Blog Post

Mobile VoIP: Killer App in More Ways than One

Stay on Top of Enterprise Technology Trends

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

By Jesse Kopelman

This year we are seeing widespread US deployment of CDMA EVDO Rev A. The next two years should bring the same for both UMTS HSPA (both uplink and downlink) and WiMAX.

What these technologies have in common, besides a lot of capitalized letters, is vastly improved performance, in terms of latency and data rate, over what has previously been available with wide-area wireless networks. A side effect of this increased performance is that the mobile wireless connection can now support high quality VoIP. It seems a certainty that this capability will change the mobile telephony industry forever.

The $80/month that carriers charge for standalone unlimited data plans sounds pretty good for an industry where many struggle to break an ARPU of $60/month, but it suddenly doesn’t look so good when one considers that high end voice plans can get upwards of $200/month. It would be a huge blow to carrier profitability to have the rate plan ceiling so drastically lowered.

Meanwhile, they are caught between the hammer and anvil, as they can’t raise data prices without hurting uptake there and data use has become a significant revenue source. The situation will get worse over time thanks to two factors: First, over the next two to three years T-Mobile, Clearwire, and Sprint WiMAX will all emerge as players in the nationwide mobile broadband market. Second, improved devices in the PDA and smart phone form-factor will allow for VoIP without compromising the expected phone look and feel.

In the end, traditional mobile carriers are powerless to protect their voice margins from the threat of VoIP. They could try blocking VoIP, but how well will that fly when they’ve got compete with both a T-Mobile eager to gain market share to recoup their 3G spending and WiMAX carriers with no legacy voice revenue to protect?

Also, do Verizon and AT&T really want to fight yet another Net Neutrality battle? Sprint Nextel is an especially interesting case, as the only way for its WiMAX efforts to succeed is at the expense of its legacy operations and vice versa. Another tack would be to market the circuit-switched service as having superior quality to any third party VoIP offering. This could be quite compelling to high end users who are often business people, but is largely undercut by years of scraping by with a service that is markedly inferior to landline.

To make the argument stick, they’ll have to invest considerably in improving the network and that will be a losing game as pressure from VoIP will only hasten the pre-existing downward pricing trend. Perhaps, since they can’t beat them they ought to join them – shut down their circuit-switched networks and go completely packet based.

A carrier’s own VoIP service will always have an inherent advantage over that of a third party – it doesn’t have to go over the Internet. It’s debatable that going VoIP will make it less expensive for a carrier to offer voice than their current highly optimized circuit-switched networks, but there is no doubt that only having to operate a single packet based network for all services will save money.

This is the direction they are heading with IMS, already. The point is that competition, especially competition from third party VoIP, is going to force them to move faster than they might like. Meanwhile, the cash cow of legacy voice may not have as much milk left in her as they thought.

Jesse Kopelman is a long time GigaOM reader and an expert in wireless technologies, who has been involved with many wireless network builds. This is his first contribution to a new feature on our site called Conversations, where we will invite folks whose comments we have enjoyed to share their thoughts on topics that interest our community of readers.

15 Responses to “Mobile VoIP: Killer App in More Ways than One”

  1. This is an area I have been researching in a lot of depth. VoIPo3G is inevitable – later versions of both CDMA and UMTS are all-IP. The question is whether VoIP will be purely operator owned-and-controlled, purely independent (Fring / Skype etc over a data connection), or some partnered version.

    Also, later variants of HSPA and LTE should get higher volumes of VoIP traffic than circuit-switched in the same capacity, owing to improvements like header compression. Latency will also improve in stages.

    My take is that it will be a mix, varying by country, demographic, CDMA vs UMTS prevalence, frequency allocations, and device type. Independent VoIPo3G will be more important in markets with lots of competing operators, plenty of fully-open smartphones and plenty of scope for tariff arbitrage.

    Another angle is whether there are more valuable applications for VoIP than basic person-to-person telephony. Embedding VoIP elements into games, enterprise apps, social network etc could enable pricing models that look very different from either per-minute, or plain flatrate data.

    I’ll be commenting a lot more on all this over coming weeks and months.

    Dean Bubley

  2. I can’t see many consumers being sold on the idea of paying high plan charges for “high quality” offerings. Such an approach reflects badly on their other plans (intimating their inferior quality). Are we all to expect an echo-riddled service or a sparser bank account? Not much of a choice!

  3. Jacomo

    The issue for the carriers is sufficient spectrum and bandwidth, whether CDMA or GSM. Data/Video and VoiceIP demands, not to mention the real killer P2P requests, are and will continue to seriously impact these carriers ability to deliver on their real business a consistant high quality voice service.
    They cannot cost effectively go to a flat rate per month for service because bandwidth, Latency, not to mention slow return path links prevent them from deploying a viable (QoS)and competitive service offering against either the emerging (yet to be proven) WiMAX mobile service and a properly designed (not these lame 1 radio nodes)Wireless Mesh Networks being deployed across the US in their major markets.

