Bonanza or bust for termination fees?
One occasionally controversial, but always lucrative, part of the telecoms business is the collection of termination fees. For example, the UK regulator Ofcom estimates that approx. 15% of UK mobile industry revenue is via termination (in other words £2bn of a £14bn industry in 2006). What might happen to this voice and SMS wholesale revenue base as we move towards more open 'Telco 2.0'-like models? We've got a surprising answer for you, based on two-sided markets.
In the very early days of the telephone network there were multiple competing access networks, which did not interconnect. You could have half a dozen lines coming into your property. This was clearly a non-scalable solution, since to be able to reach everyone you needed to subscribe to every network. This soon led to interconnected networks, with users choosing a single access provider, and fees levied to carry a call off-network. This naturally favours particular market structures, with dominant players receiving considerably more than they pay out, and with little or no price competition on termination fees. After all, if someone wants to reach you, what choice do they have except to go through the retail carrier the call recipient has chosen?
As a result, regulatory intervention has been commonplace to attempt to tie termination fees to actual costs. In a capex-intensive business, this creates a lot of creative work for lawyers and accountants attempting to allocate those costs. Traditionally telcos have done well out of this system, particularly GSM operators with calling party pays, but the pendulum might be about to swing the other way. Indeed, there are lots of arguments over whether this system is a good to a bad thing, or whether there should be a 'bill and keep' alternative without any termination fees. We're agnostic, as we just have to deal with the world as we find it.
These fees are a kind of degenerate two-sided market. Like a full-blown two-sided market (e.g. Google Adwords), you attract an audience and then charge someone to access that audience. Those wishing to reach your subscribers potentially subsidise the acquisition of that audience. For example, we saw this in the past with free calling 'on island' in many territories in the Caribbean, but with very high termination rates for outsiders wishing to call in.
Not really a two-sided market when a telco is on both sides
However, it isn't quite a real two-sided market. One of the attendees at our last Telco 2.0 Executive Brainstorm noted in the feedback:
I don't understand the comment that c.99% of [core telco] revenue [today] is retail -- 20% of revenue and 30% of EBITDA is already from wholesale termination receipts.
The problem is that those termination fees turn into retail call charges by other telcos. So the main effect is to damp down usage, not create value. This can be seen in the large gap between US and European average minutes used (but not ARPUs), due to differing termination regimes. The flip side is that Europe enjoys increased penetration (since you can subsidise receiving calls and it's worth having a handset even if you never make a call).
You're not drawing in new revenue from outside of telecom -- unless you've a single dominant telco able to hold the user base 'hostage', or a perimeter fence to shake down rich expats. So it's not really a two-sided market according to the strict definition, as you're not tying two distinct groups together who wish to transact, just individuals; and providing just a conduit, not transactions. Nice for mobile operators, nasty for fixed ones, but still just stirring the revenue pot around rather than adding to it.
There are some existing two-sided revenue streams inside telecoms, such as freephone numbers, and premium SMS. The merchant is paying the telco as a customer for providing a service (access and billing respectively) in the context of a retail relationship the telco has with the user.
So much for the history lesson. Now for the meat.
Where's the value in telephony?
Go back and check out the pyramid in this old blog post of ours. Telephony is sandwiched between two other processes:
- Rendezvous, which is the selection of the right timing, location, participants and medium of an interaction such as a call.
- Conversation, which is the process by which multiple interactions and linked together -- for example, when a merchant promises to call you back to tell you if the goods are in stock.
The margins in telephony are declining, but we can create new products and processes that incorporate these elements. Specifically, we draw upon the Communications Enabled Business Process (CEBP) space. So an example is today my utility company sends me a text message asking me to read my electricity meter and send the answer back by SMS. Problem for utility company: I'm abroad, not at home, asleep, in a call, or otherwise unwilling or unable to do so. Problem for telco: There's precious little termination revenue in a bulk SMS.
Job of the telco: help people 'get through'
So, what can we do? Well, the telco offers an API that will only forward the message to the user when they get home. That could be using location, or other presence data (e.g. dual mode phone re-associates with home hub). Even better, the message is never forwarded in the middle of a call, or when I'm abroad. And it's presonalised too. The telco knows what time of day I'm typically active. After all, I might be a shift worker. So the telco gets rewarded not for delivering SMS messages, but for getting meter readings in return.
