What’s out there?

Mobile platforms, languages and SDK’s

There are many mobile brands multiple possible platforms and SDK’s depending on the devices, so I couldn’t possible list everything here. Instead I wanted to provide a brief summary of what’s involved with each platform, devices it run’s on and places to start looking.

iOs

• Runs on iPhone, iPad
• Start at the iPhone Dev Centre
• Requires you run Mac OS X, $99 for the SDK, you must use the XCode IDE
• Write native apps using native Objective-C, or using HTML + JavaScript and a porting tool such as PhoneGap or Titanium
• Sell apps only through the iTunes App Store

Android

• Runs on a range of Android enabled devices including handsets from HTC, Samsung, LG and Motorola as well as a growing range of Tablets
• Start at Android.com
• Open source and can be built on Windows, Mac OS X and Linux, you can choose your IDE, plugins provided for Eclipse
• Write native apps in Java, or using HTML + JavaScript and a porting tool such as PhoneGap, Sony Ericsson Web SDK or Titanium
• Sell apps in the official Android Market, or unofficial stores such as Handango, GetJar, Handmark and SlideME

BlackBerry

• Runs on all BlackBerry devices
• Start at the BlackBerry developer zone
• Write native apps in Java 5, or write Blackberry Widgets for newer devices using HTML + JavaScript
• Use Windows, Mac OS X, Linux with an IDE of your choice including plugins for Microsoft Visual Studio and Eclipse. You can also use PhoneGap to publish Web Widgets.
• Sell apps in the official store BlackBerry App World or unofficial stores such as Handango, GetJar and Handmark

Symbian

• Runs on most Nokia devices and a range of others from manufacturers such as Sony Ericsson and Sharp (full list here)
• Start at the Symbian Developer Site
• Although apps can be written in a range of languages, (a guide can be found here) the best approach is to choose between Symbian C++ (which supports the widest range of devices), WRT Widgets (which are written in HTML/CSS/JS but don’t have much device access) or Qt (a cross platfrom C++ style language whose apps run on S60 3rd edition + Symbian devices as well as Windows Mobile, Maemo/Meego). You can also use PhoneGap to publish WRT Widgets alongside output for other platforms.
• The Nokia Qt SDK runs on Windows and Linux, with a beta for Mac OS X and can also be used for developing Symbian C++ apps. Alternatively both types of app can be developed using the Windows only Carbide.c++ or using custom tools. WRT Widgets can be written using the tools of your choice, or using Symbian WRT Toolkit which also now includes PhoneGap and runs on Windows, OS X and Linux.
• Sell Apps in the Nokia run Ovi Store or unofficial stores such as Handango, GetJar and Handmark

MeeGo/Maemo

• Maemo 5 runs on the Nokia N900 and Nokia Tablets, since Feb 2010 it has merged with Moblin to become MeeGo. MeeGo runs on a range on netbooks and will be installed on certain future high-end Nokia devices.
• Start at the Maemo page on Forum Nokia, from October 2010 the MeeGo project for devices should replace this location.
• Apps for Maemo are written using the Maemo 5 SDK with allows development in Qt (a cross platfrom C++ style language whose apps also run on Windows Mobile, MeeGo and S60 3rd edition+ Symbian devices). Apps for MeeGo will use the same language albeit with a slightly different (hopefully backwards compatible) SDK.
• Sell Apps in the Nokia run Ovi Store or unofficial stores such as Handango, GetJar and Handmark

Windows Phone 7

• Windows Phone development is now gearing up for the future and Windows Phone 7, a gorgeous interface for devices being launched in Q3 2010.
• Start at the Windows Phone developer portal and the Getting Started page on MSDN
• Apps for Windows Phone 7 are written using Silverlight for most apps or the XNA Framework for graphics based games.
• The free toolset Windows Phone Developer Tools Beta is required and runs on Windows Vista/Windows 7
• Sell Apps in the Windows Marketplace for Mobile

