Flashback to 2010. Naspers is still a majority shareholder in Mxit, at this stage Africa’s largest social network and instant messaging client with 27-million registered users. It is the darling of the South African technology space, proof that a digital African company can play on the international stage. Perhaps even more importantly, it is an inspiration to any number of mobile startups in South Africa and Africa as a whole. Five years later, it’s changed ownership, undergone wide-scale retrenchments, and its user numbers are believed to have fallen even lower than it’s willing to admit.
So where did it all go so wrong? And how did it happen so quickly?
As Memeburn’s investigations reveal, it has been a story of a massive drop and confusion around its user numbers over a span of ten years, intense international competition, and a smartphone revolution that appears to have left a company whose success was built largely on feature phones, behind the times.
Interrogating the numbers
At its peak, Mxit claimed to have 50-million registered users around the globe and 17-million active users in South Africa alone. By July 2014, the last time Mxit released any official user numbers, the latter number had plummeted to 4.9-million. Today there are suggestions that the service is shedding users faster than ever before and that monthly active user numbers have fallen below the one-million mark. It’s a pretty stark picture but, as anyone who’s watched the company closely over the years will attest, things are seldom that simple with Mxit user numbers.
Founded by Namibian-born mobile entrepreneur Herman Heunis in 2005, Mxit rode an explosion in South African mobile ownership and was especially popular among teenagers and young adults, for whom it provided a cost-effective means of communicating with friends and meeting strangers online. A little more than a year after launching, it claimed to have over a million registered users.
By January 2007, the service had tripled to 3-million registered users — enough to attract the attention of Naspers, which bought a 30% stake in the company. By early 2008, the numbers it was claiming had rocketed to 7.5-million. Two years later, Mxit had a claimed 15-million users and was big enough to be a source of controversy among the local media, sometimes unfairly so.
Before the year was out, the company would claim that it had 20-million users in 120 countries, sending an average of 250-million messages a day. The company, which had started out as a small team of eight people, now had well over a hundred staff and was expanding into exciting new territories like Mexico. By the end of 2010, Mxit was being hailed as Africa’s largest social network, with 27-million registered users. It was seen as a vital communication tool, with its users able to communicate seamlessly across other platforms like MSN messenger and Google Talk.
But things were changing rapidly. Early adopters were latching onto the latest smartphone models as soon as they could, and Steve Jobs had catapulted the world in the post-PC era with the launch of the iPad. Mxit still appeared to have a solid and loyal user base but it was a platform born on to Nokia-era feature phones that used the increasingly dated, and very slow J2Me development language, which was being left behind by a new smartphone wave. It wouldn’t be long before Nokia’s CEO Stephen Elop issued his famous, and brutally honest, “our platform is burning” statement to the world.
Alan Knott-Craig to the rescue?
Even so, when Alan-Knott Craig Jr acquired Mxit, buying out both Heunis and Naspers in September 2011, he brought with him the sense that he could turn things around. Knott-Craig, who also took over from Heunis as CEO, concluded the deal through his startup investment company World of Avatar which at the time had serious investor clout behind it. The company’s backers included the family trusts of two FirstRand founders: Paul Harris and GT Ferreira.
At the time of stepping down, Heunis claimed in a News24 article that Mxit now had an impressive “42- million users”, bigger in South Africa than both Facebook and Twitter combined. There was no mention of active user numbers.
In the year or so between Knott-Craig taking charge and stepping down, there were many, many headlines about the company’s bid to reinvent itself. There was brand refresh, the capital ‘X’ was dropped from Mxit’s name (plaguing sub-editors for years), and the company kicked off with an outlandish “Alice in Wonderland” party for the country’s digerati in the Cape Winelands. Few articles at the time paid attention to the user numbers, focusing on the excitement of a deal rumoured to be worth R500-million.
But perhaps we should have been paying attention.
In his first official presser as Mxit CEO in February 2012, Knott-Craig spoke about the platform’s now- 50-million registered users, eight million more than Heunis’ figure and nearly double the number of registered users it claimed 18 months earlier. He also said that the service had around 10-million active users.
Less well publicised was the fact those 10-million users, mostly South African, represented a sharp decline from the 17-million active users it laid claim to in 2010.
