Lowenstein's View: 'Wi-Fi First' - What still needs to happen

Mark Lowenstein

With the continued growth in demand for data and proliferation of private and public Wi-Fi hotspots, the idea of a 'Wi-Fi First' wireless service has become a favorite discussion topic. Republic Wireless and Scratch Wireless are two MVNOs that are already offering a Wi-First (my term) service, using the Sprint network as cellular backup when not within Wi-Fi coverage. The leading cable MSOs, who have formed the Cable WiFi Alliance, have deployed hundreds of thousands hotspots, and are turning millions of home routers into "neighborhood hotspots" by issuing a second, "public' SSID. Many believe that the cable companies will eventually offer a Wi-First service as a value-add to their broadband customers and as a way of competing with cellular.

I think Wi-Fi will continue to play a significant role in our communications infrastructure. There will be continued proliferation of public hotspots, more free or cheap Wi-Fi in cities, more spectrum for Wi-Fi courtesy of the FCC's recent 5 GHz ruling, and even creative proposals for a 'premium Wi-Fi' service using Channel 14 of the 2.4 GHz spectrum as proposed by Globalstar in its TLPS petition. This is all great for Wi-Fi, great for Ruckus, Cisco, et al, and great for consumers who are looking to offload expensive cellular data services.  

It is technically possible to offer a Wi-First service. When the Wi-Fi signal is good and the AP not overburdened, voice quality is adequate and data speeds are competitive. The density of available APs continues to improve, particularly in markets where the cable companies have invested aggressively. Republic Wireless and Scratch Wireless have done a nice job of building the infrastructure so that a full suite of voice and texting services and most value-added features can be delivered when connected to Wi-Fi. The texting piece is especially significant, since it is the "backup" communication mechanism in cases where the user is not connected to Wi-Fi but has not bought cellular minutes or bytes.

However, I think there is a long way to go before Wi-First can be considered as a viable, mainstream alternative to a cellular plan. Although it might be workable for specific types of usage profiles, or for those who are very price-sensitive, there are still a lot of issues that have to be addressed in order for Wi-First to be a competitive offering.


My experience so far is that the user still has to do a lot of work, such as choosing Wi-Fi or cellular "modes". The UI on some of the initial Wi-First phones is a bit clunky and non-intuitive, with difficult to read screens and icons, and, sometimes, lots of steps involved in set-up and configuration. In an ideal world, the service would move more seamlessly between Wi-Fi and cellular networks, choosing the 'best connected' scenario for a given context, in a way that is invisible to the user.

Second, because of the limited range and density of hotspots, any user who is not stationary is going to find themselves quickly out of Wi-Fi mode. Today, there is not seamless handoff between Wi-Fi and cellular. There is a lot of work going on in this area, and I believe that the cable companies will not launch a Wi-First service until movement between Wi-Fi and cellular is smoother.

Third, we have to consider density and demographics in the U.S. The "home as neighborhood hotspot" is a great concept, but much of the U.S. does not have sufficient population or housing density where the AP would effectively 'broadcast' enough signal to be picked up by other users. Outside dense urban areas or tightly packed residential suburbs or MDUs, usage would more naturally tilt toward cellular 'mode' when not in the home or office.


The initial Wi-First services work only off specially outfitted Android devices supplied by Republic or Scratch. There is no iPhone option. I can't help but think that eventually, that Wi-First would be delivered as an app or as a value-added service, much in the way users download alternative texting services such as WhatsApp. I see Wi-First as more of self-configuring feature, and a way of intelligently managing usage or spending, on any phone, rather than a specific service on a specially adapted device.


Anyone who uses Wi-Fi in public locations, such as cafes, hotel lobbies, trade shows, and so on, knows that the experience can be highly variable and unpredictable. Plus, quality drops off very quickly with distance from the nearest AP. My experience with Scratch's service, for example, was that it is impossible to make a voice call when there are fewer than three 'bars' of Wi-Fi service, or when the local AP is crowded. This is a familiar issue for those who use Skype or other VoIP services on smartphones.

