P2P may be Google’s biggest weakness and an Operator’s biggest asset

serverfarms.jpg

Notes: Let me say at the outset that I am a big fan of Google. This blog uses Google as an illustration in answer to the question Operators often ask; What can we do to differentiate against the Web (an euphemism for Google). Similarly, I have used Motorola’s Seamless mobility ideas since they are closest to what I am trying to illustrate on the network side. I am also reasonably familiar with the ideas of Seamless Mobility from the time I was on Motorola CTO Padmasree Warrior keynote panel at Javaone

Today I spoke at the Mobile Wimax conference in Cannes. The venue was one of the best known places in Cannes Palais des Festivals et des Congrès(location of the Cannes film festival)

I have historically had an interest in Wimax and believe in its disruptive potential. One of the themes of the conference was : ‘Operator strategies for mobile broadband’. So, I discussed the following idea in my talk

P2P may be the killer application for the mobile networks because P2P is the biggest weakness of Google(and other web based services)

To illustrate this, I used the example of ‘Seamless mobility’ from Motorola – because in it’s ultimate incarnation; Seamless mobility is a very disruptive idea if deployed by visionary Operators.

Firstly, seamless mobility, as I understand it is: A user should be able to go from one network to another without any disruption to their experience. In an IP world, intelligence will always shift to the edge of the network. Hence, to devices and services. Thus a service layer that spans different network types will be a needed in more complex scenarios.

I have been speaking of a the need for a new type of value chain in my blog Isn’t it time we put the customer at the centre of the mobile data value chain:A new value chain for the Mobile data industry

What would be needed in such a scenario from a network standpoint?

The device should be able to choose the optimal connection type from a range of connections possible. The connection /session should be managed across heterogeneous networks. Physically , that means a seamless integration between network types like Wimax and IMS/NGN.

If this scenario is realised(even partially), the next question is: How will it be used?

We can extrapolate existing business cases (for example corporate users etc ) and those are all well and good ..

However, the ubiquitous and plentiful availability of broadband connectivity may trigger a fundamental change in user behaviour and applications – the likes of which we don’t see at all currently.

Specifically, we may see true Peer to Peer applications.

To recap from wikipedia ..

A peer-to-peer (or “P2P”, or, rarely, “PtP”) computer network exploits diverse connectivity between participants in a network and the cumulative bandwidth of network participants rather than conventional centralized resources where a relatively low number of servers provide the core value to a service or application. Peer-to-peer networks are typically used for connecting nodes via largely ad hoc connections.

Sadly, operators and many in the industry don’t think in this way – if you say ‘P2P’ people think Napster ..

That’s sad because the telecoms industry has an identity crisis between content and contact .. i.e. many in telecoms forget that the industry is all about connecting people and not about playing second fiddle to the content industry (and all the baggage that comes with it – such as DRM)

A true P2P ecosystem triggered by the idea of seamless mobility and plentiful bandwidth can be very disruptive indeed …

And what has this got to do with the Web and Google?

Think server farms ..

Google is client server .. It has to be .. if it needs to store, manage and index the Web. Indexing the Web, as we know it, has to be centralised ..

Consequently, if a decentralised – P2P architecture takes off – then Google cannot match it because it is not in Google’s DNA to do so(just as the Web was never in Micosoft’s DNA) .. And furthermore .. it will be a service which people will want(Doubt that? – My Skype account shows 8,663,106 online at the moment!)

This is classic Sun Tzu ..

Unfortunately, more strategies are driven by Mickey Mouse than by Sun Tzu ..

And I mean that quite literally .. in the sense that it is driven by the content industry aka the Disney’s and the Warners of the world.

The telecoms industry sadly does not realise that it is in the communication business – and not the content business.

That should however not detract us from the vision of P2P triggered by the concept of Seamless mobility .. which, in my view, is perfectly valid and highly disruptive .. especially since it is a unique advantage which the network can foster ..

Which was the original question I was addressing: What can Operators do that is unique and different in a Web world and still add value?

Thoughts and feedback welcome ..

Image source: http://www.anticlockwise.com/farm2.jpg

Update on comments ..

Many thanks for the great feedback

I agree with Tope in the comments (and which was my point) i.e. we can’t look at the future with the ecosystem of the present. P2P is highly disruptive and no one really knows(and I dont claim to make any predictions either) how users will use all this bandwith and connectivity in a P2P mode.

I am however a BIG believer in the power of grassroots and in the belief that empowered,connected individuals can drive grassroots change(much the same optimistic view I take for Africa and mobile technology – of which I am a big advocate) i.e. the more links you can create between people, the more the emergence of an ecosystem which will be truly vibrant – and whose ultimate form we cannot see based on the views of today .. Thanks all – appretiate the feedback

Comments

  1. Embracing P2P for the sake of being different may not be important…. In your example above indexing using P2P or centralized networking. Note that indexing can be accomplished using P2P method of work distribution, but it also requires the distribution of the P2P node/executable as well. And that goes back to the logistical issue of managing such nodes. That is why centralized is simpler, more controllable, yet, it is centralized (processing power); even if quasi-centralized i.e. centralized by region.
    Carriers need other differentiators — and we are starting to see them w/ Verizon’s announcement that it is going to open the network. Other ways is to provide better services on the their network such as integrated billing, etc to developers and allow open network, any device, any application.
    ceo

  2. Ajit Jaokar says:

    Hello Enrique, Thanks for your comments. I agree that P2P should not be undertaken for itself and that there are benefits of centralization. My talk was more of a strategic approach – and I also feel that the benefits of P2P are yet not fully explored. It is also a place where the network itself can add value(and I mean a fixed and mobile network). Appretiate your comments as usual!

