Speaking Engagements & Private Workshops - Get Dean Bubley to present or chair your event

Need an experienced, provocative & influential telecoms keynote speaker, moderator/chair or workshop facilitator?
To see recent presentations, and discuss Dean Bubley's appearance at a specific event, click here

Monday, February 09, 2015

The rumoured Google MVNO: what's likely & unlikely?

For the last couple of weeks, speculation has been rife that Google is working on an MVNO-type service offering, probably based on both T-Mobile US and Sprint's networks, plus also incorporating a significant WiFi component. Apparently, it is called "Project Nova" at the moment.

But beyond that, details are scarce. It's not obvious exactly whom the service will be aimed at, or whether it's a big near-term strategic play, or more of a tactic to gain insight into service-provision practicalities with a view to a much longer-term project. Perhaps it's even thinking about owning more spectrum directly at some point, or influencing the direction of 5G "from the inside of the tent"?

Indeed, Google is not the first major Internet player to dip its toe into the mobile resale/connectivity market. Disruptive Analysis has already done a lot of work analysing the strategic relevance - and perhaps irrelevance - of Apple's SIM offering announced last October. (See the original blog post here and for a more detailed analysis it's included in the new update of the mobile broadband monetisation study, see here). Facebook's Whatsapp arm has an MVNO deal in Germany too. (And in fact, Google itself was rumoured to have a Spanish MVNO a few years ago).

In fact, it was the eye-watering overhype about "Apple becoming an MVNO, dis-intermediating the telcos" with its SIM card, that has prompted me to think a bit more carefully about Google, and what is most likely to be going on here.

I've been pondering a few scenarios, and the ones below seem fairly realistic. There are also some outstanding questions to which there are no current answers:

  • Will Google be issuing its own SIM cards?
  • Will Google be increasing its retail store presence? (Own-brand outlets or with partners)
  • Is this a US-only concept, or are other options possible, for example through T-Mobile's European OpCo's, or even Sprint's owner Softbank in Japan? Most Google services are internationally-applicable, although its Fiber infrastructure arm is US-only.
  • Will the service include classical "telephony" at all, or is it a data-only pitch? Will it be integrated with Google Voice in any way?

The one which everyone seems to be thinking about is a WiFi-first proposition, like Republic Wireless or Cablevision Freewheel. These would use cellular connectivity as a backup to "offload from WiFi" rather than vice versa. This is possible, but I can't help but think it's difficult to differentiate for Google - its own WiFi footprint isn't that large, so it would have to integrate with the same "curated" hotspot aggregators/wholesalers as various others. Aside from offering its own devices, this just doesn't seem "special" enough to go through all the cost and hassle.

One possibility for Google could be a unique form of pricing rather than standard "monthly plans" - perhaps part-subsidised by itself or others, perhaps customised on a per-user basis. (My idea from a couple of years ago was insurance-style pricing for mobile data). A "freemium" model could also work. This could be adjusted based on the balance of WiFi vs. cellular access, whether the user tended to consume advertising-rich apps, and so on. 

All that said, unless this is just another Google small-scale experiment, it would be extremely tough for it to scale to millions or tens of millions of users, without huge investments in sales and support infrastructure.

Instead, perhaps a likelier option is that this is - like Apple's SIM - a tablet-oriented service rather than a smartphone-based one. This gets around two problems - firstly, it doesn't need a conventional numbered "phone service", and secondly it can be pitched to the operator partners as a way of adding extra cellular devices to the market, rather than competing for market share of existing ones. Data-only connections also don't come with lots of the traditional perceptual baggage of being a "monthly plan".

Perhaps integrated both with Google's own Nexus tablets and also other OEMs' if they wanted it, this could allow various models for data access - subscription, PAYG prepay, sponsored, WiFi-primary, daily rates and so on. It would be a win-win-win if Google could increase "cellular attach rate" in tablets, increase hardware margins by selling more expensive 4G-enabled ones, and improve user utility when out-and-about, and, obviously, collect more user-data on behaviour. It might also be possible to make it all "black-boxed" enough to avoid the need for complex and expensive (voice-based) customer service, billing queries etc.

It's not obvious if Google is aiming at a US national roll-out, or perhaps initially just confined to areas where it has fibre assets. Some sort of quad-play offer could make sense, combining fixed-broadband access with TV, phone and perhaps tablet propositions.