The cost of someone coming to my door is ten dollars or more; that of a wholesale SMS message a cent or two. In the middle is a huge amount of margin to be made. And guess what? The only agent able to assemble the data to make the rendezvous or conversation process work well is ... the telco.
So you turn a 1 cent wholesale SMS into a 1 dollar 2-sided market platform service. It's not price capped, as it's an information service, not a regulated telco service. And you can use the revenue to subsidise the retail side to gain audience and market share.
Need to build the right rich wholesale products
The over-arching pattern here is that (i) someone wants to get through to someone else, (ii) that access needs to be timed, presented and targeted correctly, and (iii) the person who receives the message is enabled and encouraged to act as a result. For example, adverts are part of a two-sided market, and also follow this pattern.
What operators need to do is look afresh at their portfolio of products and think how they could and should operate in a two-sided market. Take voicemail, for instance. Today a call centre that wishes to contact you has a single choice: to make your phone ring. This will gather standard per-minute termination fees. The problem for the call centre is that sometimes they may not wish to talk to an individual, as it drives up call time with social chatter. The resources for outbound calling also may not be at times that make sense to the recipient. Furthermore, a significant number of calls will 'fail', in that the callee will answer, but will effectively be unable to talk ('call me back later').
What if, instead, the telco simply offers an API to deposit voice messages into your voicemail? This can be charged at a higher rate than the equivalent per-minute fee. Then you let your imagination run riot in designing this wholesale access product:
- What if the telco held back the user notification until an appropriate time? (So if you're roaming abroad, no 3am 'beep beep').
- What if the telco let the sender supersede a message, so there's no confusion if there was a change of status (great for airlines keeping customers up to date with schedule irregularities)?
- What if it wasn't just an audio file being played, but the merchant could pass on, say, a Voice XML document with an embedded IVR?
- What if indeed the user, when they play a message, is really connected live to the merchant's IVR? "The product is now in stock, press one to confirm your order."
As you can see, once you've assembled the consumer audience, the value you can extract is only as good as the range of interactions you can offer to merchants. By making these APIs blend customer data, historical behaviour and network access they can be very sticky, which translates to high margins. That is why we think telcos looking at the advertising industry are focused way too narrowly. The opportunity is in enabling a far greater range of interactions across many more business processes than just marketing.
Telco brand as trustmark
Facebook is a site that epitomises the Internet approach. It offers rich functionality, an open API, and is teeming with innovation. No telco service is ever likely to compete with the state of the art of Internet services in terms of functionality. On the other hand, with its beacon privacy disaster, lack of payment mechanism, and problems with abuse and spam, it also reflects the worst of Internet commercial culture. Or consider the massive outage at Twitter, which is fairly typical of start-up growth pains.
The telco strength is in the 'non-functional' parts of the puzzle: scalability, availability, support, billing, provisioning, etc. In particular, the role of the telco brand in these new markets is that of a trust mark, much like that of the VISA as you go into a store. You don't hesitate to enter your PIN into the terminal, because you believe you are safe (rightly or wrongly).
The telco doesn't need to compete with all the latest innovations in social media and personal communications to be successful in 2-sided markets. The telco just needs to be able to bring its trusted, secure, reliable proposition into these third party applications. That is why it is so dangerous and damaging for telcos to fail at self-regulation in these markets, or fail to be transparent with how they will use customer data. When the user sees a BT or Verizon logo, they need to feel, "Oh, good", not "Oh! God!".
A challenge for traditional regulation
Finally, these new markets challenge the traditional assumptions of the regulator. What was previously predatory pricing becomes a necessary prelude to gaining an audience for merchants and brand owners to interact with. What was previously a 'termination fee' is now a 'success fee' for brokering an interaction. What was previously 'significant market power' becomes 'insignificant market power' when that business process is viewed in the context of Google, IBM and VISA. Today's debates around next-generation access will become tomorrow's debates on who owns and controls the data.
Still, the future opportunity lies where it always has: in connecting people together. Those who execute can look forward to 'next-generation termination fees', which look suspiciously like unregulated high-margin monopolies. Indeed, what is there not to like? And why are you waiting?