By September that year, active user numbers had fallen to around 9.35-million, a decline that went largely unnoticed by the country’s media.
An investor ousting
As the numbers declined, trouble was brewing inside the company. Knott-Craig was forced to step down as CEO in October 2012, apparently over disagreements with Paul Harris around how the company should be run. A lack of user growth was believed to be one of the factors at the centre of the ouster.
The banker had apparently threatened to pull his share of the R100-million investment being ploughed into Mxit if Knott-Craig did not step down. According to Business Times, “Harris was unhappy with Mr Knott-Craig’s performance and style and was only prepared to invest his share of R100m — which shareholders are ploughing into the company — if the CEO resigned”.
Publicly, the company continued to tell reporters that its monthly active user number was moving up and down between 9.3 million and 10 million. At the same time, an industry insider told Moneyweb the number was closer to five-million.
Late 2013 saw another substantial drop in Mxit’s official user numbers. By now it claimed to have just 7.4-million active users across Africa. Mxit suggested that it was because it had changed the way it reported its figures — in 2012, it reported users who were active over a three-month period (90 days), while in 2013 it limited the figures to the more standard one month.
At the same time, research put forward by World Wide Worx and Fuseware in their annual SA Social Media Landscape study suggested that Mxit’s South African user base was holding steady. In July 2012, a 30-day active measure gave Mxit 6.2-million subscribers. A year later, in July 2013, the active subscriber base had grown marginally to 6.3-million. In August, this number had grown to 6.5-million.
Straightening out the stats
In the report, the research companies also noted the efforts of new Mxit CEO Francois Swart – who had taken over the helm in 2013 — to clean up the way the company reported user numbers. But that would have been scant comfort when just a few months earlier, Swart had said that Mxit was aiming to double its user base through 2013 — this despite having retrenched nearly a third of its workforce.
Nonetheless, Swart continued in his efforts to ensure that the numbers reported by Mxit were as credible as possible. It was under his guidance that the company joined the DMMA (now IAB SA) and brought its advertising formats in line with industry standards.
To be fair to both Swart and his predecessors, Mxit is not the only company to have had its user numbers called into question. For years, Asian messaging apps sold success stories based on downloads and registered user numbers rather than active user numbers. Things got so bad that TechinAsia, which covers the region’s tech space, placed a ban on any stories built around download numbers as the central premise. This was, by the way, in 2013. So Mxit was actually ahead of the curve: first in making its user numbers understandable and secondly in bringing them in line with industry standards.
Former FNB CEO joins Mxit and international expansion
By this stage, the company had also brought on Michael Jordaan — rockstar banker turned venture capitalist — as the chairman of its board. While Jordaan — who has close ties to Harris and was CEO of FNB for more than a decade — said that he was broadly happy with Swart’s vision for the company, his appointment only added to the sense that Mxit was prepared to go to extreme lengths to turn its flagging fortunes around.
Since then, the company has launched new versions of its app across all the major smart — and feature — phone platforms and also opened an office in India. None of these interventions seem to have helped.
In mid 2014, the last time Mxit released any official user numbers, it was down to 4.9-million monthly active users in South Africa (Swart said removing porn and profanity was partially to blame for the decline).
Read more: Hey MXit, where’s the share button?
Mxit 7, the new version it had spent so much time and effort building and which was supposed to seamlessly bridge its smart and feature phone offerings, had reportedly been downloaded by 7.4-million users but only 55% of those were still active. In late 2014 meanwhile, it undertook a fresh round of retrenchments, with 45 staff — including some in key positions — offered packages.
There are claims that, in the intervening months, things have only gotten worse. A recent MyBroadband article cites a “well-placed industry” source who said that Mxit now has fewer than one-million active users and is losing 10% of its user base every month.
One possible reason for the decline was the promotion introduced by network operator Cell C in 2014 to make instant messaging service WhatsApp free to its users.
Memeburn approached Mxit spokesperson Ben-Carl Havemann to comment on the company’s decline in fortunes, and was referred to the company’s prior press releases. Speaking to the falling user numbers, Havemann referred us to the 2012 change in the way Mxit reported its numbers and the release of Mxit 7, which included new verification technology. “This, in conjunction with a stronger focus on moderation of user profiles and content, contributed to the decline”, he told us.
