Summary: In 2008 we identified a $375 billion opportunity for telecoms to help make media and other industries more efficient. We review the impact of broadband developments here.
In preparation for our EMEA and AMERICA brainstorms,the Telco 2.0 team has been reviewed and debated some of its core theories and previous analysis.
Two important reports, which also helped launch the Telco 2.0 Initiative, are "Beyond Bundling: Future Broadband Business Models" and "The 'Two-Sided' Telecoms Market Opportunity". Together they describe a $350bn growth opportunity for telcos to leverage their distinctive assets to create interoperable platforms that enable third party organizations to optimize their everyday business processes and interactions with customers. The reports focus on mature markets, where the greatest business model pressures exist, with a 10 year time horizon.
The $375 Bn wasn't a forecast; it was (and still is) an opportunity to add greater value to the wider 'digital economy' and grow ahead of market projections.
But we all know it's hard making predictions, especially about the future. Suggesting fundamental adjustments to how successful industries make money is also not without its pitfalls. As part of a preview of two new reports on Fixed and Mobile Broadband End-Games and 'Two-Sided' Business Model Use Cases, we thought we’d take the brave step to revisit our original analysis to see how things have changed and what interim lessons we could draw out. This first article focuses on the broadband analysis. The next one will address the lessons from the 2-Sided Business Model report.
Back in the autumn of 2007, just as the great financial crisis began to unfold, we had two key messages - the first was that the current broadband business model risked seeing costs escalate without limit while revenues fell behind, and the second was that telcos and ISPs could escape this by mastering a richer suite of digital logistics skills, seeking adjacent as well as end-user revenues, and embracing structural separation.
From the viewpoint of today, some of the predictions in the report look robust - the coming wave of streaming video, for one, and the impact on ISPs' OPEX bills, closely foreshadowed what would happen in spring 2008 with the arrival of the BBC iPlayer and the second stage of the YouTube rocket. (see also the report Online Video Market Study: The impact of video on broadband business models) On the other hand, the industry is still lagging with respect to richer value-added services, both retail and wholesale, especially if you compare with the mobile world.
However, the position isn't as grim as it might have looked from early 2008, either; despite the video surge, the ISPs are hanging on. Partly this is because of the steady progress of Moore's and Gilder's laws. Upgrading has been getting cheaper, cushioning the blow - and the rapid growth of the CDN industry (something we certainly did predict) has also been a factor in coping with the crisis. In fact, the CDN business has been a primary form of two-sidedness in the telecoms industry - several backbone operators have been developing their own CDNs, and their revenues from upstream customers obviously flow directly into the industry, but even the third-party ones like Akamai help in that they reduce costs to the eyeball networks (that is, primarily residential, rather than transit, hosting, or enterprise ones).
As far as the various "slice'n'dice" options go, these have turned out to be just as difficult as expected, and have in fact been undermined by the industry's success in coping without them. Nobody takes the M6 toll road if there's no traffic jam on the M6; one major UK operator that spent heavily on a Deep Packet Inspection (DPI) infrastructure is mostly using it to enforce their original usage cap, and another that invested in Ellacoyas (switches that perform DPI) is using them to monitor botnet (i.e. machines controlled by hackers) activity.
This coping strategy, however, has limits; Moore's law only affects things like routers, and Gilder's law (that bandwidth grows three times faster than Moore's law states for processing power) only affects the network cards, rather than the wires in the ground. At some point, either the technology (traffic growth beats ADSL2+) or the economics (it becomes uneconomical in terms of OPEX to keep going with relatively maintenance intensive copper) will force the jump to fibre. Then, the industry will be faced with a massive bill for civil works as the trenches get dug up, and the question of whether it can fund the write-off of the copper infrastructure and the capital expenditure for fibre deployment from its current business model will be back.
As far as fibre goes, a trend we didn't see coming was that deployment would be dominated by the public sector to quite the extent it has been. This takes different forms - in Australia, Amsterdam, South Yorkshire, and Singapore, the role of the state has been direct, in actually building publicly owned dark fibre or wholesale networks. In France, by comparison, it has been indirect, using regulatory changes to mandate regulated access to France Telecom and other operators' ducts, trenches, and poles.
We certainly thought that a wide range of other services, as well as voice, IP, and video, would want to ride on the fibre, and in fact, developments in the US and Australia tend to bear this out. The US stimulus plan is putting serious money into both broadband and smart grid technology, while the Australian Government is very keen on the "trans-sector concept" popularised by Paul Budde, in which multiple public services and infrastructures (telemedicine, research & education, smart grid, environmental monitoring, CCTV, etc, etc) are delivered over a common fibre network. However, in practice, this has been more likely to take the form of these services being customers of the dark fibre owner, or else over-the-top applications, than being customers for sender-party pays data from the telcos.
Another issue regarding fibre, which grows out of its increasingly public nature, is that any move to next-generation access will be heavily influenced by the LLU/wholesale market. It is unthinkable in many markets that the incumbent operators will be allowed to turn the clock back - in fact, some of the incumbents agree, like Deutsche Telekom - and in others, like Australia, the incumbents' role in the access layer is usurped entirely. However, the alternative operators (like the UK's bitstream/LLU DSL operators) have made real capital investments in LLU and associated backhaul, and any migration strategy will have to accomodate them in order to be acceptable.
That includes pricing; in the UK, initial OFCOM policy is that there is no need to regulate the pricing of BT's planned wholesale fibre access product although there is only one of it. This raises an important point about the original report; did we mistake BT Wholesale (and Openreach) pricing issues for fundamental economic ones that affected the entire industry? After all, for most UK ISPs, the immediate physical effect of surging video traffic was a bill from BT, not a fried router TCAM. Arguably, what happened was that more of the industry's revenues were swept into the incumbent, and specifically its less-regulated wholesale division, just as it tried to shuffle as much of its cost base as possible into the highly regulated, low margin Openreach.
And while this industry-wide tactical firefighting operation was going on, not without success, did the operators take their eye off a major strategic trend? We've seen plenty of innovation in voice - but it's all come from disruptive start-ups, mostly originating from the nethead side of the wire. With a couple of honourable exceptions, telcos have been letting the core voice business rot and fighting with knives over a much less profitable ISP line of business. The results are visible, as Dave Burstein points out in his current newsletter:
Ivan Seidenberg, Verizon CEO, saying "voice is dying" is a defining moment in telecom history. He didn't use those words, but his comments at Goldman Sachs are clear "we have to pivot and make a shift from the voice business to the data business and eventually to the video business. ... we must really position ourselves to be an extremely potent video-centric asset."
"The issue there is perhaps it is like the dog chasing the bus a little bit. So what I need to do is get ourselves focused around the following idea, that video is going to be the core product in the fixed line business. ... I shed myself of the burden of chasing the inflection point in access lines and say I don't care about that anymore."
Verizon remains one of the most profitable companies in the world, but the wireline business is heading downhill so fast JPMorgan writes "Action will likely be necessary to support the dividend beginning in 2012." They won't be able to support $5B/year in dividends without tapping wireless 45% owned by Vodafone. Martin Peers think Verizon will buy a satellite TV company. Knocking out one of the four TV providers is unthinkable if the Obama team is serious about competition, but that's not proven...
In the future, it won't be enough being a telco to be good at voice - the link with the copper (or fibre) in the ground is going. We certainly got that right - and a few other good shots too.
(Ed - to catch up or contribute further, join us at the EMEA and AMERICA brainstorms, or to get the latest analysis email [email protected] to register interest in our forthcoming new research reports on Fixed and Mobile Broadband End-Games and 'Two-Sided' Business Model Use Cases.)
