Twenty-five years ago, four South Africans wired the internet to a mobile phone using four lines of code. Today their company runs the back end of conversations between banks, airlines, retailers and roughly a third of the planet's mobile users.
A woman in Lagos sends a WhatsApp message to her bank at 11:47pm. By 11:47pm and twenty seconds, her loan application is approved, the funds are in her account, and the entire transaction - identity check, signature, disbursement - has happened inside a chat thread. She has not opened an app. She has not phoned anyone. She has, by any honest measure, just been served by Clickatell.
Most people will never type the word "Clickatell" into a search bar. That is, in a sense, the point. The Redwood City-headquartered, Cape Town-grown company runs in the plumbing of mobile commerce, somewhere between the SMS your dentist sent you and the WhatsApp confirmation from the airline. It is, depending on who you ask, a messaging gateway, a chat commerce platform, a payments rail, or - lately - a small army of AI agents pretending to be a single very patient employee.
In the year 2000, the cellphone was a brick and the internet was a noise. The two did not speak. If your bank wanted to tell you the rent had cleared, it had to send a letter. If your courier wanted to confirm a delivery, it had to ring you, which often it did not, because you were nine time zones away.
Four young South Africans noticed something that, in retrospect, looks obvious and at the time looked vaguely impossible: a customer's mobile phone was the single most reliable thing about them. More reliable than email. More reliable than home address. More reliable, certainly, than the phone in the office.
So they wrote four lines of code. The code did one trick. It let a server somewhere on the internet send a text message to a phone somewhere on a mobile network. This was, at the time, a small miracle dressed as a utility.
The bet was unfashionable. The dotcom crash had just landed on top of the world's optimism like a piano. African tech startups were a category that mostly existed in PowerPoint decks. And the de Villiers brothers were proposing a B2B messaging gateway run out of the Western Cape, sold to whoever had a phone number and a problem.
What they bet on was that mobile networks would not consolidate. There would be hundreds of them, in dozens of countries, with no incentive to play nicely with each other. Somebody had to be the polite middleman. Somebody had to learn the local SS7 quirks of the Vodacom in Johannesburg and the MTN in Kampala and the AT&T in Texas. Clickatell volunteered.
By the late 2000s the company had quietly built itself a network of relationships with hundreds of carriers. By 2022 the number was 960. The list of countries reachable through the platform passed 220 - which is, helpfully, more countries than the United Nations admits exist.
For its first decade, Clickatell was, in plain English, an SMS gateway. You wrote a couple of lines against the API and your application could text any phone on earth that had a signal. The product was boring in the way that running water is boring: deeply useful, easily ignored.
The interesting move came later, when WhatsApp ate the world. Clickatell did not chase the channel; it absorbed it. Today the Chat Commerce platform routes the same business conversation through SMS, WhatsApp, USSD, web chat - whichever channel the customer happens to be inside.
Send a one-time password. Confirm a flight. Run a marketing broadcast. Take a payment. Approve a loan. Onboard a customer with a KYC flow that respects the regulator and the user's patience in roughly equal measure. The platform is, increasingly, less a messaging product and more a place where transactions live.
Twenty-five years compressed into seven cards. Each one was, at the time, an opinion most people disagreed with.
Founded in Cape Town by the de Villiers brothers, Du Toit and Lawson. The first commercial bridge between web and mobile.
Silicon Valley notices a South African messaging company. Sequoia Capital invests. The HQ slowly drifts west.
Carrier connectivity quietly becomes a moat. Hundreds of telco integrations, none of them glamorous.
The company stops calling itself SMS and starts calling itself chat commerce. The market eventually agrees.
The first payment of its kind. The press release is read by approximately twelve people. The product is read by millions.
Arrowroot leads. Kennedy Lewis, Endeavor, Harvest follow. Oversubscribed. US expansion accelerates.
Twenty-fifth birthday. The new bet: AI agents that finish the task inside the chat thread, supervised by a human who occasionally agrees.
A messaging company is, in the end, judged by reach and revenue. Clickatell's reach is unreasonable for an African-founded firm. Its revenue is healthy enough that it stopped needing to raise.
Sources: Clickatell press releases, TechCrunch, Crunchbase, company filings. Bars scaled for legibility, not for investor decks.
The customer roster reads like an emerging-market commerce census: tier-one African banks, global airlines, insurers, telcos, retailers. The investor list is similarly unsurprising for a company in its mid-twenties - Sequoia from the early days, Arrowroot, Kennedy Lewis and Endeavor in the last round.
Make commerce conversational. Not chat about commerce - chat as commerce. The browser tab is, after twenty years of A/B testing, a slightly tired sales channel. The chat thread is a fresh one. People type their problems in plain language. Clickatell's wager is that they should be able to finish their problems there too.
Pieter de Villiers - who, against all CEO archetypes, lives in Stellenbosch and not in Atherton - has spent two decades repeating a version of this idea. The repetition has been the point. It is hard to remember now, but in the early 2010s it was a contrarian opinion. By 2025 it is a Visa partnership.
The new noun is "agentic commerce". The idea is that the chat thread stops being a channel and becomes a small office. You text the airline. An AI agent reads the thread, books the new flight, processes the change fee, pings a human supervisor when the rules get strange, and confirms the seat. The whole experience never leaves the bubble.
Clickatell's pitch is that they have spent twenty-five years building the unromantic parts that make this work. The carrier integrations. The KYC plumbing. The payment tokenization. The compliance logs. The agent layer sits on top. The agent layer, importantly, would not work without the layer underneath - which is the part nobody wants to build twice.
Whether the bet pays off again depends on whether customers prefer typing to tapping. The early evidence in emerging markets - where data is expensive, apps are scarce, and WhatsApp is the operating system - is unambiguous.
The woman in Lagos closes WhatsApp. The loan has cleared. Her landlord is paid. She has not, in any visible way, encountered a company - only a conversation that happened to work. Behind that conversation sits a carrier integration that took two years to negotiate, a payment token that took a Visa lawyer six months to bless, a chat banking flow that satisfied a regulator nobody had heard of, and a small AI agent supervised by a human in another time zone.
All of it bills back, in fractions of a cent, to a quarter-century-old South African company that started with four lines of code.
The founders are, like most South African founders, more comfortable explaining the company than promoting it. The good news is the explanations are unusually clear.
Pieter de Villiers on Stellenbosch, scale, and the long way to Silicon Valley.
Product demosWalkthroughs of Chat Commerce, Chat-2-Pay and the SMS API.
CMO interviewPieter on why chat commerce is the next CRM layer, not a channel.
Press releaseThe 2025 anniversary release where agentic commerce gets named.
Official channels, founder profiles, the funding paper trail. Verify everything.