All this activity raises the question:  How many app stores can the market sustain?   I am going to go out on a limb and say that the answer is probably none.  In fact, I predict that within the next five years, app stores will go the way of carrier walled gardens and consumers will get most of their apps directly from the developers and brands that create them.    Here’s why…

Until now, the app stores (by “app stores” I mean Apple’s iTunes primarily) have served four functions:

• Making the Market: By generating market awareness and excitement about apps, the Apple iTunes store conjured a market for apps where there previously was none.   The iTunes store is part of a virtuous circle:  demand for apps draws developers into the fold; more developers means more desirable apps are created; more desirable apps means higher demand for iPhones; and more iPhones means even greater demand for apps.
• Curating: App stores are in a position to separate the wheat from the chaff.  Nobody enjoys paging through dozens of ‘crapps.’  In a world with tens of thousands of niche developers and few well-known brands, the app store provides a service by keeping out substandard apps (as well as malicious or poorly executed code).  The app store can also stimulate sales of selected apps by featuring them in a prominent location.  While I don’t have statistics to cite with regard to app stores, in studying consumer behavior in supermarkets, research by Wharton professor David Bell suggests that up to 45% of purchases are unplanned within the app store demographic of ‘young, unmarried adults with above average income.’   If the same proportion holds true for app stores purchase behavior, as I suspect it does, then the curation function could account for as much as $3.2 Billion of the $7 Billion that is forecast to be spent on apps in 2010.
• Feedback:  Apps that receive positive user feedback outsell those that don’t.  No secret there.  Numerous studies have shown that online reviews are second only to direct word-of-mouth recommendations when it comes to influencing online purchase decisions.   By providing a forum for user reviews and comments, the app stores give prospective buyers a valuable information source that is not currently available anywhere else.
• Collecting Payment:  Making the purchase process flow smoothly is a key ingredient for app sales.  Online consumers are loathe to enter credit card information – especially for a one-off purchase of $0.99 from an unknown seller.  And from the merchant’s perspective, dealing with credit card fraud (both the malicious andfriendly variety), and the resulting chargebacks, is as appealing as a stick in the eye.  As a result, offloading this function to the app store is the best option available.

In a closed system, such as Apple’s iTunes, the app store concept makes a lot of sense.  By locking app downloads for iOS devices to iTunes, Apple has been able to carefully orchestrate the right mix of style and curation, resulting in the great user experience that iTunes is known for (let’s not forget that none of this would be possible if the Apple devices were not compelling to begin with).  In return for their efforts, they levy a very hefty 30 percent fee on any commerce that’s conducted. Don’t get me wrong, I believe that Apple is entitled to whatever it can command for its role in creating the market for apps.   As I blogged back in 2007, smartphones, as we know them today, would not exist if it were not for Apple.  But, let’s recognize iTunes for what it is.  Apple’s 30% fee and the tight control that it exercises over the iOS ecosystem are monopoly constraints that would be unsustainable in any open market situation.

Apps are digital goods that will naturally find the path with the least friction.  And, all things equal, market participants will gravitate to the markets that are open.   Of course “all things equal” is the key.   Before demand for apps existed, and later when the iPhone was the only real smartphone available, all things were not equal.  Apple held a natural monopoly and could afford to impose its “style” constraints in any way it chose.   With the explosion of Google’s Android OS, there now exists another smartphone ecosystem that is comparable to the iPhone and iTunes.  As a result, the landscape must change.

In an open system such as Android, where it’s possible to install apps from any source (known as side loading) the logic for an app store breaks down.  Developers will market their apps through the channels that best suit their needs.  In most cases, that will mean a direct transaction between the developer and the user.  Even today, there are hundreds of new startup companies developing better alternatives to the app store model.    App recommendation engines such as Flurry, Appsfire and appSpace will duplicate, if not improve upon, the curation and feedback functions. Others, like BilltoMobile, PayNearMe, Bling, Venmo, Boku and Zhong, are giving consumers and developers better ways exchange money.   As enterprising third parties find ways to innovate around the curation, feedback and payment functions, app developers will embrace them.  And, as developers embrace open systems, new ecosystems will evolve.