I have also had a very uneven experience with the Cable WiFi APs. As much as 25% of the time, even within range of a hotspot, I get weird "cannot join" or "cannot connect" messages. I have also found the speed to be lacking at many of their hotspots, which affects both voice and data. In at least 70% of the cases, my cellular LTE speed is faster than my Wi-Fi speed at Cable WiFi hotspots. Plus, at least in Comcast's case, the customer service for Wi-Fi needs to be improved. I could not find a way to reach a representative who was knowledgeable, or could help me effectively troubleshoot, issues specific to the Xfinity Hotspot service.

There are also some software issues that need to be addressed. Battery drain on Wi-First mode seems to be faster than on traditional cellular services, which might be attributable to a "no man's land" issue, where the phone is toggling between Wi-Fi and cellular networks.


A "Wi-Fi only" plan makes zero sense. There are too many situations, or the possible emergency, where the user needs to know they can make a call (although 911 is free). This means that any Wi-First user should, or will, have cellular as part of their plan.

Republic, and especially Scratch, rely on reselling cellular voice and data services, at a markup, in order to monetize their free or low-priced service plans. But the cellular part of the plan is expensive. For example, on Scratch, $14.99 per month buys 250 minutes of voice and only 200 MB of data, which means Sprint is probably charging these smaller resellers fairly high wholesale rates. At these prices, cellular can only be used as a last resort backup, especially for data.

This will change, over time. With more spectrum and network capacity, cellular operators will be more encouraging of resale type services, as is the case in some other countries. The cable companies or other larger volume purchasers of capacity could offer more attractive pricing, such as what Wal-Mart has been able to accomplish with StraightTalk. The FCC could even get into the game, as regulators have in some other countries, by requiring more favorable and transparent wholesale and roaming rates, particularly if we're moving toward a triopoly structure. 

5.We Take Cellular for Granted

In my view, the biggest obstacle to Wi-First entering the mainstream is the fact that we use cellular services "behind the scenes" in many more ways than we generally consider. Think of all the navigation services, or apps that somehow use location in the background, for example that need some level of consistent and pervasive connectivity to the network. People are moving around, dipping into this app or that, snacking on various forms of content, without really thinking about it. They might not be consuming a lot of data volume, but they're using the network for a lot more overall time – between voice, text, and data – than they realize. Consider how many times a day you use your phone for less than a minute.

6.Bring Your Own Wi-First Plan

The final issue here is the addressable market for Wi-First. Even with the shift away from subsidies, some two-thirds of consumers in the U.S. are on some sort of sharing or group plan, and/or are tied to their operator through some sort of equipment financing arrangement. There are also some pretty aggressively priced prepaid plans, and the market is getting even more competitive with the national expansion of Metro and Cricket via TMO and AT&T.

We also have to recognize that voice and text has become, essentially, "free" or "unlimited" on wireless. Data is the real currency. The price-sensitive user could build their own "Wi-First" plan without having to purchase a specialized device or service. They could jury rig a Wi-First service by purchasing a cheap prepaid plan with or a plan with a limited amount of data, and be disciplined about using Wi-Fi wherever possible. Wi-First could also become a plan option offered by cellular operators or in prepaid plans, as long as services such as text are delivered regardless of the "mode" one is using. Sprint and TMO are already half way there, with their Wi-Fi Calling features. One could also imagine an OTT-based Wi-First type plan, combining text, VoIP, and some amount of cellular data. This is fertile ground for Facebook/WhatsApp, Google, Skype, and other key players.  

We are in early days of even thinking about the Wi-First idea. Hotspot density is going to continue to improve, and I believe that over time, Wi-Fi and cellular networks are going to be more intertwined. Most of the usability bugaboos are solvable. There is lots of room for creativity in service plan concepts, or even apps, that meld Wi-Fi, small cell, and wide area cellular services.

Mark Lowenstein, a leading industry analyst, consultant, and commentator, is Managing Director of Mobile Ecosystem. Click here to subscribe to his free Lens on Wireless monthly newsletter, or follow him on Twitter at @marklowenstein.