Google: Project Ara
BY ANDREW SHEEHY, CHIEF ANALYST - 07 May 2014 Comments
Google describes Ara as a hardware platform that “challenges the conventional world view of what is possible” with smartphone technology. According to information shared by Google at the first Ara developer conference in June 2013, Project Ara has two strategic goals
- Ara is intended to be a
ecosystem for smartphone hardware that, in terms of the speed and degree
of innovation, will be more like the ecosystem that already exists for
- Ara is seen as the only
way to deliver the mobile internet to the next 5 billion people.
Ara was conceived and developed within Motorola Mobility, the company Google acquired in May 2012. Google first announced Ara in October 2013. Following two more developer conferences which are planned in 2014 Google is planning for the first Ara smartphone to be launched in January 2015.
An Ara-based smartphone consists of a set of ‘slide and click’ interchangeable modules each of which performs a core function, such as Wi-Fi connectivity, battery, screen, camera/video, thermal imaging module, memory etc.
Google thinks that users will want to participate in in creating their own smartphones: users will order their smartphone online which will then be delivered as a self-assembly ‘kit’.
|Title:||Google Ara – Variants and parcelling scheme|
The smartphone hardware consists of two-sided ‘endoskeleton’ (which will be supplied by Google) onto which all modules are attached. Modules are held within the endoskeleton using an system of electo-permament magnets which allows modules to be set to ‘locked’ or ‘remove’
|Title:||Google Project Ara|
|Sub Title:||Styling Options|
Google plans to deploy on-demand 3D printing technology on a mass scale which will allow users to customise modules using snap-on covers that will be available in a range of colours, materials and finishes.
|Title:||Google Project Ara|
|Sub Title:||Module Configurations|
- Developers will get a
worldwide license to use Google’s IP to develop and test Ara modules;
- Developers will also get a
license to sell and distribute Ara modules;
- There is no cost to
participate in the project although it is likely that, as with Android,
Google will sell services to support develops and help them with module testing,
validation and approval;
- Developers keep 100% of
the revenue they make from selling their modules;
- No fragmentation:
developers are not allowed to create a rival ecosystem that uses Ara IP;
- Developers who do not like
what another developer is doing can take legal action against that
developer, provided they do not seek a legal injunction that would stop the
developer doing whatever they are doing;
- Google reserves the right
to suspend a developer’s license in the even that the developer attempts
to take legal action against Google in respect of Ara;
- While developers will own any IP in the modules they develop, in return, developers grant Google a license for any IP they have developed which extends beyond their module into the Ara platform more generally.
We see Ara as less of a way for users to personalise their smartphones and more of a way for Google to change the structure and operation of the smartphone industry.
Google apparently thinks the current pace of innovation in the smartphone market could be higher: smartphone vendors are not introducing new features and capabilities fast enough and they are not focusing enough on low-end devices that can serve the “6 billion users who do not presently have access to the mobile internet.”
Certainly, Apple, who accounts for over 80% of smartphone hardware profits, has minimal interest in producing the sort of stripped-down, reduced functionality devices that are needed to serve consumers that will be increasingly price sensitive as the market expands downwards into the lower layers.
Google’s future depends on delivering the mobile internet to as many individuals as possible and so the company is happy to take bets of projects than might increase the smartphone market by, say, 1 billion users – ideally more – while putting Google in the driving seat for how services are delivered to those users.
We think that Ara’s success or failure will depend on the answer to three key questions:
- Whether a change in structure
of the smartphone hardware market could substantially increase the pace of
innovation, at both the high end and low end of the market?
- Whether Ara is capable of
effecting that change?
- Can one platform, Ara, simultaneously
meet the requirements of users in developed markets who value the ability
to customise and personalise their smartphone and users in developing markets
who mainly just want a phone that provides access to the internet.
If the answer to all of these questions is ‘yes’ then Ara could change the smartphone hardware market in a similar way as how Android has transformed smartphone software market. Alternatively, if the answer to any one of these question is ‘no’ then Ara will have to accept a future as serving a part of the market rather than the whole market, as Google envisages..
