EarthLink announces earnings, changes strategy in muni market

In announcing his company’s earnings today, EarthLink President and CEO Rolla P. Huff said there is “a lot of inherent goodness” in the municipal wireless market but his company is not getting an acceptable return on its investment in it. He says munis working with EarthLink need “to step up and become a meaningful anchor tenant on completion of a build.”Huff made that announcement in EarthLink’s second quarter earnings call this morning, apparently putting his company on a path that is similar to what MetroFi has been pursuing in which anchor tenant commitments from the municipalities help to make the build-out economically for the service provider.

Anchor tenant commitments have worked well in communicities like Riverside, California, where the city partnered with AT&T to provide a variety of city services (for which it pays) as well as a free tier for community residents. Anchorage, Alaska, and Corona, California, however, recently decided to review plans for muni Wi-Fi projects after being unable to reach agreements with MetroFi on anchor tenancy questions.

One thing is clear, however. Muni wireless is not, and never should have been regarded as, a gift horse. Business plans, such as the one which was carefully developed in Toledo, Ohio, can be identified to produce cost savings that insure deployments with minimal financial impact on a city budget. Others with unique demographics, such as Ocean City, New Jersey, can even produce revenue streams for a city. But, however it is done, service providers cannot be expected to bear all the risk and not share in the rewards.

Huff’s remarks today are the first indication of EarthLink’s new direction in its ongoing evaluation of the muni wireless market. At its last quarterly earnings report, EarthLink indicated it was re-evaluating its approach to muni wireless. At that time, I suggested that EarthLink and its fortunes should not be confused with the market itself. Today’s earnings call, along with a new report from Datamonitor predicting tremendous growth in the muni wireless market over the next five years, reinforces that.

Huff, who said he spent 4th of July driving around Anaheim to observe the results of that WiFi deployment, heartily endorsed the future of municipal Wi-Fi while, at the same time, acknowledging his company’s current approach to the market was flawed.

“The access business has been delivering substantial cash flow for some time,” he noted, adding that “as we recognize that our growth platforms are not performing as expected we need to be prepared to change the underlying business models and cost structures and we’re prepared to do that.”

Huff said “in my mind there can be little question that a bypass technology which allows us to access customers with high-speed data connectivity just has a lot of inherent goodness. We can also feel great about the fact that the number of Wi-Fi-enabled devices in the marketplace number in the hundreds of millions and is growing‚Äö?Ѭ?As Apple and AT&T recently announced to the world(a reference to the release of the iPhone), Wi-Fi doesn’t have to simply compete with other wireless technologies, it can compliment them. All of these things point to a business that we should be interested in if, and that’s if, we know how to scale it in a way that provides a return to our shareholders for the money invested.”

Huff said that EarthLink is doing a detailed review of its muni business model and is “beginning a dialog” with the municipalities it has partnerships with and is considering partnering with “to explore ways we can bring this exciting technology to their communities while still providing a return for EarthLink shareholders. As in all of our businesses, we expect a return on this investment. The Wi-Fi business as currently constituted will not provide an acceptable return. We’re actively exploring ways to scale this business more economically. We’re going to look for municipal government to step up and become a meaningful anchor tenant on completion of a build. That would go a long way in our being able to get an acceptable return on this investment. Until we’re convinced that we can build new networks and get an acceptable return we will delay any further new build-outs.”

His statement came in the company’s second quarter earnings call this morning where EarthLink reported consolidated revenue of $312.2 million and income from operations of $20.2 million for a net loss of $16.3 million or 13 cents per share.

Click here to read the press release on EarthLink’s earnings.

Click here to hear the Webcast of EarthLink’s earnings conference call.

Comments

  1. Jacomo says:

    If they are to survive in this market they need to deploy real carrier grade systems that are upgradeable and have robust backhaul capabilities (between Nodes).In short what they need to do is change their Mesh Vendor and go with either a BelAir or Strix system that are robust enough to handle todays big bandwdith demands, multiple radios and spectrum and future (802.11n, WiMAX and 700Mhz systems). In addition if one wants a carrier grade network they will ultimately need to deploy Fiber Gateway links of 100Mbps to handle these new 802.11n systems coming on line in 2008(Intel using 802.11n in their existing Centrino products)
    Existing 1 and 2 radio Mesh Nodes/AP are not adequate to address the heavy demands customer will place on these Portable and Mobile networks and near term results will prove this out.
    Vendor Spin:
    Deploying 30-40 Nodes per Sq Mile and deploying Wireless Gateways every 3-4 nodes is not a viable solution and is more vendor spin to address inadequate backhaul and system design flaws.
    Ask why the big boys are beginning to releasing and or announce 3 & 4 radio modular systems in the 2008/2009 time frame.

    Jacomo

  2. “Until we‚Äôre convinced that we can build new networks and get an acceptable return we will delay any further new build-outs.‚Äù

    To paraphrase, “we are out of the muni wi-fi business.”

    There is currently no direct to consumer muni wi-fi business model that works.

  3. kimo crossman says:

    Does this mean San Francisco deal is dead?

  4. Esme Vos says:

    No, Kimo, don’t bring out the champagne glasses yet. EarthLink has not made an annoucement although it would be a perfect time for them to get out of it given that the Board of Supervisors is trying to change the terms of the contract signed by the City.

  5. Kimo Crossman says:

    Esme

    The Board of Supervisors are only asking for reasonable things that were in the original RFP which have now been avoided in the contract which was negotiated without the the BOS input.

    A Service Level Agreement for indoor and outdoor coverage, better privacy per ACLU/EFF, 8 year franchise rather than 16, 1MB/s speed in the original RFP and what Google is providing Mtn Vw for free.