Just like the iTunes app store made carrier walled gardens obsolete, open app store ecosystems will make the iTunes model obsolete.  And if market forces don’t do it, the Department of Justice will.  After all, what’s the difference between Apple’s position with iTunes and the iPhone today, and the one that Microsoft had with Internet Explorer and Windows not so long ago?  Not for nothing, the fact that Microsoft’s Windows Phone 7 will not allow side loading virtually guarantees its failure, in my opinion.

In case it’s not already obvious, it’s my assertion that there really isn’t much of a role for a third-party ‘store’ in an open ecosystem for digital goods.  That’s why you don’t find many software retailers online.   I suspect this is also why Google has chosen not to invest any resources in the curation and feedback functions of the Android Market (much to the dismay of current Android developers).  Unlike the enviable position that Amazon has carved out for itself in the world of physical goods, where fulfillment and physical delivery is a critical function, I can’t see them adding much value to the delivery of apps in the long run.   In a digital world, it’s just too easy to bypass middlemen and go direct to the source.

Flashback to November 2007, when Verizon first announced plans to open its network. Back then, Verizon was in a pitched battle with Google for control of the 700Mhz wireless spectrum, and industry watchers saw it as a conciliatory effort to appease regulators and public perception. The Wired article describing the move was appropriately entitled 'Pigs Fly, Hell Freezes Over and Verizon Opens Up Its Network -- No, Really'. In fact, few believed that they would actually follow though.

Yesterday, us Doubting Thomases were forced to take another bite of crow when Verizon announced the availability of Messaging, Network and Location APIs at the Verizon Developer Conference in Las Vegas. And, like the deal with BilltoMobile, the APIs are being provided by third party specialists . Smart move! It takes a hungry innovative start-up to understand the needs of mobile developers. Arguably the most valuable of the APIs will be network based location provided by LOC AID technologies (unlike today's location services, which are driven by GPS from the phone itself, Verizon location will be fed by the network).

The real surprise for me isn't that cellcos are opening up (not to blow my own horn, but I first called the trend in August 2006 on this blog), but that Verizon is leading the way. I would have bet dollars to donuts that Sprint would get there first. In any case, I'm not complaining.

It should be noted that similar attempts by major telcos have not been wildly successful. BT's groundbreaking Web21C initiative, the first attempt by a telco to open its core network to software developers, is struggling for traction. After launching to lackluster demand, BT scooped up Ribbit (a Silicon Valley start up that had amassed a large developer community around its telephony APIs). More than a year later, however, Ribbit developers have yet to find a ready market of consumers.

The lesson to learn from BT is that you have to prime the pump from both sides: you need an active developer community to turn APIs into compelling apps, and you need a vibrant marketplace to put the buyers and developers together. For Verizon's gambit to work, it must now follow through with appropriate incentives to the mobile development community. For starters, it is encouraging developers to submit apps to its newly rechristened VCAST store. But just in case the second time around is not a charm, Verizon will also allow API users to sell their wares wherever the market exists (such as the Android Market and Blackberry App Store).

For once, I'm rooting for Verizon.

• Although a successful app store is a good way to create a lasting relationship with the customer, it seems obvious that there is not enough room in the app store universe for all the interested parties.   Many of the current app stores are destined to fail.  It remains to be seen what other types of relationships might be created.  Yahoo! attempted an alternative approach with  Go 3  which optimistically aspired to be a device independent entry point for user applications.   The device-specific MotoBlur approach is yet another alternative.

• Apart from creating lasting relationships with end-users, it is also necessary to cultivate lasting relationships with mobile software developers.  If the developer community does not create appealing apps, then there will be nothing to attract end users.  In the case of the iPhone (where Apple controls the OS, device and store) Apple can create a controlled development environment.  While this may simplify the development process for developers, it can also prove constraining, as witnessed in the well-publicized case where Apple denied entry to the Google Voice app.  To the extent that other players do not control all three legs of the tripod, it may hamper (or enhance) their ability to attract developers.

