Telco 2.0 event: Business Model Map Q&A
At the Telco 2.0 event we presented our Business Model Map. Rashly I promised to answer all of the questions from the 250-odd participants. Given the quality of the questions, we're reproducing some of them here to help clarify the map and its implications.
In a nutshell:
- Network operators are delivery/distribution businesses: they deliver valuable bits from A to B.
- There are many ways of delivering those bits. For example, a video could be sent on a DVD, via IPTV, a peer-to-peer download, streamed from a content delivery cache, etc.
- The map documents these distribution channels for valuable bits. Which ones are you as a telco going to invest in?
- Each one is assessed on two criteria: does payment automatically flow between connectivity and the content/service ("commercial integration"), and is the delivery network hard-wired to that particular media delivery, or general purpose ("technical integration").
Q&A Business Model Map
Can you clarify the Embedded Business Model?
It's the reverse of the handset subsidy model. You buy a device which comes with connectivity embedded - a bit like those self-adjusting clocks. As we in-fill the device market between mobile handsets and PCs, we'll see a huge range of appliances, many with relatively short lifespans (under 2 years). Embedded a fixed period of connectivity into the device eliminates a huge amount of billing and revenue management cost.
In other words, as we have more tailored devices deployed, we'll see many variants on how the money flows between the hardware, service and network. In this case, it's the hardware sale that funds the rest.
Is it fair to read the matrix and bubbles in the following way: no single new business model, nor a combination, will compensate for cash flow lost on traditional business (bubbles are revenue right?)
That depends on how optimistic you are that operators can grow the "new opportunity" bubbles. In the chart their sizes all add up to the same amount in every scenario, so I'm just showing relative (not absolute) changes.
We do not understand the "protection" zone
You try to protect the existing revenue model of vertically integrated telephony and messaging. Do this by gentle feature innovation, extending the commercial footprint (e.g. non-geographic numbers, short codes) and resisting competition.
We deliberately don't call it "product innovation" because it's not about seeking new online services products (best left to Internet players) but extending the lifespan and economic model of the ones you've got.
Q&A: Tiered access and QoS
How do we deal with the situation on Tiered service (Paris Metro model) where disputes arise - proof of QOS etc.?
The point of systems like Paris Metro Pricing is that they retain the cost structure and efficiency of best-effort delivery. If you don't make a promise of guaranteed delivery/capacity (just better statistical odds of delivery), there's no dispute possible or proof needed compared to the "reserved capacity" model of QoS.
You can read more at here.
Can you detail the Tiered commercial model?
Think "multiple virtual dumb pipes". Today my computer talks to the Internet on device eth0 or wlan0. What if there were five "virtual" interfaces ("vnet0" etc.), and I could choose any one of them, but some had higher cost/lower contention? And I could dynamically switch between them? Think of it as like TCP/IP's back-off mechanism, but you step up and down priority levels as well as transmission rate.
A lot of the complexity is going to be hidden from the user. Say Skype get together with a bunch of carriers to offer Mobile Skype. Rather than re-write Skype to fit into IMS (no chance), they use a high-priority virtual pipe for the VoIP part, and a low priority one for the IM/file transfer.
One day, you might have hundred or even millions of "virtual Internets" ("VWANs", to mirror Virtual LANs). A bunch of Xbox gaming devices might talk to each other on a virtual network, safe from denial-of-service attacked from the general Internet, and with the priority and latency they desire.
Does traffic prioritisation have some value as a business model at least in the short term?
Short term, yes. Enterprises, absolutely. And there will always be some friction in the system between different generations of technology that some network smarts will be needed to smooth over.
The comment was made that it was desirable to move away from deep packet inspection - what model would be better?
DPI isn't all bad. For example, I'd retain it as a means of identifying "bad" packets like DDoS attacks, spam, fraud management. However, as a means of value-based pricing, it's kaput. Even from a physics standpoint, it's a non-starter: move all the bits from the optical transmission domain to the expensive and slow electronic processing domain. I'm sure Cisco would like you to do it, but it doesn't mean it's a good idea. You're just in an arms race with your own customers to cloak the traffic. A better model is either (i) spend the money on more capacity and stick to all-you-can-eat broadband, (ii) go to the Paris Metro Pricing style of tiered dumb pipes plus offload the heavyweight traffic to CDNs, (iii) lock down the edge devices to eliminate the problem at source.
