Pages

Pages

Wednesday, May 27, 2009

Does the mobile network standards process inhibit business model innovation?

One of the things that struck me from the LTE summit last week was that the way that some standards bodies operate (notably 3GPP) risks entrenching legacy business models for operators and others.

This is ironic, as many standards groups, staffed by engineering-type people, try and avoid the whole issue of commercial models. This is either because they have limited understanding of that side of the industry, or limited time - or perhaps are worried about regulatory and anti-trust implications.

The problem arises because certain aspects of technical architecture can act as limiting factors. Physical SIM cards, for example, need to be distributed physically. Which means that a customer has to physically go to a store, or via the post. What seems like a technology-led decision can mitigate against particular business models, such as ad-hoc usage - and add in "latency" of hours or days to a process.

Or alternatively, dependencies between otherwise separate sub-systems can cause huge brittleness overall. LTE is being optimised for use with IMS-based core networks. But not all operators want to deploy IMS, even if (in theory) they want LTE - again, restricting business model choices or forcing them towards what is now a non-optimised radio technology.

The insistence of a lot of mobile operators to only view each other as peers (through the GSMA club, and various of its standards initiatives like IPX) is another example. This reinforces the notion that alternative service providers like Skype or FaceBook are *not* peers, but instead deadly enemies. For some operators that may be true, but for others they might be critical partners or even (whisper it) in a dominant role, for which the MNO is a junior part of the ecosystem.

Freezing old-fashioned assumptions into standards and architectures, often without even identifying that those assumptions exist is a recipe for disaster.

This isn't to say that standards are bad - but just that there is often no mechanism by which seemingly-sensible technology decisions are double-checked against potential future business models. Having a cycle in which people ask questions like "Will this work with prepay?" or "What's the wholesale model?" or "What happens if 3 people want to share one 'account'" or whatever would avoid many of these mistakes. You can never account for all eventualities, but you can certainly test for flexbility against quite a range.

Again thinking about the LTE Summit, I did not hear a single mention of the word "MVNO" during the whole event. Nobody has thought what an LTE-based MVNO might look like - or whether there might be cool features which could enable such an provider to provide more valuable services. I was met with blank stares when I asked about implementing open APIs on the radio network, to make it "programmable" for developers or partners. So I guess we won't be getting latency-optimised virtual mobile networks for gaming, then.

Many speakers appeared to view the only mobile broadband business models as traditional contract and prepay mechanisms - no talk of sponsored or third-party paid access. No consideration of the importance of Telco 2.0 strategies. No discussion about where in the EPS or LTE networks a content delivery network might interface, and so on.

One option for fixing this problem is via the other industry bodies that don't set standards themselves, but which can consider use cases and business models a bit more deeply - NGMN, OMTP, Femto Forum and so forth. Perhaps that's the level to bring in these considerations, so that they can then "suggest" specifications for the standards bodies to work to.

How about "Future mobile networks MUST be able to support a variety of MVNOs of example types A, B and C" for example?

On the same theme, I'll write another separate post soon, about why the increasing desperation to get IMS deployed is a particularly dangerous risk to the industry. In my view "legacy IMS" is a set of standards that is not fit-for-purpose in mobile - in large part because it is entrenched in a philosophy of walled-garden business models, rather than built around openness from Day 1.

6 comments:

  1. Anonymous1:47 am

    Part of the reason why you are getting blank stares when talking about LTE MVNOs is that the MNO would have to offer the network on a wholesale basis. MNOs are not there yet, and most MVNO contracts (especially in EMEA) have not evolved much beyond the voice, SMS and (a little bit) of data view that they were started with a couple of years ago.

    ReplyDelete
  2. Davide8:49 am

    "This is ironic, as many standards groups, staffed by engineering-type people, try and avoid the whole issue of commercial models. This is either because they have limited understanding of that side of the industry, ..."

    I disagree with the statement above to some extent. Both vendors and operators are involved in standardization activities through their own staff. If they do not put business thinking in what they are proposing, then it is their fault, but I believe that is not the case.
    The very reason of operators and vendors to be in the standardization bodies is to steer the output towards their own interests.
    VoLGA is an example of this situation. Some operators clearly opposed to including the solutions currently promoted through VoLGA in the standard and that is why the VoLGA initiative was born.

    ReplyDelete
  3. Anonymous - you make my own point for me perfectly.

    Have you considered that the reason that data/3G MVNOs are not common is because nobody ever thought "does this architecture make MVNOs easier or more difficult to create & run?" or "what could be done to the standard to enable richer wholesale propositions?"

    Actually, there are some reasonably sophisticated MVNOs these days, such as iPass, BT, 3 when it is on 2G, various M2M providers and so forth.

    Even away from the radio side & focusing on applications - have you ever heard of anyone talking about the impact of IMS on MVNO services?

    That is in fact one of the fatal flaws of IMS - it certainly doesn't seem to have been designed with wholesale as a primary or even secondary use case.

    Dean

    ReplyDelete
  4. Davide - fair point, there's obviously a huge a mount of politics as well, ultimately driven by commercial concerns.

    Although I tend to see these as being more vendor-centric vested interests, rather than operator-based.

    Where they are operator-based, they tend to reflect "old-school" views from legacy silo business units, rather than the newer "2.0" thinking which is evolving in corporate strategy departments.

    Dean

    ReplyDelete
  5. Gareth2:09 pm

    Hi Dean,

    So MVNO's come in very different models of thickness and thinness.

    I think that LTE and HSPA+ actually lend themselves reasonably well to an MVNO model (in some ways better than traditional voice MVNO's).

    Our good friends in Sofia-Antipolis have to their credit created (probably by implication, rather than design) an interesting MVNO model due to the roaming capabilities of both IMS and Policy and Charge Control (PCC)architectures of Release 8.

    So an MVNO could now be either a co-operative MVNO with the MNO and have access to policy interfaces between the 'home' and 'visited' PLMN, where 'home' is the MVNO. They coiuuld use this interface to request specific policy for any given session, e.g. on-line gaming, video streaming, voice, where IMS or just plain 'SIP' networks can be used for session control.

    Or by just taking a data service, they can be a non co-operative, over the top MVNO.

    Additionally the UE has the option to request differentiated service classes at establishment of a PDP context, which could be used to provide conversational, streaming or best efforts sessions, again this could be under the control of the OTT MVNO's App running on the UE.

    Additionally from a standards perspective, Open Service Architecture (OSA) in 3GPP has extended web services to provide location information, messaging, session control API's for access to networks. The key issue is for MNO's to adopt these and open up the network.

    I think the key problem may be the commercial implications of doing so, rather than people not being that commercially aware.

    BR

    Gareth

    ReplyDelete
  6. Hi Gareth

    Many thanks - that has some interesting implications for future wholesale business models. If, that is, there are effective ways for them to be commercialised.

    Would be interesting to chat through some of the issues offline at some point - maybe drop me a mail via firstname.lastname AT disruptive-analysis DOT com

    Dean

    ReplyDelete