App craftsmanship

Bespoke artisan furniture; we can all appreciate it, even if we can’t afford it. There’s a place for Ikea, mass production and flat-pack, but there’s also a place for craft.

It’s similar with software. There’s a phrase “code is craft”. The mobile app revolution is perhaps one of the best examples of that philosophy.

We’re not talking business cases, global delivery, multi-platform development or sprawling requirements here. We’re talking software developed by a handful of people, or maybe even a single person, building the absolutely best thing they can. We’re talking small nuggets of software perfection.

I can think of two good examples of this: Tweetbot (Twitter client) and NetNewsWire (RSS reader). Both of these apps are something rather special. They are also the two most-used apps on my iPhone. Every day I use them to keep abreast of the world around me. And they are both little nuggets of perfection - created by software artisans.

The UI design of these apps is subtle. They are both designed to fit within iOS in a respectful way. They don’t attempt to reinvent the wheel, but use standard UI capabilities, extending only where useful. Branding is exhibited through subtle use of colour, icon design and animation. Nothing is overt, everything feels natural - as if it had been desiged by Jonny Ive himself.

Discrete might be another way of explaining this kind of design philosophy. Being discrete and subtle sometimes takes a lot more effort and thought than being brash or overt.

Take the way that NetNewsWire subtly colours the navigation bar with the theme of the news article’s source. We don’t see a brash logo, but rather just a subtle colouring. As a result, everything feels consistent. But articles are also differentiated in a discrete way.

And take the way that everything in Tweetbot scrolls at 60fps. Consistently smooth inertia scrolling at 60fps takes design effort. That effort isn’t obvious, but it’s there. This kind of attention to detail is needed if you want an app to be something special. It’s not easy to measure and justify the effort involved. But it matters if you want to delight your users.

Sweating these details matters - because it creates something a little special. And when something’s a bit special, it’s obvious - in the same way that a hand-built piece of furniture is obviously special.

I use NetNewsWire and Tweetbot not just because they get the job done, but because I appreciate the thought that has been poured into them. Using apps like this is pleasurable.

In contrast, some other apps annoy me every time I use them. Their brash colours, their clumsy styling, their inept attempts to re-interpret standard UI controls for no purpose, their stuttery scrolling. Some exceptional examples have even failed to update for iPhone 6/6+ screen sizes, continuing to operate in zoom-mode a year later - hardly an attempt to delight users.

Much has been said and written about the economic difficulties of making a business in app development, with so many free or £0.69 apps around. But I gladly coughed up the ‘premium’ price for Tweetbot and NetNewsWire.

When I say ‘premium’, I mean the price of a couple of Cappuccinos, which perhaps puts things into perspective. I can drink coffee for a few days, or I can buy an app I’ll use every day for the next couple of years, at which point I’ll gladly pay again for the upgraded version.

There’s a role for something better and that role has value. Many of us still appreciate craftsmanship and are willing to pay for it.

I cannot afford to furnish my house with bespoke furniture. But luckily, software craftsmanship is cheap. If you can afford a Starbucks, you can afford a good app. So when you see a great app that costs a few pounds, compare it to the value you get from a coffee. Apps like Tweetbot and NetNewsWire are enormous bargains.

Apple Pay - the first novel payment system that actually works for users?

I used Apple Pay last week for the first time. It’s really excellent and a huge contrast to other mobile payment solutions I’ve used or looked at. I thought it worth a little analysis of what’s going on with Apple Pay because there are some interesting lessons here.