Was the Paris Metro Pricing a way of Bell Labs ingratiating themselves with the new French owners
Bien sur! Err, no. The research was done back in the days of Bell Labs, before croissants came on the cafeteria menu.
Q&A: Low vs high integration models
Is the future in the radical advances of end-user devices (high memory, CPU power, client-based apps)? Does this accelerate the 'dumb pipe' scenario for infrastructure?
Yes - ever more applications become possible to deliver over a general-purpose network with a general-purpose handheld computing device.
How do you move from a world of one size fits all broadband into a slicing and dicing world? If broadband is unlimited, who would want to bundle it into a device? Will users be willing to move off their unlimited plans to tiered plans?
As "legitimate" P2P and video download traffic moves onto content delivery networks and separately packaged offerings (that avoid peak hour, keeping capex down) the price of "unlimited" broadband will probably rise. For the average email/web/Youtube user, they won't notice a thing. But the broadband "hogs" most certainly will. We've already seen the first steps in the UK with BT wholesale going from an unmetered to a metered product. (I'd expect a congestion-based pricing scheme to be next.)
Do consumers really dislike vertically integrated offers? It is the cost and service that matters. Given cheap offering with good service, one probably will buy everything from one place.
They love them! However, technical vertical integration makes for incredibly slow progress (e.g. voice telephony network) compared to "dumb pipe" applications (e.g. social networking). The trick is to preserve some of the commercial integration of service and delivery without the technical integration.
Can you really integrate commercially without any technical integration/leverage? Explain...
Sure - you're just accounting for the packets differently based on their origin and destination. There's no magic - as long as there's some compensation that flows between the device, content or service and the underlying connectivity provision, then you have some form of commercial integration.
Packaging service/devices etc with connectivity is a good answer, but the worry is that the examples James gave are actually about startups that help users unbundle.
These two trends are not mutually exclusive. Artificial bundles will be torn apart, but where they create convenience to users then they'll buy up packaged offerings.
What are the fundamental technical drivers of business model & structural change? For example, does Moore's Law and smarter end-devices MANDATE that a dumb pipe model will evolve?
Blame Claude Shannon, the father of information theory. The information carried by a signal is independent of how it was delivered. Dumb pipes are less complex than smart ones, all other things being equal. So if you can get the application to work over the cheap pipe, why pay more for a specialised one?
Q&A: Telco culture and organisation
How bad does the crisis have to get internally/externally at the telcos before they actually embrace change? Do we need to go through a real period of agony before the willingness to change becomes real?
We're already there. The ISP industry is already a world of hurt except for some niche players, with exits and M&A on the up. The crisis is as much for the users as for the operators, who can simply delay investment. The users miss out of the experience they require, and the economy never gets to grow new industries.
What is the future for incumbents
In the short-medium term, they're OK as there are such large barriers to entry and a regulatory tar-pit to negotiate. In the longer term most will disappear (consolidated into a few mega-operators with the necessary scale). Many will exit the application services space or become effectively different businesses (e.g. BT's global services, where the network is just an enabler, not the product). Probably a similar story to what's already happened in other utility and media industries. British Gas used to retail cookers - but no longer.
What should telcos do today? What's the first big tipping point?
Decide which core Telco 2.0 strategy you want to follow: diversification, protection, platform or pipe. Then exit the non-core stuff (but preserve the revenue streams via partnerships). "Tipping points" are a bit of a semantic illusion, it's a continuous process of evolution.
In terms of broadband and mobile who owns the customer today?
The customer owns the customer. At best, you own some expensive glass and electronic boxes, plus some databases of variable quality. If "2.0" is about anything, it's the users being in control of their destiny.
What might the cultural changes be for Telcos to become more innovative?
Innovation does not equal invention. Innovation is simply applying inventive ideas to bring those things to reality. Telcos are plenty innovative enough, just not in end-user services. Most of the innovation is in deployment and operational management of networks, which you never get to see. The evidence is in how well they've adapted to extremely rapid technological change. I don't think it's possible to create an innovation or invention culture in a mature organisation which doesn't have one.
