Universal Mobile Interface

Avoiding app stores turning into app graveyards – Part 1: So what’s the issue?

Posted in UMI by Martin Vendel on December 17, 2009

The development of app stores has been beyond anyone’s expectation. The key driver for the development the last 18 months has been the iPhone app store but similar concepts have been around for many years with services, such as Handango and GetJar, as well as many operator portals. The overall numbers are really impressive. There is today about 115,000 iPhone apps and more than 100 million downloads a month (incl. iPod Touch), while the second largest app store Get Jar, serving all types of mobile phones, has about 60,000 apps and 50 million downloads a month.

With close to 40 million iPhones sold the number of apps downloaded per user is on average quite high. Some claim that the number is as high as 11 apps per month and user, but I’m not really sure they got the math right. Anyhow, is this a gold mine for the developing community? We do have success stories but the app stores are now quite crowded. On average each app is downloaded about 1000 times a month. But looking at the usage patterns for free apps presented by Pinch Media only about 20 % return to the app a second time, only 5% return to the app after 25 days and after 90 days only 1% still use the app. Keep in mind that these numbers are all averages. Also taking into account that the top 1% of the apps in the larger app stores probably account for 90% of all downloads, then the numbers become quite disappointing for the majority of the apps not making it to the top. Most apps do probably not have more regular users than, well, those involved in developing them. Sad but probably true. The numerous launches of app stores by device manufacturers and mobile operators do expand the market but it also becomes quite fragmented. All different operating systems and specific demands result in that each app have to be made in many different versions, driving cost even further.

The likelihood that a new app after some initial attraction ends up in a close to zero down load frequency and that those initially downloading it stop using is, despite the over all success of the app stores, quite high. The app stores are turning into app graveyards for most new apps, buried under piles of more attractive apps and with very little possibility to gain visibility. Gaining visibility is a marketing issue but the limited stickiness of the services may indicate that most of the apps should never have been made as they do not meet a specific customer demand. Services with a limited but loyal segment are fine but with false expectations on future take up and revenues the cost levels are usually far too high for the development.

The term app graveyard has been used for clusters of apps that were not approved and never made it into the so attractive looking gated communities of the app stores. The question is if they actually have a better chance of succeeding outside the fence. The app stores may be fantastic and one of the most disruptive elements seen for quite a while within the mobile sphere, but please remember that they still can’t sell services that are not really marketed, nor sell services that actually do not meet a customer need.

The Mobile Internet is dead – long live the Internet!

Posted in UMI by Martin Vendel on December 10, 2009

I think it’s time to stop using the term “mobile Internet”. In the early days of mobile Internet services, these were in WAP or similar formats, and usually within a separate universe compared to ordinary online Internet services. With increased bandwidth and capacity of the mobile devices the mobile and online versions have gradually converged. Today most content accessed through the mobile is technically integrated with the online web platforms, but often adjusted to the limitation of the mobile.

The mobile has become an additional access point to more or less any Internet service. We are referring to services “in the cloud” independent of access point and ordinary web services are increasingly viewed directly on more advanced phones such as the iPhone, parsed into a more mobile friendly format or through an app. What is then the difference between the “mobile Internet” and Internet? Well, there’s actually no difference as it is all one and the same Internet accessed and viewed through the computer, the mobile or, soon also to a high extent, the TV. However, there is still a need to adjust the content depending on context and device where the computer probably will be the exception compared to other devices, as it’s the only one with a full size key board.

The network apocalypse is getting closer

Posted in UMI by Martin Vendel on December 7, 2009

Mobile internet services have gained significant increase in attraction and usage. About 20% of all active on Facebook use the mobile to access the service. 27% of the mobile phone owners browse the Internet, use applications or download content, according to comScore. 3G, smartphones and flat rate data plans are key enablers.

Apple demands that the iPhone is sold bundled with a flat rate data plan to encourage the use of data services. The success of the iPhone gives a good indication for how large the unmet market need is and what can be achieved when services are easy to find, buy, download and use in combination with simple pricing. High quality screens, improved data speed, and flat rates also open up for services that are quite data heavy such as streamed videos, TV and music, for example YouTube and Spotify. The perceived end user value per MB decreases as the users become more advanced and with the introduction multi media services. The dilemma is that the additional value is usually less than the additional cost of production. With flat rates we can expect that the usage of non-voice services will grow rapidly resulting in heavy pressure on the mobile networks, demanding further significant investments.

