Mobile operators have an SMS addiction

The margins on SMS are eye-watering, and the operators are hooked. Perhaps this explains Vodacom and MTN's bizarre statements and sabre-rattling about "over-the-top" services. By Hilton Tarrant.

hilton-tarrant-180Let’s start at a common point of departure: the mere notion of mobile operators hoping that so-called “over-the-top” (OTT) services be regulated is insanity. One can understand how an operator and its executives can think this rational, though. After all, an operator only knows how to be an operator. This explains their endless failed attempts to be anything but an operator. Think media businesses (remember Yebo Radio?), financial services businesses, music download businesses. There are exceptions that prove the rule, but the cold hard truth is that none of this is in their DNA.

From the perspective of an operator, these “pesky” OTT services mean a loss of revenue. Average revenue per user — the industry measure for how much revenue each subscriber generates in a month — has been declining steadily for years. Mobile voice tariffs have dropped sharply (thanks to hefty cuts mandated by communications regulator Icasa). Data costs have been falling, too, and messaging apps (“OTT services”) have practically supplanted SMS. Again, it’s logical that we’re seeing this response, especially from the two largest operators in South Africa: bizarre statements, sabre-rattling and hints of lobbying.

Never mind that it’s illogical, impractical, unfeasible and — quite frankly — impossible to implement. How do you regulate Skype and WhatsApp but not WeChat, for argument’s sake? Or Telegram? Is Apple’s iMessage an OTT provider? Facebook Messenger? Instagram, because it has messaging functionality? Twitter? Where do you draw the line and how do you try keep up with progress? Just look at how governments and regulators are battling to regulate Uber. And that’s a fairly definable, distinct service.

But I digress. SMS pricing remains stubbornly high. After the cratering of voice tariffs and natural decline in data costs, it’s effectively one of only two levers the operators have left (the other is the monthly base contract tariff, and you’ll recall those increased practically across the board last year).

Even Cell C, with all of its recent spin around OTT services, charges 50c/SMS on prepaid plans and contracts

On Vodacom, prepaid (and contract) SMS pricing largely remains at the 50c mark (on a few plans, it’s 80c in peak periods and 35c in off-peak). MTN also charges 50c/SMS on most of its plans (some are 75c). On contracts, it’s the same (on some, there are bundled/free messages included). Even Cell C, with all of its recent spin around OTT services, charges 50c/SMS on prepaid plans and contracts.

The margins on SMS are eye-watering. It’s a drug. And the operators are hooked.

What these stubbornly-high SMS prices have done is create a distorted market. Airtime is effectively cheaper than SMS. On many (most?) tariff plans, you’ll be able to have at least a 30-second call for the price of a text message. And don’t even attempt to compare the 50c to the underlying cost of data to send the same message. The distortions extend to post-paid plans, too, where consumers are force-fed practically worthless — and endless — SMS bundles. Who, in 2016, uses 200 SMSes a month, not to mention limitless SMSes, which operators seem to believe is a compelling value proposition.

SMS has become the ultimate digital grudge purchase.

Yet, it is still a big money maker for operators, given its high margins. In the six months to 30 September 2015, Vodacom generated almost R1,3bn in “messaging revenue” in South Africa. To put that in perspective, it’s “only” 5% of Vodacom’s total service revenue in South Africa. SMS volumes have cratered — they’ve halved in the last two years — but somehow revenue inched up when compared to the same period in 2014. This means the average price per SMS is actually increasing!


Look at that jump in the average price per SMS between 2014 and 2015 … it’s 23%! No wonder operators are bleating about OTT providers.

It’s a vicious circle. Not only are services like WhatsApp and iMessage completely entrenched in user behaviour, consumers have become conditioned to know that SMSes are expensive (relative to other options), so we avoid using them, whether consciously or not. The latter likely helped cement the former.

One other oft-overlooked dynamic in the SMS market is that a good chunk of this messaging volume is made up of bulk SMSes (those sent by banks, insurers, retailers, call centres and security companies). Here, pricing is discounted, but not by as much as you think.