• It seems logical that mobile users will tend to have a 'favorite' app store.  The question this raises is which party in the ecosystem has a competitive advantage to win this role.  Both the OS makers and the network providers have unique attributes.  The OS makers can provide APIs that give apps more capabilities and ease development.  The network providers can provide additional services that bring into play elements that are beyond the scope of the device itself (which I have previously blogged about here).  It seems to me that device makers not in control of the OS are at a decided disadvantage when it comes to attracting software developers.

• After years of poor execution, Moto is clearly against the ropes.  A failure now could be devastating to the future of its handset division.  So the decision not to play in the app store space is a bold one that will either prove prescient or another nail in the its coffin.

At the heart of the matter is the stark reality that network operators don't want to be dumb pipes. More specifically, they dread the idea that outfits like Google may generate rich service fees from subscribers, while they are left with relatively paltry access fees. Especially after spending billions of dollars to lay fiber, buy frequencies, build cell towers, etc. Yet, the proliferation of OTT services, of which Google Voice is just one example, indicates that this is direction things are headed.

So what will the network operators of the world do? The immediate answer is that they will fight tooth and nail. They will give precedence to their own services using all available means for as long as they possibly can. They will lobby congress (or the relevant governing bodies in other countries) against open access in an effort to maintain the world they know (witness the 700Mhz auctions of 2007). Along the way, innovative services like Google Voice will continue to be suffocated, if not killed outright.

But does it really have to be this way? The answer is a definitive 'No'. Not if network operators started to view the OTT application providers as potential customers instead of competitors. That doesn't mean that application providers should pay "extra" for priority over the last mile, as some network operators have suggested. Rather, it means that network operators should ask themselves "how can we provide value to OTT applications?" rather than fighting them off in congress and the courts (especially the court of public opinion).

When viewed in this way, there is actually quite a bit about network access that need not be dumb:

Payments - The bane of all digital content providers is collecting payment. Without a physical address to which goods may be delivered, digital merchants are constantly battling fraud and chargebacks. Even Paypal, the 800 pound gorilla in the area of online payments, must battle fraud when no physical delivery takes place. A network provider could very easily fill this void. They know where the customer lives, and they have a tangible physical connection to them. Much like NTTDoCoMo has been doing for years with iMode, they could act as a trusted intermediary and OTT application providers would happily pay them for this service.

Presence - Many of the promising applications just over the horizon will depend on the concept of presence. Are you in your car, or at your desk? Is your preferred method of communication a video call or an instant message? Does your device of choice have the capability to support MP3 or AAC as a file format? All of these questions can be answered by presence, and as the entity with a direct connection to your gadget, the pipe provider is in an optimal position to mediate it.

Directory Services - Closely related to the idea of presence are directory services. The ability to reproduce the good old phone book has proven to be one of the most vexing issues of the post-Ma Bell era. Just try to look-up the phone number of a mobile phone or VoIP user and you will understand the dilema. The lack of a trusted repository has caused ambitious attempts at creating online directory information, such as ENUM, to falter. Rather than evolving into online resources, however, the Bells have been steadily shedding their directory services ever since divestiture.

Location - As with presence, the network operator can best determine the actual physical location of the mobile subscriber. The current generation of location-based applications require that the application be 'turned-on' in order to transmit location information to the application provider (hogging precious CPU, battery and screen real estate in the process). A much saner approach would be for the network operator to maintain location information and then provide it to interested applications (with the subscriber's assent or course) upon request. Application providers would be willing to pay for this knowledge.

Marketing Intelligence - Since the pipe is the only common denominator between the subscriber and the many different providers of content the subscriber might access, the pipe owner is in an optimal position to provide marketing intelligence to the content owners. An astute pipe provider could build a comprehensive profile of its subscribers and offer this as a service to content providers. A Neilsen for the mobile age.

I can think of more, much more. But the above list should suffice to make the point.

Of course, there are some rough patches to be navigated. Chief among them is that all of the these services would entail drastic changes for the typical network operator (in terms of culture as much as business model). Further, a payment suitable model would have to be developed. The Internet is at odds with the traditional 'pay for access' model that network operators adhere to. A 'pay for play' model, where the network operator takes a chunk of the application providers paid-for transactions would work much better (again a'la iMode).

Yes, it's still early days. But it is evident to me that network operators must change their way of doing business in order to remain relevant. The question is which network operator will be the first to seize the mantle? Somehow, I don't think it will be AT&T.

