Q&A on 'Future of Broadband' analysis
Following our earlier post on the 'Future of Broadband' keynote presentation given at the Telco 2.0 event last month, we've responded below to a sample of the questions we captured in real-time from the participants via our 'Mindshare' collaborative technology.
We've grouped them into 7 sections (Retail, Wholesale, Regulation, User Proposition, Business Models, Infrastructure, Other Industries). (Note: More detailed reports and summaries from the rest of the Telco 2.0 event brainstorming will be available privately to event participants in the next few days).
1. Retail pricing and packaging
Q: Why is Korea going back to metered? Goes against the theory of the carrot...
A: Transit and network capacity upgrades ultimately do cost money, and low-use price-sensitive users aren't going to cross-subsidise the others. Also usage patterns diverge over time (compared to dial-up's massive spike of 2-3 hours online at a constant low bitrate) making one-size-fits-all pricing ever less attractive.
Q: Given that devices will need to be open to any network, who will subsidise the price to the end user? ... The current subsidy models in some markets are responsible for putting technology into the customer's pockets that they otherwise would not be able to afford -- cf the popularity of Nokia N95 for example.
A: Not true. An artificial bundling of a credit scheme, device and service can be picked apart. Plenty of other expensive consumer electronics are sold on credit or hire purchase.
Q: Some hotels (e.g. Marriott) already have FREE WiFi as competitive differentiator...
A: So wouldn't they like some more revenue from mobile operators to allow direct roaming, and also for the operators to avoid having to buy spectrum and backhaul for all that EV-DO traffic?
Q: In fact, today I can sign up with BT Openzone and roam onto other WiFi hotspots in lots of hotels, airports etc. Same for mobile phone use abroad. So does this model already exist?
A: Partly, since the credential follows the user, but try signing onto a hotspot with both your laptop and mobile at the same time and you might find that the number of simultaneously provisioned devices is limited. Mobile roaming is still a line-based approach, not a user-based one where every device you have is on the same account. The "Connected user" concept [see previous blog post for description of this] is a long way off unless you're willing to settle for one converged device.
2. Wholesale markets
Q: In the "connected user" sale scenario, what incentive is there for (say) a hotel to provide WiFi Internet access to each room if the guest has paid someone else for "connectivity"?
A: Rather than getting 20% of the guests paying full retail you get 10% instead and then 70% of the other guests find their laptop and mobile automatically roam onto the wifi network and the hotel gets wholesale revenues for that. Plus they don't have all the billing, care and collection costs.
Q: Who pays whom for what on what basis? e.g. I pay an aggregator a fixed monthly fee for unlimited access. The aggregator pays an ISP (on what basis?). The ISP pays the hotel chain (on what basis)... Isn't the new aggregator just another player in the value chain to share the limited pot of money?
A: No, because there's huge transactional friction in today's market and taking that friction out creates value and aggregators share in that vale creation.
Q: You don't pay for the soap or the sheets separately - it's a bundle.
A: Yes, but the soap maker is selling part of the soap at retail to users at supermarkets and pharmacies, and the rest on wholesale markets. (I bet even Marriott gets its soap via intermediaries who manage all the logistics of stock management etc.)
Q: Please further quantify how 50% market growth [through new wholesale markets] can be achieved
A: It's just based on the hypothetical question of if we removed all the friction in the wholesale markets, what would the impact be. The how indeed remains an open question.
Q: How to ensure sufficient incentives to invest in infrastructure?
A: We'll be controversial and say we need to stop funding passive infrastructure directly from services. It's a different business - part of a multi-utility plastic piping company working on a 25 year planning horizon and not driven by Moore's law. The money comes from property development businesses, business taxes, as well as user fees.
4. Business models
Q: How should ISPs cooperate with alternative distribution models such as Content Delivery Networks to maximize the value for their subscribers?
A: Probably by banding together and getting scale either as competitors (directly supplying content sites) or more likely as channels (supplying Akamai) but in a better collective bargaining position.
Q: At the moment it seems that internet content providers are getting a "free ride". How can we change this?
A: Nobody is getting a "free ride", just the delivery charges are being paid in a strange way that is separated from the value the end user value gets. ("Connectivity" is an input, not an output that the user directly desires.)