Others

• Java ME is installed on more devices than any other platform, but far from being ‘write once run anywhere’ can require different builds for each device. Write in Java, and publish on any of a huge range of application stores.
• AIR/Flash Lite is installed on a range of devices and platforms including Symbian and Android. It has limited access to device API’s and is best used for simple apps. There are multiple versions of Flash Lite with different capabilities, and all run in the rudimentary ActionScript 2 language. AIR can run on Android 2.2+ and runs ActionScript 3. All Flash Lite developers need a licensed copy of Adobe Flash, AIR developers can use the free SDK. Apps can be distributed in a range of app stores.
• Windows Mobile 6/6.5 has been replaced by Windows Phone 7. Despite there being plenty of devices out there still running the software, and no devices currently running Windows Phone 7, Microsoft seems to be discouraging developers from developing apps for Windows Mobile with all developer portals redirected to Windows Phone 7 info and tutorials. Windows Mobile apps were written in .net and require XP or Vista along with Visual Studio 2008.
• BREW is installed on devices by operators, with the majority being in the US and Japan and almost none in Europe. It has a huge install base though, 1400+ devices and 250 million handsets. Apps can be written using C, C++ and on some devices Flash Lite and Java. BREW apps are said to be costly and take a long time to produce due to expensive and slow certification processes. Apps can be purchased on the device using operators stores.
• BREWMP is a new OS aimed at low end devices and installed on HTC SMART, Samsung Reality and AT+T’s Quick Messaging Devices. Apps are written using Flash Lite, Trig, Lua or C. Application stores will be launched to sell apps, with the first launching in China.
• Samsung Bada is Samsungs new OS for Smartphones, and currently features on Samsung Wave and Omnia devices. Native apps are written in C++ with the option of using Web Technologies and Flash. Apps can be purchased through the Samsung Apps store which currently has almost 1500 free and paid apps.
• Palm and HP are still encouraging developers to create apps for webOS, the innovative but sadly ignored OS running on the Palm Pre. There’s plenty of support including a web only IDE and all apps are written using HTML/CSS/JS. Get started at the Palm Developer Center. Apps are purchased on the device.

What do other developers say?

These major reports have been released over the past 3 months, each the result of research conducted on various developers working in the mobile sector. Both are worth reading in full if you have a spare hour.

• The Appcelerator Mobile Developer Survey June 2010 is a survey of 2,733 Appcelerator developers who were asked which platforms they liked working with and which they would look to work with in the future. To be taken with a pinch of salt perhaps, given that Appcelerator developers are using a platform which makes it easy to publish to multiple platforms, perhaps giving them a bias and lack of experience in working with individual platforms.
• The Vision Mobile Mobile Developer Economics 2010 Report is a far more in depth study than Appcelerators, with developers chosen using a wider range of criteria, and asked a wider range of questions. Only 400 developers were questioned for this report however.

Preferred platforms – Android and iOs come out top

Apple’s platform came out high in almost all areas of the Appcelerator survey. Developers feel Apple’s platform has the best near-term outlook (78% for Apple vs 16% for Android with Blackberry and other platforms not getting a look in), and although 54% said they felt Android had the best long-term outlook, 40% still plumped for Apple.

Appcelerator asked developers what they would be interested in developing for, iPhone and iPad came out top, with Androids mobile and tablet offerings close behind. Blackberry and Windows 7 raised less interest, with very few developers interested in developing  for Symbian, Palm Pre or Meego.

Vision mobile asked what developers actually develop using and found Android came out on top (60%) with iPhone in second place closely followed by Java ME and Symbian. The most important reason for choosing a platform was market penetration and monetisation.

Market Penetration – Java ME and Symbian have the largest user base

Vision Mobile found that whilst developers prefer Android and iOS the number of people with Java ME and Symbian devices is far larger. Flash Lite, Java ME and Symbian all have a small number of apps available for large numbers of devices, in Q2 2010 Symbian had just 6000 apps for sale with a market share of 390 million units. iOS and to a lesser extent Android have much large numbers of apps for a smaller market share, in that same quarter Apple had 180,000 apps for 60 million units. Windows Phone and Blackberry have both low numbers of apps for low numbers of users.

One area of the market being under-served was localised apps, Vision Mobile found localised Spanish and French Apps counted for less than 3000 total apps in the crowded Android store.

