Internet Everywhere!

It appears to me as if Google is throwing things at a wall to see what sticks. Perhaps it would be better off to put together partnerships with other players.

Google is rapidly moving in many directions to provide ubiquitous Internet worldwide. First it was muni-Wi-Fi in its own backyard, then fiber to the home in Kansas City, and soon, according to Google, in up to 34 more cities. The caveat here is that the cities have to provide Google with infrastructure blueprints and a streamlined construction process to scope out construction costs. The fiber to the home battle will be interesting to watch as Verizon, AT&T, Google, and the cable TV companies circle around the issue of construction cost per house.

Next comes the Google Loon project where hot air balloons are launched and supposed to be steerable—or at least follow predictable routes to provide Internet to those within the balloons’ wireless footprint. Then to augment the Loons, drones will be used as gap fillers. And finally, the latest in the Google arsenal of wireless Internet offerings is the recently announced low-earth orbiting satellite (LEO) system.

Before looking at what Google is planning here, let’s take a stroll back in time to 1995 when Bill Gates, Paul Allen, Craig McCaw, and Saudi Prince Awaked bin Talal formed a company called Teledesic. The company’s plans in 1995 were to invest a total of $9 billion and launch 840 low-earth orbiting satellites to provide worldwide Internet coverage with download speeds of up to 100 Mbps and upload speeds of 720 Mbps. By 1997, the number of satellites to be launched had been reduced to only 288 but at higher altitudes. The company went out of business in 2002 with no satellites launched and no income.

Okay, today times are different, satellite technology is much more advanced and less expensive, launch vehicles are more readily available, and Google certainly has the resources to make something like this work. However, I have to wonder whether there is a sustainable business model in this. Suppose Google spends $5 or $6 billion to put the satellites into orbit. It will need ground stations and all of the other backend equipment necessary for multiple connections to the Internet. Coverage will be outdoor only or require a roof-mounted antenna for inbuilding service, and may or may not be available for mobile devices depending on the system design and what can be built into the mobile devices, what they will need in the way of antennas, and other related factors. Where there is no cellular fourth-generation LTE broadband the Google satellite service will be welcomed for sure. However, in most of the rural areas currently unserved by LTE, residents and business owners may not be the most affluent and therefore, depending on the monthly price, might be very slow to commit to the service.

In areas where LTE is available, chances are pretty good that there will also be cable or fiber to the home, plenty of Wi-Fi, free and otherwise, and at least three or four connectivity options. In order to gain acceptance in these markets, the Google satellite service will have to:

1)     Be price competitive
2)     Provide low-latency access to the Internet
3)     Be available 24X7
4)     Have provisions for inbuilding access of some sort

When Teledesic was first conceived, the land mass not covered by cellular systems was much greater than today. Since the early 2000s, terrestrial wireless networks have expanded to cover many areas around the world that were never covered before. This trend will continue and with the advent of Public Safety broadband systems being built around the world, wireless terrestrial broadband coverage will continue to grow and provide coverage to more of the world’s rural and remote areas. Certainly there will always be portions of the world that are not covered by terrestrial wireless but there is still the question of whether covering these areas will provide Google with sufficient income to provide a decent return on its investment.

Of course, Google’s business model has never been a “normal” business model since it is based on the number of eyeballs it can attract and how much they are worth to Google’s advertisers. But will that model sustain itself when the eyeballs it is chasing are living in rural and remotes areas of the world? Assuming it won’t, I have to wonder what is motivating Google to push into all of these wireless Internet delivery methods at the same time. Finding radio spectrum to use for broadband within the confines of the United States is difficult enough; finding spectrum to use worldwide has to be even tougher. Yes, satellite spectrum is coordinated on a worldwide basis, but then Google will have to buy an existing satellite provider or partner with one or more providers. As it turns out, that is exactly what Google has done, announcing today that will purchase Skybox imaging for $500 million. Skybox is a satellite start-up that is building small satellites for imaging purposes so theoretically these little birds could provide a new imaging source for Google Maps as well as wireless Internet to the underserved.

From muni-Wi-Fi through TV White Space to balloons, drones, and now little LEO satellites, it appears to me as if Google is throwing things at a wall to see what sticks. Perhaps it would be better off to put together partnerships with other players. I have on several occasions written about providing wireless connectivity and Internet access in rural areas both in the United States and around the world. In each of these articles I have stated that if providing these services is a serious goal, then one way to approach the problem is to enlist partners in each area that needs to be served.

There will be different partners for different portions of the world. One of the easiest places to provide rural access is in the United States and there are plenty of partners available that are ready and willing to work with Google or some other group of partners to make it happen. For example, the FirstNet Nationwide Public Safety Broadband Network (NPSBN) is required by law to cover deep into rural America. This will be accomplished by extending terrestrial networks into rural areas and also by deploying satellite communications into areas that are difficult to cover.

