Blog Post

MuniFi: Build it and they still don’t come?

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

MuniFi has now been live long enough in many cities to get some feedback. Like we predicted, this year many of these bigger networks will now be put to the test.

And well, certain initial findings are a little troubling when it comes to consumer use of some city-wide Wi-Fi networks.

The Chronicle took a trip to Taipei to check out the 15-month old, Wi-Fi network run by Q-Ware, which is one of the world’s largest. Their assessment:

The city has struggled to get subscribers to sign up for the service . . . due to some perceived performance issues, competition from free hotspots and a lack of applications.

Q-Ware told the Chronicle the network had brought in an average of 20,000 monthly subscribers ($12 a month) and about 10,000 subscribers with daily, weekly or monthly one-time passes. Those figures might sound high for some networks, but Q-Ware’s $30 million investment (some say the costs are higher) was supposed to bring in an initial projection of 250,000 average users by the end of last year, and 200,000 average users to break even.

The New York Times took a look at the network last June, and found “just 40,000 of Taipei’s 2.6 million residents have agreed to pay for the service since January.” Does that mean the subscriber numbers have actually dropped over the past several months? Uh-oh.

While small in comparison to Taipei, the Lompoc (Calif.) city Wi-Fi network is causing some concern with a lack of interest from residents, according to The Lompoc Record. The article published Monday says:

Weak signals, which delayed the city’s WiFi system for five months, are still plaguing the $3 million Lompoc Net system, although city officials say the customer base is beginning to grow.

The article says the network went live last September, and brought in 137 customers, and now there are just 281 customers — far less than the city needs to make the service financially viable.

Portland’s MetroFi network seems to be having its own problems with coverage according to an initial study by an independent team that also works with Portland’s Personal Telco Project (update: the authors of the study say their work was not officially part of the Personal Telco Project). I asked Earthlink for its latest subscriber numbers of its live networks, and the company said they are not providing those figures right now.

A major problem is the one that Wi-Fi consultant Craig Settles pointed out to us in January, that “public access of city-wide Wi-Fi networks will be widely viewed as financially the weakest pillar in the business case for municipal wireless,” by the end of 2007. Instead, mobile workforce applications will be muni networks’ big ROI generator, he says.

It looks like this situation is playing out for some of these networks. Cities that rely mainly on consumer use to sustain the network, might want to rethink things.

The gap is also in perception of what these networks can be used for. If consumers are looking for guaranteed 3G style coverage outside, as well as indoor coverage without extra hardware, they will be disappointed. If they’re looking for blazing fiber-fast bandwidth over Wi-Fi, they probably won’t get that on most MuniFi networks either.

Note to network builders: be truthful or risk consumer backlash.

23 Responses to “MuniFi: Build it and they still don’t come?”

  1. FYI.

    Kite Networks has already spun off from Mobilepro, however, Mobilepro sold Kite to Gobility, who has no money. They cannot even make payroll. Now, the employees of Kite Networks will be filing a complaint with the Department of Labor, which will most likely end up causing Mobilepro some legal issues due to their fraudulent transaction to Gobility.

  2. There’s an error in the post above: the analysis of Portland’s MetroFi network was done by Caleb Phillips and Russell Senior, independent of their work with Personal Telco (it was misreported in the WWeek article). Details are at

    The perception that the municipal network here doesn’t work comes from the difficulty of even signing on to the network, with reports from people throughout the demo areas. We’re not expecting a lot, just a base level of usability.

  3. Katie: From MetroFi’s perspective, the most important metrics are utilization reporting – how many people are using the network and for how long. With reference to our network in Portland, the actual usage of the network since its launch in December 2006 has tripled. In March alone, over 5,800 users were on our network spending over 50,000 hours online. And as more and more Portland neighborhoods go live, we expect the utilization numbers to continue to increase.

  4. Issues to address:
    1. Muni (and their consultants) selecting the Provider and not considering the Technology.No excuse after all these years.
    2. Lack of pre-award trial with a side by side by two top choices of systems. The good consultants can make this a fair test if not an independant test firm should be used to make it clean.
    3. Lack of a performance Bond requirements in the RFP, which will quickly qualify the Providers to make sure they pick the correct systems and not the big name (Providers and Vendors). These bonds (effectively an insurance policies for the Muni) will require the Provider to deliver a minimum bandwidth levels per Node/AP and latency targets between Nodes for the term of the contract. This should be required regardless of whether the provider is building a “Free Network” for the Muni.
    4. Wireless Gateway connections, as in using a Canopy Link to refresh the remote Mesh Node will not suffice to maintain the bandwidth levels required of the new portable user community. 10-20 Mbps bump to a Node every 3-4 nodes is not economical nor functional. These networks really do need 100Mbps links that can best be delivered over a fiber link or select PTP radios. Of course this assumes the single or dual radio node has the ability to use the 100Mbps.