At the 100 000 messages a month level, big messaging providers are still charging in the region of 20-25c per message. Some of this will be their margin, but that’s likely only the differential between the rate they’re paying for their consolidated number of messages across all client, and these retail prices. There’s not much discounting here from the operators. Vodacom charges anywhere between 24c and 29c in its bulk tariff plans.

What if…
But what if SMS was (near) free? What if there was as a steady a decline in the price of SMSes as there’s been in the price of voice calls? If 50c became 40c became 30c became 20c became 10c (or 5c)? Would we have avoided sending SMSes at these prices? Would there have been as much friction? Probably not.

But any increase or stabilisation in volumes would likely not offset the decline in prices to achieve the same revenue outcome. This is why operators have reacted in the way they have over the last few years. To try keep SMS pricing as high as possible to soften the decline in revenue (and to incredibly somehow increase the average price per message) has been rational and predictable. I’m sure the global management and telecommunications consultants have run countless models for the operators to vindicate this deliberate decision.


But there comes a point where the base and entire use-case is eroded to such a point that the stubborn defence of this pricing (and pricing model) no longer makes sense and it collapses. It’s called disruption (an over-used word, which I hate). The announcement by WhatsApp this month that it would allow business to use its platform to message customers will mark that inflection point, particularly in South Africa. Will banks and retailers still pay a toll of 15c, 20c or 25c per SMS message when WhatsApp offers a similar way to reach the same user at a fraction of that price? (We’ve already seen brands and businesses offload their SMS messaging to WeChat in other markets, especially China.)

Back to that tipping point for operators… We’re probably past it already.

  • Hilton Tarrant works at immedia. He owns shares in Vodacom, first purchased in June 2013
  • This piece was first published on Moneyweb and is used here with permission

Share this article

  • Greg Mahlknecht

    Overpriced though they may be, there’s still an important place in the world for SMS – it’s the only messaging mechanism you can send to someone and be 100% guaranteed it’ll either reach them of you’ll be notified quickly and given a reason if it doesn’t.

    WeChat might have an API, but it’s more of the “send it, and it might reach there some time” with no kind of guarantee or way of telling if the user got it.

    WhatsApp has a more reliable delivery mechanism, and status, but no API, so that’s off the table too. It can be a little flakey at times, mis-reporting delivery/read status, but I’ve only used an unofficial API with it, so can’t make a proper judgement here.

  • Vusumuzi Sibiya

    >>especially from the two largest operators in South Africa: bizarre statements, sabre-rattling and hints of lobbying.

    Yep! >>hints of lobbying.

    …enough said!!

  • Mouton

    In 2003 I did cost models for a local operator – one case specifically to define unit cost for sms and at the time gprs. The cost of sms was about 4c with tarif then between R1 and 80c

  • V.O.I.C.E

    This is all about not being able to make more money & greed rather than “losing money”!

  • aknit

    I wouldn’t be too sure.

    Instances abound where SMS’s sent did not reach their intended recipients. I had to re-send them. Other cases, delayed deliveries were common, some as much as three or four days. And in some cases, you cannot receive messages in certain countries due to “dodgy” roaming agreements between the mobile networks.

    This is where WhatsApp provides slightly better reliability and reporting. I guess the API will come soon.


  • Greg Mahlknecht

    No, WhatsApp doesn’t provide better reliability and reporting at all. One can take wild guesses as to what will be in the upcoming, but stands, it certainly does not.

    We send SMSes for many mission-critical and life-and-death scenarios, and I know for a fact how reliable it is; and if it doesn’t reach there, 100% of the time one can follow up with the MNO and find out the reason. In my decade+ doing this, the reason was never “we don’t know what happened to it”, even with the technically inferior MNOs like CellC.

    Some countries, especially our African neighbors, have terrible MNO infrastructure and their systems go down and are offline for hours or more, but that’s not a failing of the SMS, it’s a failing of the implementation.

  • aknit

    If you are speaking about reporting at the infrastructure level, yes, the MNO gets to see what happens.