Havemann said that the Cell C Whatsapp promotion “had not seen any direct impact on Mxit’s active users”, but that it is a situation the company was monitoring closely.
He also did not disclose Mxit’s Indian user base numbers, explaining that they were “currently not monetising in this market, and therefore don’t disclose user data”. He promised an update early this year.
A company in trouble
In addition to the company’s dramatic numbers collapse, Mxit’s two significant retrenchment rounds within the space of a couple of years, suggest this is a company fighting for its survival. In addition, the high-profile hires made by Knott-Craig, including VP of developer relations Andy Volk (who had been brought across from the US); mobile specialist Peter Matthaei; and chief product officer Vincent Maher (who actually came on board following the acquisition of Cape Town-based startup Motribe), have left the company.
So where did it all go so wrong?
When Memeburn publisher Matthew Buckland and former managing editor Michelle Atagana wrote their analysis chronicling Knott-Craig’s sudden rise and fall, they noted the threats the service faced from the likes of iMessage, WhatsApp, and even BBM. Mxit’s old J2me platform, they wrote, was burning: “It’s an old technology suitable for cheap Nokia phones, and we know what Nokia thinks of its own platform“.
It’s worth bearing in mind that Knott-Craig inherited the platform from Heunis’ Mxit, as well as the fiercely loyal community that came with it. Mess with one, the prevailing logic went, and you could very well lose the other. One of Mxit’s greatest assets had, in other words, become a liability preventing it from moving forward. Even in 2011, when the company was probably on shaky ground. Knott-Craig admitted as much in early 2012. At a press conference the unassuming internet entrepreneur told the world that Mxit had “one last chance” to get back on track or the company would collapse.
“We have this year to get it right. If we don’t crack that, then I think we’re dead. If we get it right, though, we are on the right ticket. If I’m talking to you this time next year, that means we’re still alive and I haven’t been forced to emigrate because I won’t be allowed back in the country,” he said.
His five point plan to put the company back on track was deceptively simple:
Knott-Craig would have barely nine months to lay down the markers to achieving those goals. In October that year, as we’ve already noted he was ousted as CEO of Mxit and World of Avatar, the startup investment group he had birthed in 2010.
Even if Knott-Craig had been given the time he felt he needed to achieve his vision, attempts to keep the service’s core user base onside may have thwarted his efforts.
As it is, the developers behind the new smartphone versions of Mxit felt compelled to include complex features which were familiar to its feature phone users, rather than paring down the service and taking it in a single, well-defined direction.
As a result, smartphones still only make up around 15% of Mxit’s user base.
In South African investor circles, the saying “you never lose with Koos” (in reference to former Naspers CEO Koos Bekker) has become fairly commonplace. Under his leadership, Naspers grew from a South African-focused media company to a US$45-billion emerging markets internet and media giant. Perhaps we should have been a little more wary of the hype when Knott-Craig made his takeover bid.
By Bekker’s own admission, Naspers is a “very bad seller”, meaning it has has never been keen on selling off the companies it’s invested in. Naspers’ willingness to dispose of its Mxit shares when it did probably speaks volumes.
The nature of the beast
Without the burden of Mxit ownership, the way was clear for Naspers to bring WeChat to South Africa. Owned by Tencent, which Naspers has a major stake in, WeChat (known as Weixin in its native China) was developed for smartphones from the get go. It’s pretty telling that even a company as big as Tencent felt the need to reboot its efforts with the service instead of trying to morph its older platforms onto smartphones.
The truth is, very few feature phone messaging clients like Mxit have successfully transitioned into the smartphone space. The odds, it seems, were stacked against Mxit even as Knott-Craig was claiming that the company was “pretty good value and if it’s one-tenth of the things we think we’re going to come off, then we should be smiling”.
Things move quickly in the technology space, and even more so when it comes to mobile. One day Nokia is the biggest mobile phone manufacturer in the world, then it is wiped out. Just look at how BlackBerry has struggled to keep BBM, which dominated the early smartphone messaging space, relevant in the face of stiff competition from the likes of WhatsApp, Facebook Messenger, and WeChat.
