The Fake Reversal Trick: How Kenya’s Sellers Were Robbed in Plain Sight

For years, the small trader was the weak link in Kenya’s mobile money success story. A grocer in Kawangware, a boutique owner in Eldoret, a boda boda operator waiting for a fare, all of them shared one habit that fraudsters learned to exploit: they trusted the ping of an M-Pesa notification before they trusted their own eyes.
The scam was elegant in its simplicity. A “buyer” would approach a seller, agree on a price, and send a text message engineered to look exactly like a genuine M-Pesa payment confirmation, same font, same formatting, same sender name. The seller, busy or distracted, would hand over goods or change on the strength of that message alone. Minutes later, the “buyer” would call back, sounding apologetic, claiming they had sent too much money or paid the wrong till, and ask the seller to refund the difference via a real M-Pesa transaction. By the time the seller checked their actual balance, the truth was obvious: no money had ever landed. The confirmation was fabricated. The refund, however, was real money, gone for good.
Variants of this trick multiplied. Some fraudsters skipped the fake buyer altogether and simply pretended to be Safaricom staff, calling sellers directly and instructing them to “process a reversal” through USSD codes that actually authorised an outgoing transfer. Others convinced agents to hand over a customer’s phone under the guise of resolving a dispute, then took control of the line entirely.
What made these scams so damaging was psychological, not technical. Sellers, especially those handling dozens of small transactions a day, had been trained by years of legitimate M-Pesa use to trust a text message as proof of payment. Fraudsters simply weaponised that trust.
How Safaricom Fought Back
Safaricom’s response has been layered rather than a single silver bullet, and that is precisely why it has worked as well as it has.
First came public education at scale. Safaricom has repeatedly and explicitly warned customers that its customer care team will never ask customers to visit an agent for a reversal, and that reversals should only be initiated through official Safaricom channels. This single message directly undercuts the agent-hijack version of the scam.
Second, the company built a rapid reporting pipeline. Customers who receive a suspicious message are told to forward it to 333 so the sending number can be blocked before it defrauds anyone else, while genuine mis-sent payments should be resolved by <forwarding the transaction message to 456 or contacting Safaricom directly, rather than sending an independent “refund.”
Third, and most consequentially for sellers specifically, Safaricom introduced Hakikisha, a verification layer that displays the recipient’s name before a transaction is confirmed, closing the gap between what a screen claims and what has actually happened.
Fourth, the network leaned on infrastructure Safaricom already controlled: encrypted data transmission across the network makes it far harder for outsiders to spoof transaction data at the system level, even if the visible text message can still be faked.
Finally, Safaricom paired technology with enforcement, working alongside Kenya’s Directorate of Criminal Investigations and the Communications Authority on SIM-related fraud, and supporting the mass SIM re-registration exercise that took thousands of fraudulently registered lines out of circulation.
None of this makes the fake reversal trick extinct. Fraud, like water, finds the next weak point, new templates, new urgency scripts, new confidence tricks aimed at busy sellers. What Safaricom’s approach demonstrates, though, is that the most effective defence against a psychological scam isn’t purely technical. It’s teaching an entire economy of small sellers a single, simple discipline: never trust a text message, trust your own balance. Dial your own USSD code. Check your own app. That habit, more than any encryption standard, is what finally started closing the gap fraudsters had exploited for years.
The lesson for other mobile money markets watching Kenya is straightforward: the fix for a trust-based scam has to be a trust-based countermeasure — verify before you believe, and make verification as easy as the fraud itself.
Read Also: SIM Swap Fraud: How Scammers Hijack Your Line And How Safaricom Is Winning The Fight Back
About Soko Directory Team
Soko Directory is a Financial and Markets digital portal that tracks brands, listed firms on the NSE, SMEs and trend setters in the markets eco-system.Find us on Facebook: facebook.com/SokoDirectory and on Twitter: twitter.com/SokoDirectory
- January 2026 (220)
- February 2026 (248)
- March 2026 (287)
- April 2026 (208)
- May 2026 (191)
- June 2026 (236)
- July 2026 (71)
- January 2025 (119)
- February 2025 (191)
- March 2025 (212)
- April 2025 (193)
- May 2025 (161)
- June 2025 (157)
- July 2025 (227)
- August 2025 (211)
- September 2025 (270)
- October 2025 (297)
- November 2025 (230)
- December 2025 (220)
- January 2024 (238)
- February 2024 (227)
- March 2024 (190)
- April 2024 (133)
- May 2024 (157)
- June 2024 (145)
- July 2024 (136)
- August 2024 (154)
- September 2024 (212)
- October 2024 (255)
- November 2024 (196)
- December 2024 (143)
- January 2023 (182)
- February 2023 (203)
- March 2023 (322)
- April 2023 (297)
- May 2023 (267)
- June 2023 (214)
- July 2023 (212)
- August 2023 (257)
- September 2023 (237)
- October 2023 (264)
- November 2023 (286)
- December 2023 (177)
- January 2022 (293)
- February 2022 (329)
- March 2022 (358)
- April 2022 (292)
- May 2022 (271)
- June 2022 (232)
- July 2022 (278)
- August 2022 (253)
- September 2022 (246)
- October 2022 (196)
- November 2022 (232)
- December 2022 (167)
- January 2021 (182)
- February 2021 (227)
- March 2021 (325)
- April 2021 (259)
- May 2021 (285)
- June 2021 (272)
- July 2021 (277)
- August 2021 (232)
- September 2021 (271)
- October 2021 (304)
- November 2021 (364)
- December 2021 (249)
- January 2020 (272)
- February 2020 (310)
- March 2020 (390)
- April 2020 (321)
- May 2020 (335)
- June 2020 (327)
- July 2020 (333)
- August 2020 (276)
- September 2020 (214)
- October 2020 (233)
- November 2020 (242)
- December 2020 (187)
- January 2019 (251)
- February 2019 (215)
- March 2019 (283)
- April 2019 (254)
- May 2019 (269)
- June 2019 (249)
- July 2019 (335)
- August 2019 (292)
- September 2019 (306)
- October 2019 (313)
- November 2019 (362)
- December 2019 (318)
- January 2018 (291)
- February 2018 (213)
- March 2018 (275)
- April 2018 (223)
- May 2018 (235)
- June 2018 (176)
- July 2018 (256)
- August 2018 (247)
- September 2018 (255)
- October 2018 (282)
- November 2018 (282)
- December 2018 (184)
- January 2017 (183)
- February 2017 (194)
- March 2017 (207)
- April 2017 (104)
- May 2017 (169)
- June 2017 (205)
- July 2017 (189)
- August 2017 (195)
- September 2017 (186)
- October 2017 (235)
- November 2017 (253)
- December 2017 (266)
- January 2016 (164)
- February 2016 (165)
- March 2016 (189)
- April 2016 (143)
- May 2016 (245)
- June 2016 (182)
- July 2016 (271)
- August 2016 (247)
- September 2016 (233)
- October 2016 (191)
- November 2016 (243)
- December 2016 (153)
- January 2015 (1)
- February 2015 (4)
- March 2015 (164)
- April 2015 (107)
- May 2015 (116)
- June 2015 (119)
- July 2015 (145)
- August 2015 (157)
- September 2015 (186)
- October 2015 (169)
- November 2015 (173)
- December 2015 (205)
- March 2014 (2)
- March 2013 (10)
- June 2013 (1)
- March 2012 (7)
- April 2012 (15)
- May 2012 (1)
- July 2012 (1)
- August 2012 (4)
- October 2012 (2)
- November 2012 (2)
- December 2012 (1)
