The myth of mobile design


I must admit that I have not got the exact title for this blog .. I even thought of calling it Intelligent design for mobile devices but gave up on that idea ;)

Anyway ..

I hope I can covey what I want to say accurately

Most discussions on ‘Mobile design’ start with the aesthetics and the user interface. They focus on the specifics like buttons, icons, screen size, camera etc. These factors are important but in my view commoditised and any new features in this realm are likely to be quickly copied. The device itself also grows more complex and hence also the design

However, if you take a big picture view – then you would come to these features only AFTER considerations similar to below

What I am trying to say is:

a) Mobile design is more about how it works than just about how it looks. This principle comes from Steve Jobs(Inside 5teve’s brain leander kahney)

b) If we accept the above premise, then we need a different starting point for designing mobile devices

c) Depending on the starting point – it may never be possible to accurately ‘design’ the device in advance – because the ratio of predictable use cases vs. unpredictable and uncontrolled use cases are increasing as I discuss below.

Note that the operative word is ‘design in advance” – and hence what I call as the myth of mobile design. So the design of a mobile device may be a ‘platform’ and the goal should be to allow it to evolve

Starting with three types of mobile devices based on how they work

I see three types of devices based on how they work:

a). The consumer electronics device(ex Sony walkman) which cannot be modified by the user at all

In this case, all use cases are predictable

b). The consumption device with an API which can extend the functionality. The device has an API (application programming interface) but the device itself does not (mainly) create and share content. Ex the PC and to a certain extent – the iPhone. Here, the number of unpredicted use cases increase but complexity is managed

c) The creation and communication device – this type of device is intended to primarily create and share content. Here, developers can extend the device and the users themselves can create and share content. This scenario has some predictable use cases but will have many more unpredictable use cases

Knowing this classification, then the next step would be

a) To decide the customer segment the device serves

b) To decide a series of processes that the user undertakes

c) And most importantly, to make a conscious decision about how many use cases are predictable and controllable and how many are not controllable and therefore not predictable. This needs deep technical knowledge of the whole ecosystem and not just the UI. Specifically, it needs an understanding of the device stack, the integration within the stack, the network , the business models and the ‘friction’ within the ecosystem for the user which creates unpredictability for the user experience

This approach forces us to think away for buttons and widgets and specific technology like location based services in isolation – and to work with the user(who has a choice)

The designer can ‘design’ most accurately in case (a). In case (b) the designer can set out a design but the APIs can extend it to a certain extent in a controlled manner but note that the customer’s experience is still controlled because the device is primarily designed to consume content

Case (c) is different because – two things change – developers can add functionality and users can create and share content. So the designers goal in this case is to still work with use cases but the ratio of predicted use cases vs. unpredicted(and uncontrolled use cases) is relatively lower

Some examples of the above ideas

Let’s look at some examples:

a) the iphone is a case of excellent design – but leaning to consumer electronics rather than creation and sharing. Ex – initially the iphone did not support mms. So iphone is case (b)

b) Nokia on the other hand genuinely tries to enable the content creation and sharing process. So case (c). But increasingly such a device is hard to ‘design’ in advance

c) Android is a brilliant case of integrating the software stack – hence could lead to better design in general

d) facebook applications like ‘throwing sheep’ are unpredictable use cases. They are adopted by users but the proliferation of applications on facebook make the service lean towards case (c) i.e. its primary function is communication

e) The predicament of ignoring integrated design comes from that (in)famous statement I once heard – ‘yes 80 percent of out customers have sent an MMS message – but how many have sent two!’ To design the process of sending and receiving mms messages – you must be able to manage many aspects(which are mostly not in your control) – price plans, device capabilities etc etc

f) LTE will create a whole new group of non phone devices – with different design paradigms

g) I am a big fan of the blackberry which is a well designed device for a specific purpose. In fact, a lot of this blog was written on the blackberry – drinking pints of apple juice :) having

said that -

The blackberry device I use is wifi enabled. When I am browsing at home, I want to us wifi. However, I don’t know if the device will choose the wifi connection(which is free) or the cellular(very expensive for browsing). So I disconnect the cellular connection.