Flat rates as such are a good thing for the end user giving cost control. The network operator may also initially get an increase in average revenue per user (ARPU) as customers migrate to flat rate data plans, but exponential increase in data traffic drives cost and additional network investments. There is today no incentive for a player like Apple to decrease the data consumption. This situation should worry the network operators as they may be trapped in a situation with competition on flat rate pushing the prices down at the same time as usage increases heavily.

What are then the options for the operator? Trying to avoid introduction of flat rates on the market is probably not an option. To cap “unlimited” data plans is possible but the upper usage limit has to be set at a quite high level not become counter productive. We will also see attempts to increase the price but it may be difficult depending on the competition. Most probably these alternatives will either not be possible or may have a too limited effect. The equation does not add up and we may gradually move into a situation where the networks collapse as it may be impossible to justify further investments. It may be necessary to enter into stage of network apocalypse before we can rebuild the ecosystem on sustainable business principles. Is there then no way out of this situation?

Let’s go back to the original thinking of NTT DoCoMo regarding imode. This success story was created in times of quite limited network capacity. imode was built to maximise the end user value per consumed data volume. The whole system was built based on this basic idea safeguarding that all actors in the value chain should be happy. NTT DoCoMo thereby forced, as they did set up the rules, the content providers to develop services that used as little data as possible. It is now time to once again try to optimise the value for the end user to data consumption ratio, the value density. The operator needs to incentivise the content providers and device manufactures to promote concepts with less data consumption and to develop technology and services to avoid the collapse of the system. Is it too late for this? No, I don’t think so given the options. Flat rates are needed and are welcome but all, including the end user, should be incentivised not to waste data as we today are aware of not wasting any resources in the society.

The operating system forest may turn into a jungle

Posted in UMI by Martin Vendel on December 1, 2009

First when the mobile phone became a device for more that voice and sms we started to worry about the different operation systems. Ringing tones and games were down loaded and we started to surf the operator WAP portals. The non voice services were taking off and the enthusiasm was high. This was though shown to be rather difficult to manage for the content providers and operators due to different operating systems and even variations depending on phone model and brand. Many versions of each individual game had to be developed and we had to keep track of exactly what phone model the end user had and sometimes even the software version of the particular phone model.

The kind of obvious solution to the issue was to try to agree on open standards and platforms and harmonise the different operating systems. One solution after the other was announced to facilitate the further development, new operating systems such as Symbian, operator initiatives on common device specifications, and languages such as Java. But Symbian ended up being owned by Nokia, operator initiatives never materialised fully and a language such as Java ended up in many different versions. Why? The strategic agendas of the device manufacturers and the operators as well as others in the ecosystem were not aligned and most players thought they had more to lose than to gain through cooperation. All claimed to supports open standards but the drive to differentiate and keep competition out was evidently stronger. This has resulted in a forest of operating systems significantly increasing the barriers to innovation and decreasing the efficiency of the industry.

Then Apple iPhone added further to the complexity, having a very positive impact on the take up of new services but being a vertical it added yet another operating system. Next solution that we then hoped for was Android, being a free software and open source and we finally seemed to get closer to making it all more simple for the developer community. But is this really the case?

Now device manufactures, such as Motorola, start doing their own versions of Android. Where will an Android quickly mutating itself get us? If we to the present complexity and all verticals add a number of versions of Android, similar but not similar enough, this will not make it easier but rather turn the present operator system forest into a jungle. A decrease in number of operating systems that all know would make life easier and spur innovation is obviously a too scary thought for most and will most probably not happen within the coming five to ten years. The only short to mid term solution that seems available is to add a layer on top of all existing operating systems creating a virtual common operating system interface. This will make it much easier to reach the mobile users and minimise the hassle created by the jungle of operating systems. This needs to be made in a way so that the overall performance is not set by the weakest system and so that we achieve a good enough overall user experience.