What's the underlying structure out of your perspective which makes change so difficult? No moralistic judgement -- a structural explanation
I don't think change is difficult. The telecoms industry has undergone incredible change in the past 25 years (mobile, broadband, optical transmission, regulatory) - as much if not more than most other industries. How much has Google really changed? You underestimate telcos.
Why are Telcos so low in R&D spend?
Shareholders have done really well out of gatekeeping other people's innovations being distributed over their pipes. Maybe telcos are like travel agents: high turnover as you act as a handling agent for 3rd party products, but that artificially inflates your revenue and expectations for research spending. Given the efforts of the Internet companies and telco equipment vendors, overall there's no lack of R&D in communications.
How do telco's strike a balance between commercial innovation and innovation in technology which is often their bargaining card/source of competitive advantage?
Easy - exit the services innovation space and partner. Focus on integration and overall user experience across multiple services and payment/delivery methods.
Q&A: Peer-to-peer & video distribution
Can you expand on the telcos' exploitation of peer to peer and content delivery networks?
"If you can't beat them, join them." Telco IPTV isn't going to deliver the Lithuanian shows I want my kids to watch; or give my brother his dose of British TV out in California. There are going to be many different aggregators, editors and recommendation engines. Operators need to make it easy and advantageous for those video distribution networks to minimise backhaul costs and contention. Today there's no incentive to make your P2P application operator network friendly.
Since P2P is so successful, can Martin and James suggest ways for Telcos to monetise it?
Just deliver the damn bits! Someone else made them valuable, so don't expect the extract the premium for content creation. Build content delivery networks that are cheap to run and flexible to support many different aggregation and delivery models. Work to lower capex by creating incentives to avoid transfers during peak network activity.
What should the service providers do about the cost to them of "Over the Top" content traversing their networks? Deep Packet Inspection does control this and requires intelligence in the network.
Throttling users in arbitrary ways is just going to lead to network neutrality legislation and consumer backlash. Find ways and incentives to get those P2P apps to deliver their content in cheaper and more efficient ways. Stick 100Tb of BitTorrent cache in every central office!
Operators need to block P2P traffic! The people that pay the bills for broadband won't care! I would happily pay for fast content delivery...but you need to free up the network first.
This just isn't feasible in the long run. Users don't know or care what the delivery technology is, and if you give them a bad experience (or bill shock) then you lose too.
How do we reconcile the cost of carrying bits with flat-rate in a video dominated traffic model?
You can't which is why the current Internet model will fragment and video distribution will be done via cheaper methods. Just like you send most parcels by parcel post or UPS ground service and not next-day 1st class mail.
Moving complexity from retail models to wholesale is far from simple. Peering agreements are complex - how can this be driven?
I think this is an unsolved problem, and the industry is going to have to work out new structures for aggregating wholesale connectivity of varying quantities and qualities and packaging it together. (Telecom probably ends up looking more like the mortgage re-sale market - build a network, and then leave the problem of filling the pipes to someone else.)
Please explain further the shift of complexity to wholesale from retail - give a practical example?
Easy - look at the Blackberry or Three's X-series. You don't need to sign up to a data plan, read the terms and conditions to see if you can do the things you want, and watch the megabyte meter all the time.
Is there an advantage for network operators to break themselves up (like BT)? Is there more shareholder value if operators split into retail and wholesale?
Yes.... But. In the case of BT, despite the Chinese walls, the CEO and board can effectively synchronise investment across the two halves of the business. BT has a good balance between the discipline of separation (wholesale can't rely on one captive customer, retail can't rely on a distribution monopoly) and the need for services and capacity to be deployed in sync. If there was no such synergy, we'd probably have seen a break-up already. We heard some evidence at the event that the City rather likes the structural separation model, as utilities have higher multiples than volatile telco services.
Regarding shifting of complexity from retail to wholesale: why should it be done, and how can it be done? Isn't it just about shifting the risk? The risk is still there!
The WHY is because you want to manage the capex cost associated with the delivery of low-value (and often pirated!) video content. This is generally time-insensitive and needs to be shifted out of peak hour (or where economically unsustainable, eliminated via raising metered broadband prices at higher usage levels). The HOW is documented in the business model map - go read on the Telco 2.0 blog my 4 articles on it. You're mitigating risk by taking some control over it, and aggregating risk to smooth it out. For example, with the BT Openreach structure, they make money no matter which one of Sky, Virgin Media, BT Retail, Carphone Warehouse, or anyone else turns out to be the winner.
