Wednesday, July 15, 2015

Death of Subsidy: Winners and Losers

Friends, Americans, countrymen, lend me your ears; I am here to bury device subsidy, not to praise it!

Alright, burying is probably a little premature, but, clearly, device subsidy (in the form of 2-year contracts) is on its death bed. Ever since T-Mobile pioneered equipment installment plans in the US in 2013, Americans started going the way of the rest of the world. Today, equipment installment plans are the norm for the majority of customers upgrading their phones. So, how does it change the mobile industry landscape and who does it benefit most?

What Was Good about Device Subsidy?


Before we jump to how the landscape has changed, I want to give a little thought to what was good about device subsidy. (Even though it's not dead, I will stick to the principle of not speaking ill of the dead. ;) )

If you agree that supplier competition is good for the consumer, then, more choice in original equipment manufacturers (OEMs such as Samsung, Apple, LG, Motorola etc) and more operating systems (OS) is better for the consumer. Device subsidy allowed the operators to level the playing field somewhat for the smaller OEMs and less adopted OSs by offering more subsidy to the devices offered by the weaker players. Why would operators be interested in doing this? Higher diversity in OEMs and OSs increases operators' leverage and helps them differentiate themselves from others. So, the interests of the consumer and the operators are well aligned in this case.

Obviously, life wasn't perfect even with this benefit of device subsidy. Operators vied with each other to carry the most popular phone brand, ever since it was first released in 2007. Operators were willing to provide more subsidy for devices from Apple that was able to bring in higher value customers. This clearly tilted the playing field against some of the same weaker players that the device subsidy should have helped.

The other benefit of device subsidy was that it sped up the technology adoption cycle for the mobile industry. With subsidy, most customers upgraded like clock work every two years. So, OEMs and other players could plan for the demand for new technology based on this upgrade cycle. This was also good for mobile operators in managing their networks since transition between technologies (2G to 3G to 4G) could be managed smoothly with device technology transition via the upgrade cycle. Unfortunately, this resulted in higher costs for the consumers because the costs of upgrades were ultimately borne by them.

So, How Might Future be Different?


I see three main ways in which future could be different from today - likely lengthening of phone upgrade cycle, a focus on long term value of device and potential play by OEMs to get closer to the customer with their own equipment installment plans.

Lengthening of Upgrade Cycle, Especially in the Premium Segment


With the consumers directly responsible and cognizant of phone pricing, some have theorized and I tend to agree that the consumers are likely to upgrade less frequently. The evidence for this theory till now has been scant. Driven by their sudden ability to upgrade on a whim, consumers have actually pushed up the upgrade rates at most operators since the introduction of the equipment installment plans. But, I would hypothesize that the consumers who want to upgrade more frequently (than every two years) are a small fraction of the total consumers.

With AT&T's completion of two years of its launch of Next (equipment financing plan) program and high installed base of iPhones, a large majority of its customer base will be off their 2-year contracts. If the hypothesis is correct, I would expect to start seeing a slow down in upgrade rate in AT&T's Q2 results.

One possibility is that the upgrade cycle will be different between the premium segment of phones such as iPhone 6 or Samsung Galaxy S6 and lower segment phones driven by the difference in the monthly cost of ownership between the two tiers. At higher monthly costs, consumers might be more sensitive to upgrades than otherwise. This would be contrary to the past where the premium segment upgraded much more frequently.

Focus on the Long Term Value of Device to the Consumer


Reduction in upgrade rates will spur some of the OEMs to provide a device with a better long term value for the money. But, unfortunately, not all OEMs will be able to play this game. Apple, with its tight coupling of software and hardware, is much better placed in this aspect. We can already see this in Apple's plans for iOS 9 which will be compatible all the way back to iPhone 4S. At the time of the likely official release of iOS 9, iPhone 4S will be four years old. Assuming that Apple would want the device to work for at least another year, the life of iPhone 4S will be almost five years.

Now, one might wonder if this would be a good strategy, even for Apple! With a saturating smartphone market, a longer upgrade cycle will decrease the volumes even further. Apple is already preparing for this eventuality by pushing harder into services such as music and news. Even though the customer might keep the device longer, app and subscription sales will provide a steady stream of revenue that, hopefully, will compensate for revenue lost due to lower device sales.