    I’m speaking about reporting at the user level. Unlike WhatsApp, you are not always guaranteed of a delivery report when an SMS is sent. Some networks support it, some don’t. Some claim to support it, it works half the time or not at all.

    If you are talking about reporting within WhatsApp’s infrastructure, well, you don’t know how that works anymore than I do. I’m more concerned with user-level reporting, although I am reasonably confident WhatsApp have lots of data about the messages their network handles (as do Apple with iMessage).

    Businesses use SMS to send important messages about their service because it is easier to plug into the local service provider than it is with an OTT. Moreover, SMS was there long before OTT-based messaging, so those relationships, infrastructure and protocols are well-established. However, I would not rule out the possibility of OTT-based messaging making a real play in that space. I mean, a lot of IP service providers now use Twitter as a primary means of network status communication, for better or worse.


  • Greg Mahlknecht

    >Businesses use SMS to send important messages about their service because it is easier to plug into the local service provider than it is with an OTT.

    Not really, at the end of the day you’ll just be using a REST service. I make a living from this (SMS, WhatsApp, WeChat, etc – not just SMS) and have coded the platform we use, so have deep knowledge here. Business uses SMS because if you have someone’s number you can get a message to them. With WhatsApp/Wechat, they have to have the app installed. And have added you as a contact. And have an active data connection.

    I 100% agree that IM (let’s just drop the OTT term, it’s playing in to the MNO’s FUD 🙂 ) services will make inroads here, but they’ve been talking about it happening for 10+ years and it hasn’t made any appreciable impact in business-to-consumer SMS, in fact that section of SMS usage is still steadily growing. When WhatsApp releases their API (it appears to be in a limited release with select businesses now) it will be the moment of truth.

  • aknit

    I do not disagree, but you left out the part where I also said:

    “Moreover, SMS was there long before OTT-based messaging, so those relationships, infrastructure and protocols are well-established.”

    So yes, its resiliency of being embedded within the fundamental design of mobile phone communications is a real convenience to businesses that use it for this purpose. Needless to say, as you point out, the fact that customer numbers are fair game without prior permission.

    While SMS works, I do not agree that it is 100% reliable, as I’ve mentioned already, particularly from a user standpoint (as a developer using it to deliver services, you get to see more than the user does, which is good). For example, living in South Africa on a Vodacom service, even though I can roam for voice and data while traveling in the U.S., I cannot receive SMS messages, on all U.S. GSM networks. This is a known-issue which Vodacom customers just have to deal with – imagine trying to bank while there. When I leave the U.S., I start to receive some of the messages that were sent to me, others never make it. This does not seem to affect MTN, with whom I also have a service.

    IM, which is delivered over IP, could be more superior here, if they simply worked out the convenience for businesses as they have for users.


  • Greg Mahlknecht

    >Moreover, SMS was there long before OTT-based messaging, so those relationships, infrastructure and protocols are well-established

    Well, isn’t this the real problem – the IM’s are so busy competing with each other they go to great lengths to ensure there aren’t relationships and protocols which allow a universal competitor to SMS. For a while it looked like XMPP might be that protocol – 5-7 years ago we were far closer to having an SMS competitor than today, but we’ve gone backwards a long way since then in that regard.

    > While SMS works, I do not agree that it is 100% reliable

    I didn’t say 100% reliable, I said if 100% “reliable or get a reason why not” – in this specific case, which I am familiar with, we get back a status saying the SMS won’t be relayed to the handset, and we use an alternative (more expensive) route which will support it. IMs have no solution in this respect, if someone was roaming and didn’t have data, apart from maybe falling back to SMS.

    We do SMSes for one of the country’s major banks and our customers have no problem receiving their SMS anywhere in the world, because we implement what’s necessary to have it 100% reliable – if 1 of a million messages “go missing”, we have to (and can) account for it. SMS really is that reliable when used correctly.

    >if they simply worked out the convenience for businesses as they have for users.
    100% agreed – WeChat has the API, but it’s “best effort” – not good enough for important businesses.
    Don’t mistake a cheap and nasty SMS provider for a shortcoming of SMS. A cheap and nasty IM provider will have exactly the same problems when that starts to catch on.