Which is a behaviour pattern that designers will never know. In other words, applications are not bearer aware, there is no clear way to know which bearer will be chosen(especially since there are cost implications) and the choice is not communicated to the user.

h) Like many people who start from a Unix background, I started working on the vi editor . If you have never known what it is – vi takes design to new levels of minimalism. When you start it – a cursor blinks at you. You think its loading or something .. And after a few mins – you realise that ‘this is it’!

You then press a few keys(F1 for help!) And you realise that they don’t work either. Then you make a decision to either abandon it or to explore it. If you stick with it – you find that it is an incredibly powerful editor designed for relatively advanced users and it is designed to evolve through its powerful set of features like macros

i) The LG versa is also a device designed for a unique set of features including a 3D user interface and haptic feedback. So, it takes a ‘maximalist approach’ i.e. provide the best of everything and a richest possible experience to the user – also a valid philosophy which starts with consumer electronics but blends in other elements as well

k) Tomi Ahonen discusses an example of an SMS oriented phone – also a good example of a communication oriented device

l) The issue is complex since the designer does not REALLY know what software the device is running. For example – the user’s entire mobile experience could be based on Opera Mini which they downloaded. Capability exchange software would help here(ex use of XMPP for Android)

k) RIM Co-CEO calls the new reality for Blackberry storm

To conclude

So, to conclude – to me there is a myth of mobile design since many of the factors which constitute a truly excellent design are currently not in the control of the designer. This can be mitigated somewhat by making an upfront choice about the type of device, minimizing and by choosing a niche customer segment.

The minimalist alternative is also valid. But whichever way we look at it, design has to extend far beyond the existing emphasis on UI and has to consider deep integration especially going forward when devices will be more complex with LTE and beyond and services will be deeply integrated into devices

When we look at Mobile design – we often work with one or two extremes: Either that it’s all about the UI or that the ‘designer’ can control/influence the user in most cases.

The first is a common fallacy. The second is a myth.

At the moment – the designer has limited influence on many factors – as I list above. Hence a deeper understanding (and indeed evolution) of mobile applications is needed – ex Capability exchange, Bearer aware apps etc etc.

So, we will need to take the idea of Mobile design to higher standards than we see now

As usual, comments welcome


  1. Ajit, this an interesting thesis. The title suggests that what I do doesn’t exist; the article essentially lays out a case for interaction design.
    Many design disciplines, including industrial, interaction, and architecture, start with a deep understanding of who will be using the product or system. We’ll design for content creation. There are more options than “minimalist” and “”maximalist”.
    You allude to “knowing the customer segment” for whom you are designing. We couldn’t agree more … but we need to know their behaviors. Our research focuses on needs, goals, and behaviors of use – not purchase.
    We will turn down a project in which we are just doing visual presentation. System behavior is so critical, and few marketers or product managers know well how to design. (by “few” I mean “less than 20%”)
    To put it another way: never, ever, hire me to design an icon for you. But if you want your system to behave well, we’re a great match. But only if your organization is ready to really rethink your whole service.
    So this isn’t about the myth of mobile design. It’s the fallacy of design as something to bolt on to the end. We refer to this as putting lipstick on a pig.

  2. Ajit Jaokar says:

    Thanks Barbara. I did say that I struggled with the title. Having said that, I have not seen the argument laid down in terms I describe above i.e. I see way too much emphasis on the UI – and I think especially going forward a deeper knowledge of the holistic ecosystem and integration is necessary(example bearer aware etc as I mention). Thanks for the comments kind rgds Ajit

  3. Ajit, What I hope your takeway is: the UI is not just the visual layer. It is also behavior. And beyond that, the UX includes features, ecosystems, business models, and more.

  4. Ajit Jaokar says:

    Barbara, The key takeaway is -
    Design incorporates more than the UI and furthurmore, the designer does not control all the factors.
    Hence, a rethink is needed with an upfront decision as to the type of device i.e. consumer electronics device, consumption device with an API or creation and communication device.
    I could not think of a fourth type of device so far. kind rgds Ajit