Another way OEMs are likely to demonstrate long term value of the device is by supporting the prices in the secondary markets. We have seen this in the automotive industry for a number of years whereby some of the premium car manufacturers have participated in the secondary market by acquiring customers' vehicles and offering used certified vehicles. Apple, again, has shown early signs of intervening in the secondary markets and we should expect much more of this behavior in the future.

Equipment Installment Plan as a Means to Enhance Customer Relationship


One another advantage due of the proliferation of equipment installment plans and the gradual death of device subsidy is the ability of customers to bring their own devices. Clearly, consumers have been taking advantage of this - AT&T announced in its Q1 results that 313,000 of its gross adds were bring-your-own-devices. This provides an avenue for anyone that desires a closer relationship with the mobile customer to realize it through their own equipment installment plan.

Traditionally, Android OEMs, despite their best attempts, have had tenuous relationship with their customers because the customers identified themselves first and foremost to the App Store with their google ID. Reasons for signing up for an account with the OEMs weren't very strong. Equipment installment plans, if offered by OEMs, might now provide a stronger reason to be truly contract free. Motorola has already gone down this path. Using this financing relationship, OEMs could potentially understand their customer needs better and drive better customer retention. Again, Apple is ahead of its competition, even without offering equipment installment plans of its own because of its App Store and OS relationship with the customer.

So, What Does This All Mean?


Despite potentially being at a higher risk due to lengthening of upgrade cycle, Apple is likely the best prepared and positioned to take advantage of the future with equipment installment plans. While other OEMs might be able to replicate some aspects of Apple's strategy, they would do so without the cushion of services revenues (that Google is likely to claim in the Android ecosystem) and the strong customer relationship Apple has enjoyed. The net result is likely a continued race to the bottom in the Android ecosystem while Apple continues its domination of the higher end. Apple will continue to expand the reach of the higher end by selling its older generation devices to those that can't afford the latest devices and by improving the long term value of the devices with OS support and secondary market intervention. Equipment installment plans are likely to strengthen status quo in favor of Apple!

Tuesday, June 16, 2015

Project Fi: What Might Google's Long Term Plans be?

Ever since Google announced Project Fi, many have wondered about Google's motivations. Google mentioned wanting to make WiFi to cellular hand-off seamless. But, why does Google want to make the WiFi to cellular hand-off seamless?

Driving the Mobile Data Costs Down


The ultimate goal of almost all projects that Google takes up  is improvement in its ability to target search ads. Project Fi is likely no different. I believe Project Fi is likely trying to achieve that goal by driving the mobile data costs down and thereby, increasing usage of mobile devices and Google's knowledge of the customer where ever they happen to be.

The obvious way Project Fi might be trying to drive the data costs down is by upending the relationship mobile operators have with their customers. With Project Fi, Google has been able to insert itself between the mobile operators and the customers, albeit in the form of a MVNO (Mobile Virtual Network Operator - for example. US Cellular). As a customer, you now buy your phone and service from Google. How Google gets you the service is immaterial to you as long as the network coverage is good where it matters to you and the price is right.

So, how is Google delivering the service with Project Fi? Google intends to keep the consumer device primarily on the open WiFi hotspots or ones that you have access to. It has developed a  database of open WiFi hotspots around the world in the course of its Google Maps efforts. It is now putting that database to work to ensure that the mobile device stays in a WiFi hotspot most of the time. When there's no available WiFi hotspot (or if the available connection doesn't support quality of service necessary to support a voice call), T-Mobile and Sprint networks will serve as backups. With eSIM (embedded SIM that is programmable), Google is able to query the two mobile networks and find the one with the best signal (or probably just the cheaper) network.


Suddenly, T-Mobile and Sprint (the MNOs - Mobile Network Operators) aren't relevant to you, the customer, any more because of this arrangement. While T-Mobile and Sprint are happy now to get the wholesale business from Google, Google has the opportunity to negotiate lower rates, if volume picks up in the future. (Under supposed terms of service, T-Mobile and Sprint also have the power to renegotiate the rates. But, really, are the rates going to go up with higher volumes?!)

So, is this the extent Google can push the game?

I think not. There are at least two other ways Google is likely to continue to pursue reducing the mobile data costs. The first avenue Google can pursue is by sharing the WiFi-cellular transition technology with MVNOs to reduce their costs, without itself putting up capital to run a large retail MVNO operation. Large amount of bandwidth available for WiFi could make spectrum a commodity. The second avenue is by using Project Loon to reduce the leverage of MNOs in rural areas where low-band spectrum (600, 700, and 800 MHz) has been essential for coverage.