    If EarthLink/Google pullout now – it is because they could not meet the original RFP requirements which by the way, other bidders did agree to provide.

  6. Esme Vos says:

    I am sure that none of the Supervisors ever let politics, especially their personal feelings for the Mayor, ever get in the way of their assessment of the EarthLink project. By the way, the Tour de France has finally eliminated doping from the race.

  7. KATULONG says:

    Jacomo says:
    “In short what they need to do is change their Mesh Vendor and go with either a BelAir or Strix system that are robust enough to handle todays big bandwdith demands”

    What is the difference between bel air,strix,tropos,firetide and skypilot? they all offer multi radios. Bel ir is so expensive that it actually requires more radios per sq mile. so as strix.
    All these vendors use 2.4 for client and maybe 5 ghz for backhaul. So therefore, theoretically they all should be robust enough to handle big cap. Bandwidth throtlling also depends on how you optimize the network isnt it? (putting more nodes on a dense population as long as the unit has automatic gain control, or lower the attennuation on the TX via software, I dont think bel air and strix has that.)
    Just my 2 cent opinion. Enlighten us on this please.

  8. The issue for Earthlink is not which canopy technology Mesh, WiMax, BelAir, Strix, etc. provides the best cost performance balance. The business model component that Huff is focused on is revenue. Earthlink needs more of it. They have to capture paying customers to offset the costs of building out the network. We discussed this on our panel a year ago in Minneapolis. At that time everyone was talking about ad revenue being the engine. But, the problem with ad revenue is that you need people on the system to see the ads, and it costs money to capture new users before they even begin to generate revenue (not to mention that you have to build out the network first as well.) Earthlink was novel in their approach because they also included a wholesale element in their business plan, where they could wholesale connectivity to other branded networks (like DirecTV) that already have customers. This would reduce the customer acquisition cost. However, Earthlink has now discovered that even this approach does not produce revenue soon enough. They need guaranteed revenue day one. It is no surprise that Huff uses the term “anchor tenant” in describing the relationship with the municipality. Simon Properties (the big shopping mall REIT) will not break ground on developing a new mall unless they have pre-signed several large “anchor tenants” like Macys and Sears. Earthlink has come to the same conclusion. It does not make sense for their shareholders to take on the economic risk of “developing” a municipal network without having pre-signed, revenue generating agreements with at least one anchor tenant, in this case the local municipality.

  9. Question: What is the difference between bel ir,strix,tropos,firetide and skypilot?

    Answer: Night and Day difference: Strix and Bel Air have 3-6 Radio options per Node. Tropos still has a single radio delivering Access and Backhaul no matter how they spin it. Their new second radio performs a second task as in 4.9Ghz backhaul, in short they are still using 1 radio (2.4Ghz)for basic Mesh.

    Question/Statement: They all offer multi radios. Bel air is so expensive that it actually requires more radios per sq mile. so as strix.

    Answer: You need to update your information and get each vendor to propose a single 2 or 4 radio node. You will probably find BelAIr is more expensive per Node than maybe a Tropos but that is apples and oranges. You will find Strix is competitive with most of the other vendors even when comparing their 4 radio OWS2420 with a 1 or 2 radio node from Tropos or SKypilot-Not very familar with Firetide.

    Statement: All these vendors use 2.4 for client and maybe 5 ghz for backhaul. So therefore, theoretically they all should be robust enough to handle big cap.

    Answer: As mentioned above Tropos does not, they may have a 5Ghz or 4.9Ghz radio but it is used for something other then backhauling the customer Access between nodes. They still use the same radio (2.4Ghz) for both Access and Backhaul.
    As an example, Strix 4 Radio Mesh Node uses 2ea 5.8Ghz Radios for backhaul (total 40Mbps out of each Node) and provides 2.4Ghz radios(2 each)for Access. Most Strix Mesh nets use one of the 2.4 for Access and reserve the second for a spare or uses it for a Point to MultiPoint extension of the Mesh bandwidth to other customers.
    Their 6 Radio Node has 3each of the 2.4 and 5.8Ghz radios-all modular and upgradeable in the field. WHo eles does that today??
    This is a true Carrier Grade product.

    Statement: Bandwidth throtlling also depends on how you optimize the network isnt it? (putting more nodes on a dense population as long as the unit has automatic gain control, or lower the attennuation on the TX via software, I dont think bel air and strix has that.)

    Answer: I love this marketing spin by vendors with limitations in their systems operations. Placing more Nodes per square mile may well improve performance, especially in the backhaul area. But what it is really saying is that you need to spend far more money with the 1 & 2 Radio nodes to get adequate Access and Backhaul performance in a given area. Not to mention the fact that these limited systems are keeping Motorola in business as they require multiple Wireless Gateways-Moto Canopy radios (in many cases 1 every 3-4 nodes). All this does is add far more costs at start up and even more as the network becomes congested.

    Statement: Just my 2 cent opinion. Enlighten us on this please.

    Answer: This is still an emerging market and what the Munis really need is a true Carrier Grade mesh product that can be upgraded in the field (ie.adding 802.11n or WiMAX radios)and use FIber backbones for Access vs. wireless radios. Only then will these Metro Area Wireless Mesh networks be able to address the massive demands (VoiceIP, Video Uploads, P2P & GAming etc.) customers are going to place on these wireless nets.

    Jacomo

  10. To Jacomo: Firetide has 2 backhaul radios. The mesh nodes provide backhaul only; client access is supported via modular access point (single radio). You can attach 3 outdoor access points to an outdoor node, making the combined radio count 5 max — 2 radios per mesh node, 1 radios on 3 APs each.

  11. “Does this mean San Francisco deal is dead?”

    Yes. And I will put money on it if anyone feels confident enough the other way.