1. The essence of the offering is a 'follow-me' service that allows you to filter, screen or direct inbound calls to the phone of your choice.  Google Voice provides a local number at no cost.  Calls to the local number can then be handled in several different ways:  forwarded to one or more phones (including SIP phones, see below); sent to voicemail, screened, or recorded.
2. The best part of the service is that it is free (at least in pecuniary terms).  There is no charge for the local number that gets assigned, or for outbound calls to the U.S.   Every other follow-me service that I am aware of charges a monthly fee for the former, and a per-minute fee for the latter.   The quid pro quo is that you must sacrifice yet another level of privacy to the great Google machine.  Google will undoubtedly use this service as another means of targeting advertisements (more on this later).   If you are still operating under the delusion that anything you do is private, then this is not the service for you. 
3. The user interface closely resembles Gmail, so it is both familiar and intuitive.  It seems obvious that Google plans to merge the Voice interface into Gmail at some point in the future, which would  conveniently place all incoming messages in a single screen. 
4. The interface displays all of the contacts from your Gmail account.  This makes it very easy to manage your contacts as well as to select a contact name and create handling rules.  You can also create custom greetings so that every caller in your address book will hear a unique and personalized message.  The screenshot below shows the special greeting I created for family members.
    
   
    
5. The call handling screen gives you an option to forward calls to one or more phones, and one of the options is a Gizmo SIP account.  Gizmo is a paid service, but offers a free version that allows you to register a SIP phone and receive calls from other on-net users.  The SIP option promises to open up a lot of flexibility, although I was unable to make it work with my own Gizmo account (It's possible that Gizmo disables inbound calls from Google Voice since this would compete directly with their own inbound call service, which is not free).  It would be nice if Google added an option to forward calls to any SIP URI.
    
    
    
6. The most useful new feature is voice transcription, which will convert an audio message left to voicemail into text.  The resulting text transcript is then visible in your inbox and is searchable.  While the voice recognition software used for the transcription is not perfect, it is definitely good enough to facilitate keyword searches.  In the example shown in the screenshot below, the brand name was garbled, but I was able to locate the message by searching for "sale" and "tractor" (the lawn tractor is still for sale, in case anyone in the Fairfield county area is interested).  
   
    
7. It is possible to make outbound calls from within the user interface.  Calls within the U.S. are free, while international calls can be made using a prepaid account.  Rates for international calls are lower than you would expect to pay your cellco, but not as low as are available from specialty service providers.  Calls originated through the Google Voice servers will first ring your phone, and then the phone of the called party.  This gives you the very convient option to record outbound conversations (although it is not currently possible to get a transcript of these calls).
8. For good measure, Google Voice also includes an SMS short text service.  This allows you to receive SMS messages into your Google Voice inbox, which are then searchable, just like voicemail messages.   You can also send SMS messages from inside the user interface using a lightweight ajax interface.
       
9. All of the web UI functions, including placing an outbound call or SMS, can be used from a mobile browser.  The already utilitarian interface is rendered more sparse to make access from slow mobile data connections more tolerable.  Testing with the Blackberry mobile browser and the Opera mobile browser was surprisingly easy.   And, I would be very surprised if access was not made even easier from future Android phones.

With Gizmo5 you cannot build a true peer because the source code is not available. Its method of establishing the P2P overlay is proprietary and so without the source code you cannot build a true peer (it does not use the open P2P SIP standard). Therefore using the same yard stick used to denounce Skype as closed, Gizmo5 is also a "closed network."

 

Maybe [Gizmo5] thinks that it's open because it supports the SIP URI names/address space? But I think that would be a weak argument for two main reasons. Firstly because Skype supports not only the private Skype namespace (i.e. a Skype ID) but also the E.164 namespace (i.e. telephone numbers). There is little consumer demand for the additional support of the SIP URI space ... You can call between Skype and Gizmo using the E.164 namespace. I see no reason Skype should have to support the SIP URI namespace to help bolster a competitor! But again, this argument completely lacks vision of the long term evolution in communications, sticking to telephony calls over IP (yawn) being the future.