Driving Down the Costs for Other MVNOs with Call Transition Technology


As I mentioned before, Google's stated objective is to make the call transition from cellular to WiFi and back seamless. While some MVNOs such as Republic Wireless, FreedomPop etc have offered similar service to Project Fi's, the user experience during transition is less than ideal - call drops are frequent.

We can safely assume that Google is better positioned to accomplish this task than Republic Wireless, FreedomPop etc. Ensuring seamless transition likely needs changes to WiFi standard to allow low latency transitions, among other problems that need to be solved for a seamless transition. Google is likely in a much better situation to push the changes needed than a small operator such as FreedomPop.

Once Google is able to perfect the network technology necessary to ensure seamless transitions, Google plans to share the technology with the mobile operators. Thus, all operators (MNOs and MVNOS) will have access to hundreds of MHz of bandwidth with just the current WiFi bands in 2.4GHz and 5GHz. Given that the heaviest current data use typically happens to be at locations where customers send most of their time (homes, offices,coffee shops, malls etc), and for a lot of users, these locations are also likely to have WiFi, MVNOs will be able to offload most of the traffic for which they would otherwise have had to pay the MNOs. This will drive down the costs for the MVNOs which are likely to pass those savings to the customers.

Using MVNOs, Google can scale up the use of call transition technology and WiFi without putting its own capital at risk.

More WiFi Bandwidth Could Weaken the Spectrum Advantage of MNOs


Current allowed bands (2.4GHz and 5GHz) have a total (shared) bandwidth that is comparable to the total spectrum holdings of all the top tier MNOs put together. Proposals under FCC consideration to include 3.6GHz and 4.9GHz bands will increase the available bandwidth for WiFi even more. Armed with WiFi transition technology, MVNOs can hope to stay in WiFi most of the time while ensuring quality of service while the customers are even less likely to encounter bandwidth issues in the future.

In this scenario, bandwidth becomes less of an asset to the MNOs and more of a commodity. (Like land, there is little new spectrum available - so, it can never really become a commodity, but, in urban areas with high WiFi coverage, it will be close to one.) MNVOs will have the opportunity to compete with the MNOs on more even terms and go up the value chain. This has the potential to increase the competition even more today and drive the prices and margins down further.


Project Loon Could Further Reduce MNOs' Leverage


Google has been experimenting with providing internet from hot air balloons aloft in the air. Using solar power and taking advantage of wind variations with altitude, Google has been able to keep its balloons up for 100 days continuously. These balloons are akin to mini-satellites that can provide WiFi/LTE coverage. One very good way to maximize the potential of Project Loon is to use free spectrum (such as WiFi or ISM bands) and high band spectrum (that is abundant at Sprint and DISH) to provide broadband internet service.

Till now, rural cellular coverage required low band (600, 700 and 800 MHz) spectrum because of the high density of cell towers necessary for high band spectrum. With Project Loon, coverage can be provided on demand using free/ high band spectrum. While the coverage still won't rival that of low band spectrum, the quality gap will certainly narrow. This again can lead to commoditization of spectrum.

Other Benefits for Google


Are there other benefits for Google? The short answer is more information on mobile customers. With Project Fi, it will "own" the customer more than ever and understand even more of what they do - call details, apps used and usage details etc. When Google "shares" the call transition with the MVNOs/ MNOs, one has to assume that the obvious barter is for the operators to share the same information that Google now has access to with Project Fi. More customer information, better targeting of ads!


So, What Does This All Mean?


Google's ultimate goal with Project Fi is likely to drive the cost of mobile usage down, not running its own MVNO. Project Fi and Project Loon have the potential to strike at the very base of MNOs' value pyramid and commoditize spectrum, thereby, reducing the cost of mobile usage. The MNOs and regulatory bodies (like FCC) will be faced with choices that are very different from today. MNOs will be challenged to show differentiation in service more than ever before.


Regulatory bodies around the world will be more challenged than ever with balancing revenue goals and providing free spectrum for WiFi. Also, if the WiFi spectrum is meant to allow low cost internet service, should the MNOs be able to charge for unlicensed use of the spectrum for WiFi? If so, what's a reasonable amount?