A month and a half ago, I wrote a series of posts around Google’s announcement that they would become an MVNO and offer wireless service. The final post in that series was titled “What Might Google Really Do?” and it included my predictions on Google’s potential plays, based on what Google had actually said, and what they had historically done. Now that Google has officially “launched” Project Fi, it seems like a good time to check in on those predictions.

It’s important to note that, at this point, Google is launching Fi with an “Early Access Program” that is by invitation only. Some aspects of how the service will be delivered in the future will likely be quite different from how it is delivered today (undoubtedly based on lessons learned during the EAP) and some details aren’t yet announced.

But here’s what we do know. Google announced Fi via their official blog on April 22. They said “today we’re introducing Project Fi, a program to explore this opportunity by introducing new ideas through a fast and easy wireless experience. Similar to our Nexus hardware program, Project Fi enables us to work in close partnership with leading carriers, hardware makers, and all of you to push the boundaries of what’s possible. By designing across hardware, software and connectivity, we can more fully explore new ways for people to connect and communicate. Two of the top mobile networks in the U.S.—Sprint and T-Mobile—are partnering with us to launch Project Fi and now you can be part of the project too.” They then outlined three specific areas of focus and innovation.

High-quality network connections: “We developed new technology that gives you better coverage by intelligently connecting you to the fastest available network at your location whether it’s Wi-Fi or one of our two partner LTE networks.”

Communications across networks and devices: In addition to working across WiFi and LTE, Google says “With Project Fi, your phone number lives in the cloud, so you can talk and text with your number on just about any phone, tablet or laptop.”

A simple service experience: “We offer one simple plan at one price with 24/7 support. Here’s how it works: for $20 a month you get all the basics (talk, text, Wi-Fi tethering, and international coverage in 120+ countries), and then it’s a flat $10 per GB for cellular data while in the U.S. and abroad. … Since it’s hard to predict your data usage, you’ll get credit for the full value of your unused data.”

Here are the predictions I made, and a comparison with what we now know about Fi:

1. “Google would effectively be proving out new/unconventional approaches to connectivity offers (e.g. unlimited) in a way that proves out to the operators that there’s market demand (enough to be a threat) and that the economics can work (so that it’s attractive)” – This clearly seems to be the case. Instead of unlimited, the real innovation around the plan is refunding customers for unused data. T-Mobile’s CEO has welcomed Google’s “fresh thinking” implying openness to learn from Google’s experiment.
2. “I also would expect the scale to be limited, meaning it would have relatively limited retail impact on the operators” – this clearly is the case with the EAP and Google seems to continue to signal limited scale and the operators don’t seem threatened.
3. “I also wouldn’t be surprised to see Google want to move it around, so maybe each new Nexus device launched is a new MVNO on a different operator or set of operators” – Time will tell.
4. “I doubt they’ll try Google’s original Nexus web-based distribution” – For the Early Access Program (EAP) Google is using web-based distribution.
5. “They might try using their physical “stores” in Google Fiber cities” – Not yet anyway.
6. “They might also strike a distribution deal with big box retailers, like Best Buy or WalMart” – Again, not yet.
7. “I wonder if Google isn’t actually negotiating with the mobile operators to sell the service in their own stores or through their distribution channels” – Again, not yet.
8. “I doubt that Google has a desire to employ tens of thousands of customer service reps in both owned and outsourced call centers around the world” – Google has said that customers can call 24×7 and speak to a live US-based agent, but hasn’t indicated how they are providing this support.
9. “They may be able to leverage the care resources they’ve put in place to support Fiber” – We don’t yet know.
10. “Perhaps, they are going to leverage the mobile operator’s existing customer care infrastructure” – We don’t yet know.
11. “They will likely pair the service with a new Nexus device” – The service is only available with the Nexus 6 which has specific hardware and software to support the network switching unique to the service.
12. “Google’s issue will be ensuring that only the right customers for their experiment are the ones that choose their brand for wireless” – The invitation-only EAP will help Google target the right customers.
13. “Providing openness and choice, managing the network in an open, non-discriminatory, transparent way and giving users a choice of multiple service providers, may be an objective” – This hasn’t been emphasized in Google’s announcements.
14. “I can’t imagine that Google would see enough potential upside from [a full competitive entry going head-to-head against Verizon, AT&T, Sprint, and T-Mobile] to offset the serious downside it would have on their core business.” – There’s no indication that Google is pursuing an aggressive attack against the existing operators.
15. Maybe “it’s really all about IoT” – so far, it seems to be a smartphone plan, without any IoT elements.