FirstNet is permitted to team with partners that will assist in building out the network and then share the FirstNet spectrum on a secondary basis. Rural network operators represented by the Competitive Carriers Association (CCA), rural power companies, tribal entities, and others are all interested in taking part in building out rural America with FirstNet. All it would take is for Google to stand up and put together a group of interested partners, making use of their financial might, perhaps their existing fiber assets, and their willingness to work with others. Rural America could be blanketed with LTE and satellite broadband fairly easily. With the number of potential partners sitting in the wings waiting for someone to take the lead, it could also be done fairly quickly.

Each country around the world offers some type of potential partnerships for Google. All Google has to do is to have a mindset that it is willing to work with partners in order to provide Internet services. Deals can be made, investments defined, and wireless coverage provided to benefit those in areas with no Internet access today. Google and its partners can make money by providing access and services, and because there would be local partners in each area, it should be easier to work with regulators and others to accomplish their goals.

Yet Google seems bent on going it alone. It is a big company for sure but every company has finite resources even if they sometimes appear to be inexhaustible. To provide Internet access to every citizen in every country in the world, even using satellites, is a massive undertaking. Even though Teledesic failed in the 1990s, it started out right by bringing in a number of different partners, sharing the costs, and sharing the rewards if any. In the case of Teledesic, it was simply too early, but that has been the fate of a number of wireless ventures.

How many of you remember Metricom, which was located in Los Gatos, CA, just outside San Jose? Metricom built and deployed the world’s first muni-Wi-Fi system and deployed it in a number of cities. It battled for permission to mount its boxes on street lights and take its power from the street lights, and it battled the permitting process and citizens who did not like the look of the boxes, but it did provide wireless Internet very early on to a number of cities and college campuses. The company was in business from 1985 and finally shut down in 2001 after a number of setbacks. The founders lost control of Metricom to Paul Allen (co-founder of Microsoft) in 1997 but had pioneered the muni-Wi-Fi concept. Its Ricochet product was available in more than fourteen major cities and surrounding suburbs and at one point it boasted more than 51,000 subscribers.

Also early to the party, Metricom broke the ground that has been plowed many times since it showed the way. Even today I can still spot Ricochet devices hanging from some street lights where they have not been removed. During Metricom’s heyday I lived in the mountains on the Santa Cruz side of Silicon Valley and owned a communications site at the top of the mountain range. We set up a Metricom device on my tower and I ended up with wireless Internet access from my mountain home, where until then I only had dial-up phone access. It worked and it worked well. Metricom’s problem, other than being too early in the development of wireless, was that it did not offer to partner with the cities where it was installing equipment. Had it done so, it perhaps would still be around providing broadband speed access instead of the 128 Kbps data rates we were experiencing then.

The wireless roadmap is littered with companies that were before their times but whose ideas have been dusted off, redesigned, and put back into play. But even the largest of the large network operators turn to partnerships in order to provide extended coverage. For years, Sprint entered into partnerships with operators in smaller areas because it did not have the resources to provide coverage. These partnerships, for the most part, worked out well. The local company built and operated the network, sold Sprint devices, made use of the Sprint spectrum, and paid Sprint a percentage of the income derived from the partnership. Some of these deals were better than others but they all worked. More recently, Verizon has entered into a number of partnership agreements with smaller network operators in order to provide LTE coverage beyond where Verizon felt it was economically feasible to do so directly. This too has worked out well for both Verizon and its partners.

Iridium, the satellite cellular company built by Motorola, was a partnership of many different companies and several countries. It too was way early and experienced some rough times but it has been reborn and is still in business serving mostly government and maritime users that need voice and slow-speed data connectivity regardless of where they are. Partnerships share the cost and share the profit or the loss. In either case, the success or failure of a venture does not necessarily mean the success or failure of a single partner.

There have, to be sure, been some massive partnership failures. Microsoft and Qualcomm’s joint venture known as Wireless Knowledge, LLC, that was created to provide wireless connectivity to corporations was one example. The company was formed in 1996 and Qualcomm bought Microsoft out in 2001, then shut the company down in 2003. There were also several partnerships between Sprint and the cable companies; I say several because Sprint started life in the wireless space during the 1996 spectrum auctions with cable companies as its partners. Later, Sprint again formed partnerships with several cable operators and then again a few years later, each of these partnerships fell apart. Clearwire was also a partnership. While Craig McCaw was the mastermind behind the then WiMAX broadband play, the shareholders included Sprint, Comcast, Time Warner, Bright House Networks, Google, and Intel. Today Clearwire is simply part of Sprint.

Forming partnerships does not always work, but I don’t see how any single company can cover the world with wireless access. Perhaps Google can but I, for one, won’t bet on it.

Andrew Seybold