Temitope Michael Ajayi examined more than 200 scam emails and 50 SMS or related messages collected between 2018 and 2022. That 2022 linguistic study identified techniques involving false alarms, reassurance, social conventions and religious references.

The two research sources answer different questions: ExAIS_SMS concerns spam classification, while Ajayi's work examines manipulative language in scam communications.

What Nigerians should check before responding

The lesson is not to trust every promotional message. Instead, assess what the sender wants you to do. A suspicious message deserves additional scrutiny when it unexpectedly asks you to disclose a password or OTP, send money, open an unfamiliar login page or contact someone through details supplied only in the message.

TechView Africa's guide to fake customer-care accounts explains how criminals exploit people's need for assistance. Our investigation into fake support numbers appearing in Google Search illustrates the same verification problem through another channel. The wider financial consequences are explored in our report on 67,518 Nigerian digital-payment fraud cases.

Our Recommendation

Do not treat every unwanted SMS as a scam but never treat familiar branding, polite wording or a realistic-looking notification as proof of authenticity. Our 200-record investigation demonstrates why the distinction matters.

An unsolicited promotional message, an unwanted subscription notice and an attempt to steal banking credentials may share vocabulary while representing very different risks.

The safest habit is to identify the requested action, verify the sender through an independently obtained channel and refuse unexpected demands for sensitive information or payment.

The most useful question is not simply “Does this message look suspicious?”

It is “What does this sender want me to do, and how can I verify the claim without following their instructions?”

Verifications

ExAIS_SMS — Nigerian SMS Spam Dataset

Ajayi — Discursive-Manipulative Strategies in Scam Emails and SMS: The Nigerian Perspective

Central Bank of Nigeria — Fraud and Scam Awareness

Nigeria Electronic Fraud Forum — Safety Tips

Economic and Financial Crimes Commission — Red Alert on Scam

Nigeria Inter-Bank Settlement System — 2025 Digital Payment Fraud Findings