• Open Source means that the intellectual property used to create software is freely available for anyone to use and/or change.  The Asterisk PBX is the best known example of open source communications software. There are many open source SIP software implementations. Neither Skype nor Gizmo5 is open source in this sense.  
• Open Standard refers to a format or specification that has been agreed upon by concensus, is publicly available for reference, and can only be changed by a governing body through an established process.   The SIP protocol (IETF RFC 3261) is an open standard in this sense.  The Skype protocol, was created by Skype and is known and controlled only by Skype.   There may be some truth to belief held by many that it is a "better" protocol than SIP.  But an open standard it is not.
• Open Platform refers primarily to a business model where the owner of a network (be it a physical network like Verizon, or a social network such as Facebook) allows third parties to access its users for the purpose of pursuing commercial activities. Openess in this regard can vary greatly depending on the level of access that the network owner grants to third parties.  Skype is an open platform to the extent that it publishes an API that allows third parties to create complemetary products/services and conduct commerce with Skype subscribers.  Gizmo is an open platform to the extent that it's users are free to call and be called from other VoIP networks.  It makes no sense to talk about SIP in terms of an open platform, since it is a protocol.  

Comments on this Entry:

(Rich Tehrani on Jan 2, 2009 5:17 PM)

Eric,

Great post. Very thought provoking and useful. Rich

One of the primary problems that service providers face when managing VoIP and multimedia calls is the separation of the signaling and media streams.  In other words it is quite possible that the two streams may take completely different paths through the network.  In addition, even when they do pass through the same device, it may not be aware of the relationship between the streams.  Some devices within the network are however specifically designed to understand and manage the separate signaling and media streams - session border controllers (SBC)[8].

SBC is a session-aware device that manages VoIP calls at the borders of an IP network.  Unlike most network devices, SBC are aware of the relationship between the two parts of a VoIP call: signaling and media.  Session Border Controllers handle both media and signaling, intercept can be performed in a completely undetectable manner.

SBCs usually sit between two service provider networks in a peering environment, or between an access network and a backbone network to provide service to residential and/or enterprise customers.  They typically are deployed at the border between two VoIP networks, and they offer an ideal location to implement a Lawful Intercept solution.

Whilst the detailed requirements for VoIP Lawful Interception may differ from one jurisdiction to another, the general requirements are the same.  The LI system must provide transparent interception of specified traffic only and the subject must not be aware of the intercept.  The service provided to other users must not be affected during interception.

As part of the effort in developing a broadband VoIP lawful interception architecture, we implemented a prototype and conducted evaluation experiments on the prototype system.  In this document, we first describe the prototype solutions and then report experimental results.  We hope that this document can provide useful information to those interested in the subject.

Dear Eric,

Tell the FCC to Stand Up to Scare Tactics and Open White Spaces for Everyone. Sign Our Halloween Action Card

"They're Ba-ack..."

The NAB's hired guns have bombarded policy makers with false claims in a desperate, last-ditch attempt to hoard these airwaves and to disrupt a critical FCC vote taking place in just six days.

The FCC's five commissioners must not buckle under the intense lobbying pressure:

Tell the FCC: Don't Give in to NAB Scare Tactics

Here are the facts:

1. If we open white spaces now, we can bring the social and economic benefits of a fast Internet connection to tens of millions of Americans now on the wrong side of the digital divide.

 

2. FCC engineers have tested white spaces devices and determined that the technology can deliver high-speed wireless Internet, without interfering with adjacent TV broadcasts.

 

3. The NAB and Big Media are doing everything in their power to close off access to white spaces because they fear competition from new innovators and losing control of the public airwaves.

The NAB is furiously spending millions of dollars on dirty tricks and political intimidation to scare the FCC away from white spaces. They have high-priced lawyers and lobbyists, but we have you.

Take just one minute to sign this Halloween action card and forward it to your friends. Free Press will deliver your cards to the FCC on Halloween and make sure they "treat" us by opening white spaces for eveyone's benefit:

This Halloween: Stand Up to the Lobbyists' Scare Tactics

With your support today, we'll expose Big Media's fear-mongering and make certain that white spaces are used to make fast, affordable Internet service a reality for everyone.

Thanks,