How much greater then was the challenge that Mxit faced in 2011, and how much bigger is that challenge today? Think about the amount of time and money needed by major players in the smartphone messaging space to make a real, substantive mark. Would Mxit’s investors have been able to take the revenue hit necessary to reinvent and take on those players?
The fact that it hasn’t managed a resurgence, even with an investor who has as much clout as Harris and a chairman who commands as much respect as Jordaan, suggests not.
Could things have happened differently?
That’s not to say that Mxit hasn’t done the best it can on the path it’s chosen. Mxit 7 is testament to that. As Memeburn columnist Hilton Tarrant wrote shortly after its November launch, “Mxit 7 on iOS is gorgeous. Its reimagined J2ME client is perhaps even more impressive”.
And it is. It’s impossible to open up Mxit 7 on any of the devices it’s available on and not see that its had a massive amount of effort put into it. But, as Tarrant also wrote: “a gorgeous product does not a successful product make…”
Think about all the features crammed into Mxit 7 and compare them with the minimal innovation introduced by WhatsApp, the world’s most popular instant messaging service. Could Mxit have learned any lessons from WhatsApp’s success?
In order to even begin answering that question, you have to break down the Mxit product in the simplest terms possible. At its heart, Mxit is a hybrid of three aspects of social networking:
Suffice to say, blending all three of these facets together is far from simple. That’s why services such as WeChat, Line and KakaoTalk have gone for options one and two, while WhatsApp and Facebook Messenger have stuck with option one. There are few, if any, messaging-based social networks that have managed all three.
Bearing in mind how crowded the space around the first two aspects is, perhaps Mxit should have focused on the third aspect: anonymous public chatrooms.
Matthei, whose job at least partially entailed thinking about this kind of stuff, certainly appears to think so. In a recent appearance on the #o42t podcast, he said that thing that he always believed made Mxit special was the chatrooms.
Anonymous chatrooms are, of course, a far more difficult sell than private messaging and open platforms. The potential market size for a mobile service offering anonymous chatrooms is also probably much smaller than one allowing people to connect with family, friends and brands.
That said, Mxit probably stood a better chance than most of making that kind of platform work in the new smartphone space. It had managed it in the feature phone space. For many people, the chatrooms “were” Mxit.
Thing is, that’s not the vision that the World of Avatar shareholders bought into in 2011, and it’s not really even the product that Mxit was any more at that point. By now, it’s probably too late to change course, especially with so much expertise having left the company.
Days of future past
If Mxit can’t look to its past for inspiration, it beggars the question: where to next?
That’s a difficult question to answer without looking like a side show psychic. But barring a user explosion in India (something it surely would’ve told us about if it was happening) or immediate and widespread of adoption of whatever it launches in the near future, things don’t look all that promising.
One area where Mxit has seen recent success is with its Mxit Reach offering, which is centered around using Mxit technology to create free mobile educational, health care, agricultural and community applications.
There’s been evidence that the company would be shifting its focus to reach for some time now. When Jordaan came onboard as chairman, he didn’t seem interested in Mxit’s ability to make money so much as the company’s ability to help people. “One specific aspect of Mxit that I believe is unique is how Mxit is leveraging their technology to help people, particularly in the fields of mobile education and health care,” he told Memeburn at the time.
It’s also made a number of announcements regarding Reach in the past couple of years, the most recent being the launch of Ukufunda, its virtual school offering.
And in the wake of the 2014 retrenchments, Mxit said it would “place a bigger emphasis on its social services”.
It claims that around 1.2-million people a month use the various Reach apps and that it has seen 20% year-on-year growth.
It’s difficult however to imagine it bringing in serious revenue for Mxit. It was never meant to. As worthy as the stuff being done by the people building Mxit Reach apps is, it also seems unlikely that it’ll bring a mass of new users to the platform.
It also seems inevitable that even the most loyal Reach users will upgrade their devices (smartphones are now easily within reach, with a base model costing as little as US$50). And once they do, retaining them becomes that much harder, just as it has with the social network.
Whatever happens, it’s almost certain that there are a number of people who were involved with the company — both in its early high-growth days under Heunis and during Knott-Craig’s brief tenure in charge — who will be left wondering what might have been.