The smartphone device business is increasingly concentrated around Apple, Samsung, Huawei, LG Electronics and Lenovo who are together responsible for nearly all of the innovation in the market, along with 60.7% of worldwide unit shipments (figure relates to 2013, source: Gartner).
Generally, innovation is the result of competition and so more smartphone vendors should theoretically mean more innovation.
But when one takes into account the capital needed to cover R&D, tooling, marketing, manufacturing commitments and retail placement the barrier to enter the smartphone vendor market is several billion dollars.
Even then, the competitive landscape is ferocious with Apple and Samsung controlling the market and using very aggressive legal actions to target rivals who threaten their market position.
It is possible to enter the smartphone market using a white-label / OEM strategy but while this might expand the market, it does not advance the market in regards to innovation – which only happens when vendors commit significant capital to introduce new features.
With the current market structure, it is extremely hard to see how any company board would approve a business case that proposed entering the smartphone vendor market in order to compete head-on with Apple and Samsung. As things stand, investors would likely regard this as a suicide mission.
If we move up the value chain and think about smartphone component vendors then things are different: here we see many more vendors, lower entry barriers, and capital commitments for entry that might, for some types of component, be as little as a few 100 thousand dollars. We also see a more favourable competitive landscape (e.g. rival competent vendors tend not to routinely sue each other).
Many component vendors might see opportunities to improve on key smartphone components such as a new Wi-Fi antenna, a better batter technology or a display technology that consumes less power. But as the market stands today, the only route to market for these vendors is to persuade one or more of a very small number of smartphone vendors to adopt their innovation. If this is unsuccessful then the innovation is going nowhere.
Ara will, in principle, provide a route to market for these vendors by allowing them to pursue their innovations independently of Apple, Samsung and the other leading smartphone vendors. Not only that, they would have control of their own distribution, retail and marketing strategies.
With Ara, the route to market for these vendors would be decoupled from the mainstream smartphone vendors.
Component vendors would, therefore, be able to develop a parallel direct-to-consumer business by using an open platform where only barrier to reach the end-user market would be passing certain technical criteria which would apply for each module. Nobody would get involved deciding whether a particular innovation would be attractive to end users or not: that risk would be taken by the module developer, not Project Ara.
There seems little doubt that Ara would provide a new, more attractive route to market for existing smarphone component vendors, plus a range of new entrants.
But the number of developers participating in the Ara ecosystem would be far lower than the number of developers participating in Android. The reason for this is that even a simple hardware module would require the investment of a few hundred thousand dollars, going up to several million dollars, and some developer swould have to plan to invest a lot more than that.
Although smaller-scale Ara module development projects could be funded by crowdfunding platforms like Kickstarter, there would be no situations like those with Android where a lone developer writes a smartphone app in a few weeks which is then hugely successful, and who then sells the app for USD 30 million to a corporate player (e.g. Summly).
However, in principle and in time, we can see Ara working very well.
Specifically, other than the mode envisaged by Google - which is where module developers design and sell modules their modules directly to consumers, while retaining 100% of the revenue - we see two additional possibilities:
- It would be possible for
several developers to work together to create a ‘whole’ smartphone that
was aimed at one market segment. So instead of just developing and
marketing a single module a developer could also get involved in
co-development projects aimed at developing a whole smartphone, which
might then be marketed by a branded player who wanted to enter the
- Building on (2), we see that the problem of the ‘next 6 billion’ mobile internet users could be partly addressed by an approach that took a number of key modules and then integrated those modules into a single, integrated smartphone – which would reduce the size and cost of the device to less than what the modularised version would achieve: we do not think that a modularised approach will deliver on the sort of price points that the lower bands of this market will require. This specific usage is, however, not explicitly covered by Google’s licensing terms and so it is unclear what role Google would play in the commercialisation of such a device.