So, out of 15 predictions (most of which were “mights”), I would say that five were aligned with what Google has announced (1,2,11,12,14), three predictions were wrong (4,13,15), and for the other seven, we just don’t know yet. We’ll have to keep watching.

Google’s entry into any market is cause for existing players to pay attention and potentially be alarmed, so it’s no surprise that the news that Google will become an MVNO and provide wireless services has many forecasting doom and gloom for the existing mobile operators. Before we can jump to those conclusions, I think it’s wise to consider the different scenarios that, given what Google has said, and what they’ve historically done in mobile/telecom, have some level of credibility.

Let’s start by reviewing, briefly, the challenges that MVNO’s have traditionally had to solve. I think they fall into four buckets: distribution, customer service, devices, and brand. I think Google is in a very different place than the vast majority of MVNOs when it comes to these four topics, given their objectives and their starting point.

For distribution, Google’s original Nexus web-based distribution experiment failed, I doubt they’ll try that again. They might try using their physical “stores” in Google Fiber cities, although this isn’t likely to get them enough customers to provide meaningful scale and impact. They might also strike a distribution deal with big box retailers, like Best Buy or WalMart.

However, given Sundar Pichai’s comments, I wonder if Google isn’t actually negotiating with the mobile operators to sell the service in their own stores or through their distribution channels. This would be unusual, but not unprecedented.

When it comes to customer service, mobile operators employ tens of thousands of service reps in both owned and outsourced call centers around the world. I doubt that Google has a desire to establish that kind of customer care infrastructure. Again, it’s possible that they may limit this experiment to Google Fiber markets, in which case, they may be able to leverage the care resources they’ve put in place to support Fiber, or, perhaps, they are going to leverage the mobile operator’s existing customer care infrastructure, as with distribution. Again, this isn’t typical for MVNO’s, but I imagine the operators would seriously consider the potential incremental revenue this would generate.

MVNOs have often struggled to get deals with OEMs for devices because they can’t commit to enough volume to make it work. In recent years, Sprint, for one, has tried to help MVNOs overcome this challenge with their BYOD program and their custom-brand, white label program, but if Google wants to innovate in software, hardware, and connectivity, this won’t be an option. Of course, for Google this also isn’t the same problem as it is for other MVNOs, since they will likely pair the service with a new Nexus device, which gives them a unique position with OEMs. This likely is easily solvable for Google.

Most MVNOs in the market are new brands that must invest significantly to establish a position with a narrowly targeted segment. Google doesn’t have this problem. If anything, Google’s issue will be ensuring that only the right customers for their experiment are the ones that choose their brand for wireless.

Second, I think we need to clarify Google’s objectives with this experiment. Google wouldn’t be investing in this experiment if they didn’t think it would create direct or indirect value for their business. That being said, I doubt that Google believes they can make money competing with Verizon, AT&T, and the others with traditional cellular service.

As with Google Fiber, they may believe that Mobile Operators are constraining use of the Internet and applications and that they can introduce “innovations” that the existing players need to respond to, changing the overall trajectory for the industry.

Net neutrality, or to use the Google Fiber terminology, providing openness and choice, managing the network in an open, non-discriminatory, transparent way and giving users a choice of multiple service providers, may be an objective. Clearly Verizon and AT&T are going to resist the FCC’s new rules and Google may want to have market pressures to combine with regulatory pressures to ensure that the operators adopt “open” policies.

Another target may be the strong trend away from unlimited plans. The FCC’s new rules actually are likely to accelerate the move away from unlimited since it takes away the option for Mobile Operators to throttle unlimited plans. Any customer that doesn’t have unlimited has to stop and think about whether or not to watch that YouTube clip while on the go, or before they do just about anything bandwidth intensive when not on WiFi. This constrains use of the Internet and therefore impacts Google’s core business.