One slightly cynical view is that this is a great way for Google to better understand, model and learn from mobile user behaviour at first-hand, before perhaps getting more involved in the network-building business in the (much) longer term - 5G and beyond, perhaps. This could give it deep insights into real-world patterns of mobile usage, without seeing them through the filter of third-party data services. Potentially, this could be a "real world" R&D play, that could feed back into product design, bundling, and new application opportunities in its core business. It's also notable that Apple is now part of NGMN, Facebook is involved with GSMA - so getting "closer to the network" makes sense for Google as well.

One other option is that the "traditional" smartphone/tablet dataplan concept is completely off-beam here. This could all also be a prelude to a verticalised M2M play. For example, Google might be launching an integrated in-car mobile platform with pre-embedded SIMs and bundled connectivity, either to be sold to car manufacturers, or as an aftermarket option for drivers. That would also fit nicely with its future aspirations for driver-less cars and the Google Express delivery service.

Taking that theme further, it could be aimed at IoT rather than smartphone/tablet use-cases. Maybe a less-awkward version of Glass or other wearables, again with pre-provisioned connectivity.

Now for the contrarian side of the post. Looking at this from an anti-forecast perspective, the things I'm not expecting include:
  • Ordinary, mainstream, standalone, "phone-centric" cellular plans competing with both the host operators and AT&T and Verizon. Too competitive, too scale-driven, too likely to annoy existing partners - and too low-margin to be of interest to Google's shareholders.
  • Realtime dynamic switching between mobile operators based on coverage, capacity etc. I can imagine an Apple-style provisioning process, or something with a multi-IMSI SIM card, but I'm deeply unconvinced that it'll hop from T-Mo to Sprint and back in realtime.
  • If it's a data-only service (eg for tablets), it might not bother with 2G/3G connections on GSM or CDMA.
  • Unlikely to be a SIM-only service, like many MVNOs outside the US. The idea of putting a cheap Google SIM in an unlocked iPhone doesn't seem something that any of the companies involved would want.
  • Sadly, I suspect that Google isn't likely to anything truly innovative with voice or video, like contextual communications, intention-based communications, hypervoice/hypersense - or especially, anything to do with privacy-invading "deep conversation inspection". There might be some WebRTC extensions for Hangouts though - especially if Chrome takes a larger role in future devices' UI.
  • It's definitely not going to have anything to do with balloons, satellites or drones, at least for the next few years.
  • Anything that causes difficulties in terms of Net Neutrality is likely to be avoided. Forget about Fast Lanes, differentiated/charged QoS - and I don't think it's in Google's interest to play games with zero-rating on its own MVNO either. Although it could go the opposite route with "loyalty" points, eg "Use X, Y, Z services and get an extra 1GB of data", without hypothecating it for use with just a specific app.
  • Unlikely that Google will have any form of deep network integration / policy-control - especially if it needs to work with two separate hosting networks with different infrastructure and platforms.
  • It's not going to have a SoftSIM. (Although it may well have a programmable SIM of some sort)
  • It's unlikely to be some sort of fixed-broadband service using mobile. Although the idea of using LTE to connect really distributed mini-servers/data-centres is an appealing, if slightly sci-fi, concept. It could be a huge network of mobile-backhauled WiFi hotspots, if I'm pushing that angle further.
Overall, just as with Apple, I don't think this is going to be a Big Bang by Google in mobile. It will likely have one or two tactical near-term wins, but will be part of a cautious, long-term experiment and fishing-trip. While Google might have a "big strategy" in mind, with an eventual vision towards 5G, WiFi / satellite / drone overlays an so on, it cannot be certain of any given scenario at the moment. It also seems highly unlikely that Google just wants to be "another carrier" - although it might try to exert some pricing leverage. I'd say there's also a 70%+ chance that this will be mostly/totally aimed away from conventional smartphones and plans.


Zheng Li said...

I do appreciate the comprehensiveness of your analysis and realistic prediction. Here is my view wishing for more disruptions:

Google MVNO: Elements for Success

Neal said...

Good overview, thanks.

One question: is there any use for Google Voice in all this? Aren't Google already provisioning numbers through this service already?

Dean Bubley said...

Hi Neal

Good question - I imagine that Google Voice could well factor into this in some way.

Maybe a single number for mobile & fixed connections for consumers, dual-ringing etc? Or other forms of shared capability. The only issue is that would limit it to a US-only model, as numbering plans are very different elsewhere in the world