    These nets need to be designed (day one) to deliver serious Broadband services such as Audio/Photo/Video sharing (symmetrical) services as well as MultiPlayer Gaming (P2P) and VoiceIP. The days of just surfing the Internet and Email checks is gone and should be replaced by robust systems that can be upgraded (in the field)to add in new radios (802.11n, WiMAX and 4.9Ghz)and other capabilities that extend reach and bandwidth.
    Wait until the Police department wants to add 30 Video Surveillance cameras to these Mesh Networks for backhaul to their Center.

    WIreless Mesh Networks if designed properly and based on the correct wireless technologies (multiple 4+ radios)will be dominant in delivering links in Portable and slow Mobile Broadband Metro Area services. They will also be a compliment (thru Fixed Mobile Convergence) to both the Narrowband Cellular Mobile data networks as well as emerging WiMAX nets.


  5. alan schneider

    has anyone looked at the small towns’ results? i’m thinking about (for instance) towns under 5k population with REA electrical co-ops providing the service.
    i don’t know where in europe esme vos is talking about, but during my recent visit to family in belgium, i saw most folks very satisfied with service most of us here in the US would have us jumping down a WIRED phone to cance service!

  6. Most cities and counties realize that to get as much of a return on their investment, they need to make maximum use of the network: public Internet access, public safety, automated meter reading, traffic management, parking control, enabling municipal employees to work from the field, etc.

    I’ve been running since June 2003 and the RFPs I used to see three years ago asked for single use networks – either municipal use or public access ONLY. But now, they know that these networks can be used for many things and they are asking for that.

    As for “quality” of service, cities and their partners (the service providers) are now doing more pre- and post-installation testing. Still if you do a “test” today, you are also just taking a snapshot of what’s happening in one day.

    People are not after perfection. Just look at the mobile phone networks in the US. Dreadful coverage, dropped calls. I am totally appalled (as Om is) with the “quality” of mobile phone service compared to Europe. But I have a T-Mobile prepaid subscription which I do use when I am in SF.

    So we’re back to weighing the benefits against the costs/pain. I keep thinking of my friend who incurred a 10,000 EUR mobile phone bill because she uploaded a lot of video from her mobile phone to her blog. If she had access to a wide-area Wi-Fi network she would have used her Wi-Fi enabled Nokia phone to do that. If you had a choice between a 10,000 EUR bill and a bit of inconvenience (not so perfect network, slightly faster than HSDPA upload speed), yeah, I’d go with Wi-Fi.

  7. Larry McNeill

    I don’t think it really matters how good a service you supply, you will have people trying to find something to gripe about. What we really need are for some of the people who are happy with it to also post comments. Right now all we see are negative comments and many of those are sour grapes from people who didn’t get any particular contract. I think the report that slammed MetroFi i n Portland was done by a group who actually lost the contract to do the testing, so they may be a little jaded. We also see a tendency for companies that do a deployment to tell raving stories of success and then you see a few reports come out that say not all is as cherry red as they say, such as the St Cloud project. Bottom line, we need feedback from the end users in the areas the deployments are taking place.

  8. It’s still premature to deem a network as a success or failure. While free access may be available in Portland today, the network hasn’t even been officially launched yet. The public report on coverage was unofficial.

    EarthLink still doesn’t have any of their networks covering an entire city (maybe Milpitas but I’m not privy to that info) so you still have to cut them some slack on revenue figures.

    MobilePro publicly released information about the take rates they were getting in Tempe – several consecutive months of double digit growth. When Kite Networks spins off of MobilePro we may get to see those numbers.

    Applications are just starting to be developed which will contribute to significant revenue generation. I’ve seen some of them (some really cool ones are still in development so I cannot speak about them on a public forum.)

    Ad-subsidized models are still in the early stages of development, but they will undoubtedly deliver enough revenues to support a network (my opinion).

    CPE devices for extended range are finally coming available – and are getting better.

    It’s very easy to slam any company in Muni WiFi at this stage. It’s just as easy to slam EVDO, HSDPA, WiBro, and WiMAX right now. Of course, this is all just my personal opinion.

  9. Alex Witherow

    Another thing to look at is the node density of these cities. Wi-Fi requires heavy node counts for it to be even remotely functional in an outdoor environment. The higher the node count, the more subscribers are needed to break even, which will challenge this city and others…

  10. OK, here is a plug on a comment Katie made in her blog about Craig Settles: “mobile workforce applications will be muni networks’ big ROI generator”.

    Oregon is thinking about how mobile Internet is an economic driver now and in the future. You can learn more at, and if you can make it…please come to Oregon!