Finally, let’s not ignore what Pitchai presented as Google’s objectives during the interview. Although improving WiFi to cellular interworking and making problems like dropped calls less painful are noble goals, I don’t think that pressuring Operators to implement those types of improvements would truly justify Google’s attention. I think, more likely, as Pichai hinted, maybe this isn’t about traditional cellular service at all. Maybe this really is about the Internet of Things – clearly a space that Google is investing in at the device and software level. Maybe Google wants to make sure that the beyond-WiFi connectivity is being developed in a way that serves Google’s objectives.

So, with that as a framework, let me propose three different potential scenarios for what Google might really do.

First, this really could be like Google Fiber – disguised as an “experiment” but really a new business, competitive entry into the mobile service space. The biggest challenge with this scenario is that Google will be dependent on the mobile operators for at least network capacity, and that’s never the position you want to be in when you’re trying to disrupt the operator’s business (just ask the CLECs of the late 1990s who tried to resell RBOC service under the Telecom Act of 1996). Next, if Google were to pursue this approach, at least all operators not providing Google’s underlying service, would drop or deprioritize Android devices in their portfolios, seriously hurting Google’s momentum and leadership in the smartphone OS space. I can’t imagine that Google would see enough potential upside from this approach to offset the serious downside it would have on their core business.

As a second scenario, let’s take Pichai’s comments at face value and assume that this truly is a smartphone- and/or tablet-centric experiment, working closely with the operators. In that case, it would look a lot like Nexus. I wouldn’t be surprised to see Google rely heavily on their operator partner(s) for distribution and customer care. I also would expect the scale to be limited, meaning it would have relatively limited retail impact on the operators. I also wouldn’t be surprised to see Google want to move it around, so maybe each new Nexus device launched is a new MVNO on a different operator or set of operators. Google would effectively be proving out new/unconventional approaches to connectivity offers (e.g. unlimited) in a way that proves out to the operators that there’s market demand (enough to be a threat) and that the economics can work (so that it’s attractive).

The third scenario is that this really isn’t about smartphones and tablets at all, but it’s really all about IoT. Google obviously is making big investments in hardware and software for IoT, so it would be natural for them to invest to get the “beyond-WiFi” connectivity to work for them as well. AT&T has had meaningful success with IoT, and I think Verizon still has serious hopes for the space, so they might not be the first to open the door to Google’s entry into being a connectivity service provider here, but I think other operators may be more than happy to have Google’s wholesale business and to help define the de facto standards that others likely need to adopt.

Of course, all of this is pure conjecture. I have not been privy to any discussions between Google and mobile operators. There’s more that we don’t know than we know, at this point. However, I think these three scenarios outline a solid framework for anyone to consider the impact on the industry as a whole, or their particular business.

This should be fun to watch!

Just as Sundar Pichai did, I think it makes sense for us to look historically at Google’s forays into mobile and connectivity. I think there are three historical precedents to consider: Android, Nexus, and Google Fiber.

Android
Google followed Apple into the smartphone market. You can either say that, together, they created the smartphone market, or you can say that they significantly disrupted an existing market dominated by RIM (Blackberry), Microsoft, Palm, and Nokia (Symbian). Google had virtually no meaningful relationships with any of those four, but Android was a key element in the destruction of what had been a very strong relationship with Apple.

Including Apple, four of the five market leaders all had an integrated hardware/software approach to the market. Google chose an “open” or “ecosystem” model, similar to Microsoft’s successful approach to the PC market. In fact, the initial announcement of Android was made by the Open Handset Alliance, made up of 34 companies including OEMs, Operators, Developers, and Chipset companies.

Today, by far, Android is the dominant smartphone operating system. In his talk last week, Pichai claimed that 8 out of every 10 phones shipping around the world are running Android. Google has built a strong relationship with OEMs and, somewhat less directly, with Mobile Operators, to get Android to market. It is important to remember how critical Android was for Operators to have a competitive response to AT&T which had the exclusive on the iPhone. Verizon particularly rode the Droid horse hard until they gained access to the iPhone.

It is also important to note that Google’s Android play has always been focused on their core business model – increasing how much time each of us spends online, with Google providing web-based services and enabling monetization by 3rd party developers that ultimately drive advertising dollars for the company. (Advertising represented $59B of their $66B in 2014 revenues.)