Q: There seems to be a recognition that new business models are required yet a lot of investment is going into existing business models - such as for IPTV. When is this going to change?
A: As soon as the markets understand there's a "2.0" alternative the focuses on upstream (partner) revenues and not downstream ARPU.
Q: Does the [STL Partners] presenter suggest that the retail access should be turned into wholesale given the container comparison?
A: Only in part; for example, your ISP might offer a bundled (or itemised) access to Joost with a business arrangement to host/cache the content very efficiently and not count usage against a monthly cap. Joost would be party to this business relationship. Thus some of the traffic is moved from the ISP retail bucket to the BSP [Broadband Service Provider] wholesale bucket.
5. User Proposition
Q: Need to address privacy issues of user identity
A: We're going to make a rare prediction and say that OpenID (or something similar like Microsoft's card technology) will become a de-facto standard over the next 5-10 years and many of these identity issues will be solved. Remember that once upon a time things like directories were a "hard" technology problem.
Q: Is there a business model related to people who will not want to be hyperconnected? Is it possible to make them pay for this disconnection?
A: Yes! A great paradox of low cost distribution is an ever-increasing market in data and tools to filter out unwanted messages. One day you will pay more to filter calls than to make them; just as today you can easily spend more on software and hardware to manage email spam than to route the good messages.
Q: How do you define market dominance in a market that needs to be developed, before it can be dominated. Article 82 = headache
A: Today's unbundling regimes favour oligopoly over monopoly; the real need is "packet unbundling", where the pipe may be dumb but the accounting is smart, and every packet can be billed to a different (upstream or downstream) customer according to a multitude of different business models.
Q: As a user I buy connectivity from an end to end provider. They pay hotels, cafes, broadband providers for connectivity. So, it would be a bit like Oyster (London public transport payment) card then, yes?
A: Yes, a little. Many bus companies are part of the Transport for London system, but their operating license states they have to participate in the common payment system. Likewise, smarter regulation could make a big difference here in telecoms. For example, look at how roaming works: rather than bidding for my business when I land in a new country, operators are forced to access me indirectly through an uncompetitive wholesale market. A particular market structure was baked in with the SIM technology. These new B2B wholesale markets are immature enough that the playing field could be tilted towards more vibrant competition than the typical two fixed access incumbents or four wireless ones would suggest.
Q: Given the pace of technological and market change, isn't it unrealistic to expect a regulator to have some sort of oracle-like view of the future and to be able to steer the industry in that direction?
A: Yes, but given that today's industry is highly regulated, it's inevitable that even a more laissez-faire approach means interventions, even if simply by omission, to act against those with historic market power. However, a great deal of regulation does appear to have the accidental effect of entrenching existing industry structures.
Q: What are the implications for regulation? Given this much commercial disruption, should the regulator just sit back and wait to see how the market evolves, or should the regulator play a more interventionist role, helping to facilitate the shift towards greater exploitation of wholesale markets for example?
A: Unfortunately, there's no "neutral" bit of land for the regulator to squat on whilst waiting. Even doing nothing is in effect taking a position favouring the status quo.
7. Beyond telecoms
Q: What are the forces that drove the changes in the shipping model and what are the parallels in the telco business?
A: The standardised container/packet enables easy and efficient interchange between modes of transport. So the margins move to the interchanges (e.g. Googleplexes, edge cached set top boxes) rather than the interconnect between them. Also the value-based pricing regimes break down as it's impractical to open up every box/packet.
Q: What upsell opportunities do you see in the network business if you compare it to a transport or container company selling insurance for what they are carrying?
A: The opportunity is to be like, say, BT Global Services where you have in-house delivery capabilities (scale and low cost) but will put together highly tailored "logistics" solutions for delivery of business processes that cross many transport (network) types.
Q: Where is the UNLIMITED issue in the container industry... does the analogy really work
A: Precisely - you buy capacity on demand. However, you can bet that large logistics companies do commit to capacity in advance, which is effectively the same as "unlimited*" (i.e. capped).
These questions, and many, many more are analysed in the our forthcomng report. More details here.