  3. Ajit,
    I can agree with the notion of seamless mobility – the ability to simply transition across networks.
    As for P2P – I would disagree. Why? Because P2P is really nothing more than a bandwidth saving for “large” files. And typically those files are either music or video which are not typically going to be downloaded OTA (Over the Air) on mobile.
    You’ve been using the Web for awhile. Tell me how many P2P clients you have installed and use on a daily basis. My bet is none.
    Google’s brilliance was not building a better search widget and getting lots of people to use it, but rather unlocking the economic value of search.
    The key to mobile is unlocking it’s economic value. P2P doesn’t do that. Seamless connectivity could do it, (it’s like having a highway available anywhere you want to go) – what makes that valuable is the services available for the customer.
    P2P is not a service – it’s a feature which only techies use.
    So back to your original question – What can Operators do that is unique and different in a Web world and still add value?
    Start offering tools that help web designers make it easier to build and support mobile content for their installed base.
    Peter

  4. Sebastian Thalanany says:

    Hi, Peter,
    Yes, your point is well-taken.
    Seamless mobility support together with an open access policy, in concert with a service environment that leverages third-party application development, would be a catalyst for service-oriented revenue streams, and market expansion. At the same time, P2P applications do have their niche, and will continue to be attractive
    Thanks.
    Regards,
    Sebastian

  5. Tope Omitola says:

    Hi Ajit
    Great article. Good insights. And PeterC, nice rejoinder.
    PeterC’s rejoinder reeks of short-termism and the inability to look forward 5 years from now. Especially this:
    >P2P is not a service – it’s a feature which only techies use.
    Techies may be the one using it now, but dont you think they use it because they’re a few years ahead of the curve? I read somewhere that 40% of Internet traffic is P2P. Is this true?
    Isnt Skype (and service slike it) P2P?
    5 years ago, video on the Internet was mainly done by techies, now everybody’s doing it. Sometimes it’s good to lift one’s head from the current milieu and gaze into what may be coming.
    So, I think Ajit is on to something here.
    Tope

  6. Ajit Jaokar says:

    Thanks Peter, Seb and Tope. I agree with Tope (and which was my point) i.e. we cant look at the future with the ecosystem of the present. P2P is highly disruptive and no one really knows(and I dont claim to make any predictions either) how users will use all this bandwith and connectivity in a P2P mode.
    I am however a BIG believer in the power of grassroots and in the belief that empowered individuals can drive grassroots change(much the same view I take for Africa and mobile technology – of which I am a big advocate) i.e. the more links you can create between people, the more the emergence of an ecosystem which will be truly vibrant – and whose ultimate form we cannot see based on the views of today .. Thanks all -
    appretiate the feedback

  7. I agree with Ajit and Tope
    P2P is a concept that puts the power of end nodes to work, reducing (if not eliminating) dependence on centralized systems.
    What is needed to unlock the power of internet-connected intelligent computing devices is, at the most basic, a table with pointers. E.g. Napster, e.g. Skype: here’s the IP that will provide access to (file/person)
    And even this task can (and is) distributed to always-on PCs (supernodes).
    If we look at it this way, what can be done using this principle, besides sharing files and connecting VoIP end nodes?
    Example: It is quite likely that cellphones will be the music repository of choice. Could a smart agent on my phone register how often I listen to songs? Yes. Could that agent put songs I haven’t listened to in a year for sale? Why not. Would that be P2P?
    A mobile device, as my most personal computer, could host a smart agent that provide presentation services regarding things I want to sell. And automatically negotiate with potential buyers.
    Just my 5 cents…
    Regards, Martin

  8. Ajit Jaokar says:

    Thanks Martin. I like your thinking i.e. ..
    >>>
    If we look at it this way, what can be done using this principle, besides sharing files and connecting VoIP end nodes?
    <<<
    Thats the key here! rgds Ajit

  9. Ajit,
    I think that p2p and social networkings will be the future of mobile content services. Operators are the kings in this esenario but they have to start knowing there user. Today for example in latam 85% of the mobile phones are prepaid and operators really dont know how are there users. By the other hand 75% of the mobile content is download thrue the operators master portal. Today things are changing and we are starting to see that social networkings that knows there users ares starting to sale mobile content with great succes. Tomorrow these social networkings coul be MVO and offer to there users the best prices from different operators. But for all that you must now your user very well and that is very expensive in resources. That is why we develop a soluction call MOOGA ( http://www.infinitemoco.com) MOOGA is a self learning mobile entertainment ecosystem incorporating Artificial Intelligence techniques to understand, track, predict and recommend mobile content based on individual user tastes, downloads and popular contents. MOOGA is a platform that was conceived to be used in operators because it brings to the mobile world concepts which has proven to be successful in the internet such as the long tail and the wisdom of crows.
    Here you can see a demo:
    http://www.infinitemoco.com/english/mooga.htm

  10. bernard lunn says:

    interesting thread. I agree that P2P is the only alternative to giant server farms (which give a few firms with $$$$ billions to invest a big barrier). There does seem to be a link to social networking in the sense that we relate through content that we share through P2P networks.