Nexus
In January 2010, Google partnered with HTC to launch the Nexus One smartphone running the latest release of Android. The phone introduced some new features, but mostly it was an attempt by Google to demonstrate how strong a “pure Google” device could be. At least to some extent, it was an attempt to get the OEMs to stop modifying the Android platform. As you may recall, at the time, there was a fair amount of noise in the marketplace about fragmentation in Android (multiple operating system versions, different screen sizes, user interfaces, etc.) relative to the monolithic iPhone.

With the Nexus One, Google also tried to introduce a new approach to the market, selling an unlocked phone at full price, only available for purchase via a website, and with customer service only available via online support forums. None of these experiments were successful and undoubtedly contributed to the lack of success for the phone itself.

The second Nexus handset, the Nexus S (based on Samsung’s Galaxy S platform) was more successful. It introduced the Gingerbread version of Android (2.3) and had hardware specs that were impressive, including NFC. In fact, the Sprint version of the Nexus S became the launch device for Google Wallet. For this second Nexus device, Google stepped back from selling only on the web, selling as a full price unlocked device, and providing support through forums. Instead, they adopted the traditional industry models – sales and support primarily through the Mobile Operator channels.

Google has continued to partner with OEMs to introduce new Nexus phones, often using each new model as an opportunity to introduce new capabilities that perhaps the OEMs and Operators weren’t yet ready to place a bet on otherwise. It’s important to note that Google had to work hard to make sure that this program didn’t alienate the OEMs and Operators on whom the company was dependent. With each Nexus, Google partnered with a different OEM, and made sure that versions were available for the major operators.

To some extent, Google has used the Nexus devices to continue to push openness and capabilities that can enable mobile devices to be used for more and more applications, ultimately driving their core business.

Google Fiber
On February 10, 2010, Google announced plans to build an experimental fiber network, delivering 1GBPS, which they characterized as “100 times faster than what most Americans have access to today”. In their press release, they said “We’ve urged the FCC to look at new and creative ways to get there in its National Broadband Plan – and today we’re announcing an experiment of our own.”

As with Nexus, they made a big deal about the scale being not too small and not too big, saying that they would deliver the service to as few as 50,000 and as many as 500,000 people. They said their goal “is to experiment with new ways to help make Internet access better and faster for everyone” and they specifically called out enabling developers to come up with next generation apps, test new deployment techniques that they would share with the world, and provide openness and choice, managing the network in an open, non-discriminatory, transparent way and giving users a choice of multiple service providers.

They seemed (at least initially) to not want to offend existing broadband providers, saying “Network providers are making real progress to expand and improve high-speed Internet access, but there’s still more to be done. We don’t think we have all the answers – but through our trial, we hope to make a meaningful contribution to the shared goal of delivering faster and better Internet for everyone.”

With that initial announcement, they invited communities to express interest and more than 1000 did, with many doing crazy things to try to win the network for their community. I live in the Kansas City area (the winning city), and although Google Fiber is not yet available in my neighborhood, it has been a big catalyst for innovation across the metro area.

As has been well documented, Google’s entry into broadband also forced the existing broadband providers to improve their offers (speed, capabilities, and/or price). As Google Fiber has pushed into new neighborhoods and suburbs, the competitors have had to respond. Google is coming to my neighborhood this year and that has caused AT&T to expedite construction on their GigaPower infrastructure and for Time Warner to build out outdoor WiFi using streetlight mounted antennas. Everyone is offering special deals with multi-year commitments. We’ve seen similar competitive responses as Google has announced Fiber projects in additional cities.

Of course, Google Fiber is no longer a friendly, sub-scale experiment intended to help the broadband providers. In December 2012, Eric Schmidt said “It’s actually not an experiment; we’re actually running it as a business,” and he announced expansion to additional cities.

As with Google’s other telecom initiatives, the primary focus continues to be the core business. Google Fiber, both directly and indirectly, is driving more overall Internet use, and that helps drive Google’s services and advertising revenues. It’s also important to note that Google has traditionally not had a strong relationship with broadband providers, so they likely felt free to take a more disruptive approach to the market than with Android and Nexus.

In my next post, we’ll take this historical perspective, combined with Pichai’s comments, and combined with an understanding of the challenges that MVNOs traditionally face, and try to speculate on what a Google MVNO might actually look like.