Monetisation – App stores lead the way

Less than 5% of developers Vision Mobile interviewed used the once ubiquitous operator portals or preinstallation through operator to distribute their apps, the majority use app stores, with Apple’s and Androids the most popular. App stores offer a faster route to market, and faster time to payment.

The leading app stores are reputedly hard to make money from. Both are very crowded. Android has only 13 of its 46 countries able to accept payments for apps, users can ‘return’ apps within 24 hours without reason if they don’t like them. A standalone developer in Apple’s App Store can hope to sell an average 1000 – 2000 apps at $1.99. (Vision Mobile).

Ease of development – Android beats all others

Vision Mobile assessed Android, Symbian, Java ME and iOs platforms for ease of development. Based on their test data they concluded Android took the least time to develop and debug apps with, followed by Symbian, Java ME and with iPhone coming in last. I personally wasn’t totally convinced by their testing methodology, so I would suggest you read this section of the report and study the data to draw your own conclusions (Appendix 2). They also only focused on 4 platforms – 2 of which (Symbian and Java ME) are clearly on the wane.

My conclusions

Android,  iOs and Web are the best skills to sell to others

As a freelancer or employee looking to sell your skills to others, Android, iOs and Web are safe bets to keep you in pocket.

Android and iOs are popular platforms and I predict there will continue to be a demand for developers of these platforms. As the field gets inevitably more crowded you can differentiate yourself by being reliable and producing good quality code (not as common as you might think). Releasing your own apps to impress potential clients (but not to make money themselves – see below) would be a good idea at this point.

Gaining experience in a range of Widget platforms will allow you to transfer web skills between platforms and enable you to offer a range of platforms to an employer. With web you can target Symbian, Blackberry, iOs, Android and even Palm. Familiarise yourself with each platforms own widget set up and install process as well as the PhoneGap tool.

To make a quick buck or gain fame as a single developer focus on untapped markets

I feel strongly that lone developers wanting to make money or gain fame from a mobile app should team up with a great designer, but it has happened in the past. The key is to avoid over populated areas where it will be very hard for even the best teams to stand out, unless they have an existing brand.

My feeling is the iOs and Android stores are far too crowded already, but if you must enter the iOs or Android stores, think about targeting a niche. As mentioned before, localised apps is an untapped area for many…

A great way of making money from mobile apps looks to be to target the long tail of platforms. Samsung Bada is out but with few apps yet released, Windows Phone 7 is unreleased giving plenty of time to be one of the first apps to launch in that field, and for slightly busier stores but perhaps safer long-term skill investments get into Symbian Qt and BlackBerry.

To launch an rich app for an existing brand test the water with web, then go multi-platform

If you have an existing brand and want to launch an app which doesn’t require lots of device API’s I would launch on web platforms first. This will be cheaper in terms of development as the skills are often easier to find, and you can write once and publish on multiple platforms easily thanks to tools like PhoneGap.

Once you reach the limitations of web, assuming there is a demand for your product, you should release in iOS and Android stores to get the app lovers interested, and then consider Qt for Symbian (ensuring the widest range of future devices) to target the Nokia masses and BlackBerry for the young elite and business customers. Consider Windows 7, although be aware of the risk it might not be a success.

To tackle emerging markets or for simple apps for non smart phones consider Java ME, BREW and Flash Lite

Most people don’t have a phone that can run the platforms we’ve discussed, with the exception of Symbian. Most people have a phone that can run Java ME. If you want to tackle the non smartphone market, you will need to start dabbling with Java ME. For the US markets consider BREW also, and try BREWMP for newer low end devices. The nicest interfaces can be produced using Flash Lite, but this prevents you from accessing the device API’s in the same way, limited memory can cause problems for you, and less phones run Flash Lite than run Java. Expect to spend lots of time and money tailoring your build to each device, and consider that many people without smartphone’s don’t even realise they can install apps.

Open to comments, and coming next

With my work at ribot I get to play with a wide range of platforms, and I’ll be experimenting with more of those listed over the coming weeks. Expect more info on the various languages and my thoughts on differences between them. Please let me know your opinions below on these ideas.

This is a smaller event than usual but it will be a good one.