    An interesting question here is what role will Google Play if they teamed up with Muni Wireless Mesh Network (or new WiMAX)providers and handled all their VoiceIP Processing over an end to end Google Data Network (No need for the best effort internet here). They are about to launch their open standard VoiceIP Softphone and VoiceIP central management package that will reside in and be operated by them in their Fiber linked Data Centers(will be suprised if they do any Phone hardware).
    So when combined with Intels new line of Centrino Pro CHips for Laptops and bluetooth links for headsets they have a built in soon to be ubiquitous customer base.


  4. The mobile carriers control the transport medium for VoIP. Ergo, they can regulate or even charge specially for use of VoIP, or excessive streaming. Random throttling of connections will have no noticeable effect on ‘normal’ browsing, but will make a VoIP call unusable.

    I think there is an overzealous use of terms such as “disruptive”, which although can be applied in some cases, I don’t think it flies when the technology is disrupting the very service it relies on in order to function. TASTAAFL*

    • There ain’t such thing as a free lunch.
  5. samab

    This piece is totally non-sense.

    Clearwire — drop 20% since its IPO (at one time it went down by as much as 40%). Financial basketcase — losing hundreds of millions of dollars — and nobody is going to fund them the second time around.

    Sprint Nextel — another financial basketcase that got much of their AWS spectrum auction funding from cable operators. It is the cable operators that are gaining big on the VoIP front (killing Vonage in subscriber net adds). Vonage’s sharp fall from its IPO — signals the end of pure-play VoIP companies.

    Even the WiMAX trade group (the official cheerleader) is hyping only tens of millions subscribers world-wide for their technology — against (by that time) 4 billion mobile users.

    T-Mobile USA — well if VoIP is such a killer — then its German landline parent DT will also be in the same boat as AT&T and Verizon. And if DT gets weak then TMO USA will also be weak as well.

  6. Hi, I would like introduce you the sec.un.dum ( It looks like a Brazilian Technorati which pays for post’s bloggers. In the last month, sec.un.dum paied 2.4 dollars for each post in the 100 more visited list. This month it will pay 5 dollars. The site is a blog community for portuguese content.

  7. T-mobile in the UK have a $15 a month unlimited data package supporting over a HSPA network. I don’t believe they restrict voip. Vodofone and Orange are a bit more reserved, when the Nokia N95 came out a couple of months ago they removed the voip client.

  8. I’m quite skeptic about the mass-market uptake of VoIP over wireless, at least for three reasons:

    a) It will be difficult for end users to install use applications;
    b) The QoS will continue to lag behind normal calls;
    c) The operators will put some barriers

    Here some post from Dean Bubley’s blog on how European operators are limiting VoIP usage on 3G handsets.


  9. tomcoseven

    About 7 years ago I spent untold hours trying convince US and European MNO’s to introduce fixed price data plans for precisely the reason you describe — to increase take rates. At the time they were using per minute and per byte pricing that retarded adoption because subscribers couldn’t figure out how much they might be spending before the bill came.

    The MNO’s biggest concern was VoIP, both because it drained revenue from the voice business and used so much more BW. Circuit on mobile is best case 4kbps, whereas VoIP was about 26kbps (16kbps plus overhead) best case and sometimes significantly more.

    My suggestion at the time was fixed price downlink and per byte pricing on the uplink (above some reasonable limit). That way they wouldn’t have to block VoIP or P2P. The heavy users would end up paying $3-400 and all the normal users would get the benefit of fixed pricing. With the net neutality push, I wouldn’t surprised to see variable pricing again. It’s neutal and accomplishes the same thing… shut down Skype, Vonage and Joost.

  10. We are in the process of rolling out an Avaya VOIP solution hosted on a private network. I had just a conversation with a Sprint WiMax representative about being able to deply the data cards and the Avaya softphones. The represantive wasn’t resistent and actually thought it could be done — limitedly — with their current EVDO Rev A network. It seems Sprint’s business solution division seems to know which way the road is curving.

  11. As Tom ( reported, VZW does allow VoIP, presumably within the monthly quota that is “unlimited”. As I commented in Tom’s post this translates into a lot of minutes, but one would think while using a soft client. At first blush it looks like circuit switched voice is under threat, just as in the wireline world (, people will prefer to stay with circuit switched voice in the wireless world as well. Comparison to dinosaur turns out to be premature. But what will be appreciated is some form of enhanced signaling (without the intervention of the carriers) over the data connection to augment the voice session. VoIP proponents missed the boat for 10 years; one hopes that history doesn’t repeat itself in the wireless world.