Last month I’ve mused about Mobile VoIP, first suggesting IMS will kill it and then stating it is not going to happen any time soon. I think it’s time I stand behind my big words and explain why.
Here are my 5 cents reasons:
1. Mobile VoIP is (Still) Downloadable
Mobile VoIP is yet to be embedded as part of the natural user experience of a mobile phone. There are some handsets that support mobile VoIP, but finding it, configuring it and using it is way harder than simply calling someone.
In most cases, the only way to use mobile VoIP is to download a third party application to the phone. The need to download an application for daily communications is something that most users won’t accept. And from the developer’s perspective, the need to develop a specialized application for every other handset out there is a real hassle.
And if you think that the “millions of users” who are already using mobile VoIP are a movement, think again: there are 4.3 billion mobile subscribers worldwide today. A few millions is a drop in the water, nothing more.
2. Mobile VoIP by Whom?
Decided to use mobile VoIP? Great! But which one?
Skype? Google Voice? Want to use Fring for your calls? Or maybe their arch rival Nimbuzz?
And there are dozen others that I will not bother to mention here.
Mobile VoIP client vendors vary widely in offering, user interface and quality of experience. Not to say in the “coverage” of the people you can contact using it.
The bottom line is that if you want to use Mobile VoIP today, really use (as an alternative to the “plain old” mobile calls), you’ll probably need more than a single client, switching between clients depending on whom you want to contact.
3. Mobile VoIP – On What Network?
Mobile VoIP can use either a WiFi connection or any network using its data connection (3G, for instance).
Today’s mobile networks are already quite optimized when it comes to their bandwidth use. They are based on the old faithful circuit switching technology, where a “channel” is reserved for the call over the network and used only for the call itself. This allows reducing the overhead on top of the actual voice data to a bare minimum. It also what makes it way more bandwidth-optimized than mobile VoIP.
This means that carriers have no incentive to move to a VoIP network, as it may require more bandwidth per call on their side (and bandwidth is scarce). It also means that carriers will do what they can to reduce the number of people using Mobile VoIP on their networks (tweaking data network latency and packet loss, forcing people to download the applications, not approving VoIP applications into their specialized app stores, etc).
So Mobile VoIP is left with WiFi, where coverage is getting better every year, but it sure isn’t as robust as the cellular network is. WiFi also requires a roaming service of some kind, making it a hassle of checking each time which WiFi network exists in the vicinity and trying to connect to one that you can actually work with.
4. Voice Quality
Circuit switched mobile calling takes care of voice quality simply by guaranteeing bandwidth end-to-end. It is also what cellular networks were built and optimized for – passing voice calls. Data came later.
And data didn’t take voice calling into account. As Mobile VoIP is sensitive to packet loss and latency, it is not suitable for the 3G data networks it rides upon (and its suitability for WiFi access points can also be questioned). This is why, in a lot of cases, and with a lot of the different mobile VoIP applications out there, voice quality won’t be as good as that of regular mobile calls.
5. Battery Life
Mobile VoIP requires that the data connection will remain open at all times. This is because it has to “listen in” on what goes on over the IP network, “trying” to catch packets that are routed to it. It is the case even when you are not participating in a call.
Why is that an issue? Because baseband chips in mobile handsets today are optimized for circuit switching – up to the point of reducing their power requirements on standby mode. But standby mode in most handsets (or all that I know of, at least) don’t allow for having an open data channel (or WiFi connection), listening in on IP packets.
So if you want to go all-Mobile VoIP, expect to recharge your battery more than once a day. Not that usable. Or green.
Conclusion
If this seems to you that I am against mobile VoIP then you are wrong. All of these issues will be fixed. All that is needed is enough time to improve the networks and the technologies that are used by mobile VoIP.
For now though, Mobile VoIP gets a resounding #FAIL.
Tags: 3G, App Store, bandwidth, battery life, Circuit Switched, Clients, communication, Fring, Google Voice, handsets, IMS, latency, Mobile, mobile phone, mobile VoIP, Nimbuzz, packet loss, quality of experience, Skype, User Experience, VoIP client, wifi
Comments and trackbacks
1. ARJWright | October 13th, 2009 at 2:08 am
Ooooh, good points here and many of which that I didn’t consider. Your numbers 1 and 2 though have been the nail in the coffin so far for me. I really think that manufacturers need to push the contact/address book envelope more and enable VoIP services (SIP, APIs, whatever) to minimize this end of things. Because until those two are settled, VoIP will indeed stay a niche application/service on the mobile end.
2. Henryk | October 17th, 2009 at 11:43 pm
This is a matter of perspective. Definitely VOIP will not dominate calls within one network or even within one country, because financial incentive is not sufficient. But clearly this is a great method of reducing cost of international calls. For all S60 Nokia devices Fring would be enough and depending on a specific model of a phone it gives between 2 and 5 hours of essentially free of charge international 3G calls on one charge. I did 200h of 3G calls and so far I have not noticed any action on the side of 3G providers, but probably it varies from country to country.
3. Tsahi Levent-Levi | October 18th, 2009 at 6:03 am
@ARJWright – true. problem is – how do you settle the issue of downloadable apps and the issue of too many players?
Other than an operator stepping in and adopting it wholeheartedly, I see no way.
@Henryk – this is what I like most about people who preach mobile VoIP – they provide the value of the service and ignore the fact that for 90% or more of the population of cellular customers aren’t going to use it because of the reasons I stated here.
I think it’s time for mobile VoIP to become a reality – I would certainly want to see it grow, but I don’t think the market is there yet.
4. Antoine RJ Wright | November 20th, 2009 at 3:42 am
Downloadable app = open address book. If you will, instead of installing apps for dedicated services, the contacts app is essentially a VoIP, IMS, and IM browser (think Firefox out of the box), and all of these services are simply extensions to it. It would be easier to develop this model, but a pain in the butt to do carrier testing with it.
Carriers would have to adopt it to be successful though; they’d instantly see that such a model means that they could be easily shut out of a user’s address book.
The other issue is to figure out a QoS metric for VoIP providers that’s consistent. Right now, its not a nice proposition to figure this out, and many of these providers give nothing other than a statement (if you have a data connection, we work) towards their service quality. This needs to change for the consumer’s interest, even if it means fewer low-cost services.
5. Tsahi Levent-Levi | November 21st, 2009 at 8:29 pm
Antoine,
The firefox concept is a very interesting concept. I’ll really need to think it over and see how it works for me for voice communication related services – is it more suitable than the AppStore paradigm? I don’t know, though it does sound like it.
It means that Mobile VoIP clients will fade away, and be replaced by Mobile VoIP service providers (what they really are today). But then again, will the service providers let that happen, or will they just provide 99% of these services out of the box on their own?
6. ARJWright | November 21st, 2009 at 9:11 pm
Its not an app store, its a plugin to an existing system. To that end, its sustainable as long as the core platform evoloves – like Firefox has. The key is making sure that the base stays forward moving, and then the extensioned-services can have something to work with.
Here’s the idea in a bit more detail at ideas.symbian: http://ideas.symbian.org/Idea/View?ideaid=2786
It can definitely work; the question is, can folks see it simply, before adding complexity (marketing, etc.)
Trackback this post | Subscribe to the comments via RSS Feed