Q&A Customer data and relationship
What about the future of data ownership and control? Isn't this the future cash cow that perhaps telcos, if they can become platform players, can excel in?
Yes, in principle. Just as Google extracted the latent value from the collective effort of hyperlinking, telcos could feed social network applications with the squeezed juice of CDRs. Whether they can finesse all he legal, branding and cultural issues is another thing.
Google has been successful by storing, analyzing and monetizing what users are actually doing. Telcos have/own lots of data, what are the challenges in building business models around this data versus charging for access?
Privacy, regulatory, brand, encryption technology, user experience,...
It's going to be tough.
How to provide user-centric solutions when telco model still largely views the household, not the individual, as the customer
That's both a problem and an opportunity. Yahoo! don't have a clue who I am at all, so the telcos should see their glass as half full here.
Q&A Partners, portals and services
When will web services stop being free? What will be the catalyst
They're not free already: you have to buy an expensive PC and broadband connection, and then watch a bunch of ads (or give away valuable private data to enable those ads). We're just moving to a world of many different business models linking the devices, services and connectivity.
Should telcos partner with Google, MSN, Yahoo on Consumer apps and focus on Enterprise / Business opportunities?
Depends on each telco. These aren't mutually exclusive.
How can the telco better leverage the customer's various contact directories?
Sync with network, offer an API, enable 3rd parties to access it, take out the barriers to usability!
How about this as a business strategy: rather than development & investment in new services, adopt a simpler strategy of "seed and buy" successful small competitors and using the telco capital for acquisitions?
Yes, telcos could become like Cisco and vacuum up the good services ideas. However, what's the point if they are then limited to one operator's distribution network? It's like building supermarkets for Toyota car owners. Totally unnatural. Although maybe Cyworld and SK Telecom is the exception that proves the rule.
Partnering is obviously a critical success factor in the new world - how should this be embraced?
Big question. The first step of anyone embarking on a "platform" route is probably to separate partner management into its own exec position independent of BizDev.
Q&A Service innovation
Isn't innovation with big Telcos more like "we are because we are expected to"? Is it still a "show business" or real one? Why are there no ground breaking innovative products, if the above case is not true?
It's been very comfortable selling vanilla voice and messaging. The risk/reward for telco managers doesn't make it worth trying to jump that shark.
Where is the innovation coming from? Generate it from the inside or stimulate external creation? How to drive, channel and capture innovation?
All the above - depends on each operator's situation and what opportunities present themselves.
My hypothesis: the endgame will be: there will be money in Access, in Devices, but not in the three C: Communication, Communities, or Content. These will be free, paid by ads or thrown in for free by the telcos.
I think all five of them will be fine businesses. Content will become more personalised and interactive, meaning much of the copyright crisis will naturally subside. Is World of Warcraft content?
Q&A ...and the rest
What value or cost are users attributing to access the internet to make use of all these services?
They don't care about the Internet per se. Campaigns to "save the Internet" have zero resonance with the public.
How quickly will regulatory bodies and policy makers react to this change?
About five years too late, if history is any guide.
To what extent can the overall telecoms spend purse be grown? How many other purses can telecoms raid?
Banking is ripe for disruption, and DoCoMo are probably the ones to watch.
What of the telecoms vendors? Especially the big complex integrated ones?
I think I'll need consulting dollars to answer that one.
It's going to be a whole lot smaller as an industry. Again, other industries have led the way, e.g. GE into "power by the hour" rather than turbine sales and blending services with hardware.
What are we going to do with all the copper?
Put Chilean copper miners out of business by flooding the scrap market.... There's almost 70m tons of it in the legacy access networks.
What are the payment models for future fragmented services?
Equally fragmented, which will probably drive a user experience crisis in the medium term as the user switches between metered, free, ad-funded and "bundled" service connectivity.
What tech features telcos need to have on client platforms (PCs) to enable Telco 2.0 services?
The Virtual SIM that Intel described would be a great start, so we could provision applications with network access, rather than whole devices.