One of the major factors that contributed to the success of Android was very strong initial demand: following Apple’s announcement of the iPhone in 2007 rival smartphone vendors were desperate to produce a rival device. Blackberry was not available for licensing, Microsoft Windows was not attractive and Nokia’s aging Symbian was seen as being paralysed in regards to further development. That left new smartphone operating systems such as MeeGo (Intel) and Maemo (Nokia), both of which were unproven, uncompelling and therefore seen as risky bets.
When Google announced Android – which would be open and free – the company had to deal with a flood of interest from mobile handset vendors, all of whom were desperate to get back into the market after having been depositioned by the iPhone.
We do not see the same level of demand for Ara: for one thing, completely unlike Android, smartphone vendors will be either strongly opposed to the project or unsure about about how they should participate in the programme. Secondly, there is little visible sign of component vendors screaming for a platform like Ara.
Another important difference between Ara and Android is that Android was ‘just’ another mobile operating system: the only way it challenged the industry was that it was being offered for free: Android slotted neatly into an existing industry structure and it was not a blatant attempt to change the structure of the industry.
In terms of the barriers facing Ara, we think that the mobile network operators will be generally supportive: these companies are mainly interested in selling network services and they do not really care about how they achieve that goal. The only concern they have on the handset side is to ensure that there is enough competition so they do not have to accept deals on a ‘take it our leave it’ basis which has been the case when one handset brand has accounted for too large a share of the market.
On this basis, Ara does not face any of the ‘external’ market barriers that are are serving to block the development of some of Google’s other high profile projects, such Google TV and Google Wallet, for instance.
Nevertheless, on balance, we think the initial demand for Ara will be an order of magnitude less than that which greeted Android:
- There is nowhere near the
same level of pent-up demand for Ara as there was for Android;
- It will take time for existing
handset component vendors and new entrants to understand the platform, to
develop business plans and raise the necessary capital to enter in the ecosystem;
- Existing component vendors,
who are best placed to develop the initial modules, are used to operating
in a B2B setting: not all of these players will decide that they have what
it takes to develop a direct-to-consumer business;
- Distribution and retail
infrastructure will need to be established: mobile operators, for example,
will have to learn about the platform and how to sell Ara mobile phones,
alongside side regular smartphones: while modules might be sold via Google
Play, mobile network operators still control the sale of network services
(e.g. SIM cards).
Ara is a very interesting project that could have a major, beneficial impact on the structure and operation of the worldwide smartphone market.
But there are major differences between Ara and Android. In comparison with Android, Ara has:
- A much lower level of demand: handset vendors are not falling
over themselves to get Ara, as they were with Android and there is no sign
that component suppliers are desperate for a direct route to market;
- A higher level of entry capital: There will be no cases where
an 15-year old Ara developer can knock out a module design in a few days
and then enjoy a big success, as has been the case with Android;
- A fundamentally higher level of friction: As soon as any
hardware is involved, the pace with which the market can develop is
These differences will mean that change will not happen as quickly as was the case with Android.
As regards to the project’s main goal – which Google says is to deliver mobile internet to the next 6 billion users – we think Ara is way out of whack with what what is really needed, which is a smartphone that costs $10 to produce – not $50, and certainly not $100.
At such a low price point users could not care less about colours, self assembly, or other gimmicks: they just want a device that works. If Ara is really going to deliver on its main objective then Google will have to adapt the platform and acknowledge that there are at least two market segments: existing smartphone users who are seeking value-added features, and those who are seeking their first smartphone, but only have $20 to spend.
At this point, if the objective really is to deliver an entry level smartphone for $10 (and there is presently no roadmap that can achieve this), then it is very clear to us that an integrated product approach will get there well before a modular approach like Ara.
Providing 6 billion additional users with mobile internet is mainly a hardware problem that involves rolling out network infrastructure and producing extremely low cost devices. As it stands, Google’s project Ara acknowledges the problem but only hints at the solution.
Our analysis is that Ara is a welcome, highly innovative project which may go on to be very successful, but in its current form, it will not catalyse change to anywhere near the same degree as Android.
Contact Andrew Sheehy at firstname.lastname@example.org or follow him at twitter.com/sheehy_andrew