Jan Srutek: Information visualisation – more than just eye candy

Information Visualisation (InfoVis) is a domain that has gained a lot of attention recently. The mass media seem to have fallen in love with information design too – feeding us with an infographic a day. But despite the great potential for representing data visually, a lot of the potential is currently being wasted. Just because something is graphically depicted does not mean it is useful and usable. The challenge is to design InfoVis tools and infographics that will allow its users to actually gain valuable insights from the data. The talk will present some of the advantages of visual communication, along with the challenges for design and evaluation of InfoVis tools.

Jan Srutek works as User Experience Consultant at User-Centred Design agency Flow interactive in London.

From Jan’s bio on Flow’s website: Before joining Flow Jan worked as an Information Architect designing Web2.0 online community services and corporate websites. He also has experience as a Business Analyst for Online Marketing at SITA, a global IT company operating in the air transport industry. Since at Flow, Jan has completed research and design projects for a number of clients including Transport for London, Tesco, HAYS, GAVI Alliance, and Standard Life. Apart from his capabilities in user-centred design methods, Jan is interested in Information Visualisation, Web Analytics, and Search Engine Optimisation. Jan has a BSc in Information and Communication Management and an MSc in Human-Computer Interaction from University College London.

Twitter: @JanSru

Normally we think of bad design as consisting of laziness, mistakes, or school-boy errors. We refer to these sorts of design patterns as Antipatterns. However, there’s another kind of bad design pattern, one that’s been crafted with great attention to detail, and a solid understanding of human psychology, to trick users into do things they wouldn’t otherwise have done. This is the dark side of design, and since these kind of design patterns don’t have a name, I’m proposing we start calling them Dark Patterns.

I’m preparing a short talk on this for the UX Brighton Conference in September, and I need a bit of help coming up with some examples. Here’s a taste of what I’m talking about:

• Low cost airlines that put insurance in your basket without you asking.
• Social networking sites purposefully make it hard for you to shrink your social graph or move your content into private realms. I’m looking at you, facebook.
• Email sending systems that require you to log-in (using a long forgotten password) in order to unsubscribe. (This is actually forbidden by the revised CAN-SPAM 2008 rules, but it’s widely ignored)
• Systems that ask for your email / twitter credentials on the grounds of finding you friends, but then send messages as if they are directly from you, without your express consent.
• Supermarkets (in the real world) that prevent you from comparing products on price, by putting items in different sized bundles.

Can you think of any good, contemporary examples to go with this list? Add your suggestions in the comments below. I will, of course, credit you in my slides.

To be clear, I’m not looking for outright scams (which are clumsy and easy to identify), I’m looking for techniques used by above-board products and services that trick users into doing things.

Well, I’ve just noticed they’ve iterated the design, and guess what? It’s in line with many of your suggestions.

First let’s look at the old design:

Original times.co.uk paywall design (May 2010)

Your comments included:

• Simon Thulbourn: “Remove the giant image in favour of clearly defining what the user gets in return for their money”
• Paul Gordon: “…the registration page itself should have pricing options and free trial information”
• G Mulder: “what is this huge helicopter image trying to tell me? That SAS is gonna come and get me if I don’t sign up?”
• Johan D: “…neither the content nor the period is clear”

Now, check out the new design:

New times.co.uk paywall design (July 2010)

They seem to be iterating in the right direction, though some of your best suggestions haven’t been incorporated (clear pricing, lazy registration, teaser excerpts, etc), and bizarrely, they don’t mention the free trial until after you click through to sign-up. But let’s not forget, a UX designer’s job is never done. By keeping a close eye on their KPIs and iterating regularly, it’ll be pretty hard for them not to improve this… Right?

Danny and James have really pulled a rabbit out of a hat with the upcoming UX Brighton 2010 Conference. What a great line-up!

• Rory Sutherland: you may have seen his TED talk, Rory is Vice-Chairman of Ogilvy Group UK.
• Eric Reiss: heard of him maybe? He’s one of the most well known figures in European UX.
• Claire Rowland: an accomplished Design Researcher, Claire has spent the last couple of years building up Fjord London’s UX research offerings.
• Jeroen van Geel: Interaction Designer and founder of the hugely popular Johnny Holland magazine.
• Julian Hirst: Head of Electronic Ink in the UK (They’re UX heavyweights: think financial trading UIs and fraud surveillance visualizations.)
• James Page & Sabrina Mach: owners of Fera Labs, makers of Webnographer, the remote unmoderated quantitative user research platform, and also co-organizers of this event.
• Plus another mystery speaker: who could it be?

Early bird tickets are only £69, but you can get a further 15% off if you follow @uxbri and tweet about the event.

Get your early bird ticket* now: http://2010.uxbrighton.org.uk

*£69 rather than £99

Many companies use analytics as a starting point for user research but get stumped when it comes to gaining genuine understanding about why users do what they do.

For July’s event we aim to show how applications like Google Analytics can be used to both formulate and answer questions. We’ll also show how these quantitative techniques can be combined with qualitative methods such as traditional user testing.

Talk 1: Alex Tarling: Combining analytics with user testing – a case study

What attendees will learn

• Why would you want to combine analytics and user testing?
• The benefits of triangulating between quantitative and qualitative sources.
• Some practical tips, how to get started and details of approaches.

Alex is a freelance user experience consultant, focussing on research into how customers use technology in their everyday lives. Particular domain expertise in mobile, e-commerce, public sector, education and healthcare.

Alex has worked on projects for Intel, BBC, Nokia, Orange and the Open University among many others.

Talk 2: Phil Guilfoyle: Get under the hood of Google Analytics

Phil will briefly explain some of the key principals behind how Google Analytics works. Then provide step-by-step walk throughs for top analytics tasks. Finishing with setting up and distributing reports to stake holders.

Phil specialises in Conversion Rate Optimisation, Usability Testing, AB Split Testing, Multivariate Testing, Behavioural Modelling, Information Architecture and Wire Framing. He is the founder of Online Logic.

When you visit a top UX agency in the heart of London, what do you see when you walk in the door?

Lots of empty desks. Not because of layoffs, but because most consultants are working in labs or doing off-site consultancy projects. This means there are a lot of overheads for space and for kit that’s just sitting there, unused. When you hire a UX agency, this is one of the things you’re paying for, and it’s not money well spent.

Another problem is billing efficiency. UX research engagements can be very short – sometimes as little as two or three days (for expert reviews), usually a couple of weeks long (for face-to-face user research and analysis). The shorter the project, the bigger an issue you get with gaps between projects. In an agency, resourcing becomes a crazy Tetris game that’s almost impossible to win. Consultants end up with a lot of bench time, where they’re waiting around being paid but not bringing in any money. If you hire a UX agency, you’re having to cover this cost, and again, it’s not money well spent.

Also, agencies often try to include a research assistant and some senior consultant time into the costing. The fact is, if you get an experienced consultant, they usually don’t need support from anyone else. UX research is often well suited to “lone wolf” consultancy.

Enter the UX freelancer:

• She’s nomadic, and works on-site with clients. No overheads there.
• She has Morae and a couple of laptops. She can set up a research lab in any room in seconds, no need for an expensive facility with half-silvered mirrors.
• Having worked at an agency for a couple of years, she has all the credentials she needs.
• She’s able to charge less than half what an agency charges, and still make a very comfortable living.
• What’s more her low rate also means she’s more cost effective to hire for long engagements – which means better billing efficiency for her.

Today, clients seem to be waking up to the value of freelancers, where historically they seemed to be more risk averse, gravitating to well known agencies. Just last week I was talking to Be Kaler Blake (Director of Futureheads, a London-based UX recruitment agency), who described the market for UX freelancers as “buoyant” – an unusual word to hear in the midst of a worldwide financial crisis.

So, how do agencies feel, knowing that every time they train up a new consultant, they are paving the way for a new freelancer who may ultimately become a competitor? And what does this mean for the industry in general?

One way or another, things are changing. Charging £10,000+ for single round of usability testing used to be such an easy way to get by.

The trouble with being a User Experience specialist is the amount of wall space you need. In an ideal world, you’d set up a war room for each project, where all your materials can stay permanently stuck on the walls. Most of the time, that’s not possible – other people in your company need to use the meeting rooms too! This is where the need for portable wall space comes in. Here are some suggestions:

Foam board (aka Foamcore)

• Sheets of white cardboard with polystyrene sandwiched in the middle. Normally used by artists to mount photos, etc.
• Strengths: light, but rigid. Reusable. Check out this neat ikea-hack by Daylight (pictured above) if you need a stand for your boards.
• Weaknesses: not cheap but stationary shops may have slightly damaged items at a reduced price. (A small dent in the corner makes a mounting board useless for mounting art, but it’s still great as a portable wall surface for your needs.)

Butcher paper (aka Kraft paper)

• Huge rolls of thick paper (traditionally used by butchers to wrap meat).
• When you get kicked out of the meeting room, you roll up your paper and take it with you. It’s easy enough to unroll and stick to another wall – provided you have the space!
• Strengths: it’s damn cheap
• Weaknesses: heavy, so needs strong anchoring to the wall using tape or loads of blutack. This can ruin painted walls.

Pattern cutting paper

• It comes on a huge roll, like butcher paper, but with a dotted grid marked on it. (Recommended by Paul Thurston of Think Public)
• Strengths: Cheap, and the grid is helpful for sketching UIs
• Weaknesses: It’s heavy, like butcher paper.

3M “Self-Stick Wall Pads”

• I’ve never used these, but they are basically giant, flip-chart sized post-it notes.
• Strengths: no messing around with blutack / etc
• Weaknesses: surprisingly expensive.

So, what do you use for portable wall space? Suggestions in the comments, please!

The Frankfurt Kitchen in use (from Genius of Design, Ep.2)

What’s amazing about it is the huge amount of research and exacting attention to detail that went into the design. One of the main driving principles was efficiency, and Lihotzky conducted extensive time-and-motion studies to optimize workflows in the kitchen. She was motivated by the desire to make life easier for people.

A kitchen-usage time and motion diagram (from Genius of Design, Ep.2)

However, here’s the bit you’re going to love if you’re a UX researcher – in her zeal to make the kitchen efficient, Lihotzky completely forgot to consider real-life context and the social implications of the design. Quoting from Genius of Design Episode 2:

“10,000 of her kitchens were installed in a large-scale social housing development in Frankfurt [...] But despite all of the well laid plans, many first time users were apparently baffled by the layout. They found the inflexibility of the design frustrating, and they proved to be disappointingly undisciplined when it came to using the carefully labelled food bins. [...] When people actually went in to use the Frankfurt kitchen [...] people found them very cold”

Furthermore, the carefully designed food bins (intended for flour, rice, etc.) were easily reached by small children, making spillage very common; labels were pre-printed on the bins, meaning that owners inevitably ended up with all the wrong labels on things; and worst of all, the kitchen was so small that only one person could be in it at a time, making the kitchen user feel socially isolated from the rest of the household.

Designated storage bins in the Frankfurt Kitchen. Source: Wikipedia

Lihotzky’s ideals were born out of the Tayorism movement – which was all about making factory floor & repetitive labour activities more efficient. In that sort of workplace setting, it is possible to prescribe user behaviour (i.e. “Do this and you get paid”). Where Lihotzky went wrong was that she assumed you could prescribe user behaviour in the home. With hindsight being 20:20, it’s not surprising this idea was doomed to fail.

It’s interesting to consider that an ethnographic study would have uncovered problems with the design very quickly, while dogfooding would most likely have failed to uncover the problems. Why? Because Lihotzky and her colleagues would have been so heavily engrossed in the ideals of the design, they would have used it in a highly disciplined and rigorous way – unlike their target users (german working class families who had just emerged from the ravages of WW1). This is a lesson worth remembering – dogfooding is today a very popular method in big corporates yet it can be disastrous if there’s a big difference between the test users and the real users. Google, for example, used dogfooding in the design of Wave and Buzz, and both of them had adoption issues when launched.

If you’ve seen the off-beat norwegeian Swedish comedy Kitchen Stories, the time-and-motion diagram above might look familiar to you. That’s because Kitchen Stories took a lot of inspiration from Lihotzky’s work, as you can see below. Personally, I love the idea of writing a comedy about field research and time-and-motion studies – it’s such a geeky idea.

A still from the movie Kitchen Stories