To start things off, it’s worth recapping why other mobile payments solutions have struggled to succeed. I’ve previously been quite dismissive of those solutions because they all have at least one of the following drawbacks:

1. They aren’t secure. e.g. Barclays bPay thing is very convenient, but relies on a £20 limit on transaction value as there's is no way to authenticate a transaction. I can’t accept the idea of so little security on a banking transaction, regardless of the imposed limits. bPay is just a digital version of cash - if you lose the physical bPay device, you quite possibly lose all the money loaded onto it.
2. They are actually less convenient than our normal European chip+pin solutions. Unlocking a phone, finding and launching an app and entering a secondary security code is a surprisingly common series of steps. Or loading money from your bank account into a special new “mobile wallet”, that comes with yet another passcode to remember. These approaches are less convenient than a plastic card. And the £20 transaction limit on UK contactless payments pretty much limits its use to coffee+sandwich only - hardly convenient.
3. There’s no critical mass of acceptance at retailers. If I can’t actually use a new payment method to buy the things I need, it’s pretty useless. There’s a profusion of interesting “fintech” payment solutions - but you have to really try hard to find a retailer that accepts them and there’s no realistic prospect of that changing.

I think that Apple Pay looks like it might be a winner - because it ranks well on all three of my problem fronts:

1. It’s secure - you can’t pay without authenticating with TouchID. And TouchID is almost certainly more secure than a 4-digit PIN code that anyone can steal by looking over your shoulder.
2. It’s convenient - just hold the phone whilst touching TouchID. No codes to enter, no buttons to press, no apps to find. It’s about as simple as it’s possible to be. Because TouchID means the Apple Pay transaction is secure, some UK retailers are already removing the countries £20 contactless limit for Apple Pay.
3. It’s gaining critical mass. Apple Pay works at the growing number of standard contactless points that are being rolled out across retailers. Because UK banks had already invested in NFC contactless features in debit/credit cards, Apple Pay exploits that industry standard point-of-sale infrastructure. And in the USA the launch timing was perfect, as industry rushed to renew point-of-sale infrastructure in order to increase security in the light of scandals like the Target hack.

It’s still early days for Apple Pay (especially in my native UK where it’s only just launched), but it’s the only novel payment solution I’ve seen that has a chance of meeting my three criteria. It has to be secure, it has to be simpler than chip+pin and it has to have a demonstrable likelihood of critical mass acceptance by retailers.

I don’t care about how novel it is or isn’t, if it saves retailers a few fractions of a penny or how it might disrupt the world. The only thing that will make a new payment mechanism successful, is if the experience is better for users. If it’s not, people won’t use it and everything else is irrelevant.

Of course Apple Pay will always be a relatively niche solution because it’s maximum market size is a subset of iPhone users. But American Express seems to have survived for many years on a similar basis.

The way that Apple Pay is integrated deeply into iOS is a critical part of the overall user experience. Others could use the TouchID API to build fingerprint authentication into their payment solutions, but that would still require the user to unlock the phone and find/launch the appropriate app before starting the payment process. In contrast, the that Apple Pay is integrated deeply into the iPhone and iOS requires none of this, just needing the phone to be placed near the payment terminal.

This experience could only be created by an organisation who has the ability to influence not just the banking ecosystem (Apple had to galvanise the banking industry into delivering on the promise of Tokenisation as part of the Apple Pay solution), but also the phone hardware and operating system design.

Lets look at what was needed to bring Apple Pay to fruition (this is far from a complete list, I am sure):

• Bring TouchID to market, ensuring the availability of a seamless authentication solution
• Wait to ensure a critical-mass of TouchID-enabled phones being used by customers, so that when Apple Pay launches there’s enough people with TouchID-enabled phones to use it
• Influence Visa and Mastercard to implement Tokenisation
• Get the retail banks on board to implement their end of the solution - at the time of writing I can count over 380 banks currently supporting Apple Pay, so no mean feat
• Strike deals with payment processors like Stripe to ensure that Apple Pay can also be used for in-app payments
• Integrate Apple Pay deeply into iOS so the end-to-end user experience is simpler than anything else
• In the USA, sign up a broad array of retail partners who will implement the required point-of-sale infrastructure to support Apple Pay
• In the UK, start influencing retailers who’ve already implemented contactless payments to remove the £20 limit for Apple Pay transactions
• And since it’s Apple, do most of this in secret

Let us not forget that Apple had to purchase Authentec for $356m in order to obtain the underlying technology needed to create TouchID in the first place. On top of that were the presumably significant investment costs to perfect and bring to market that nascent technology. If we then add in the costs of creating Apple Pay, it’s easy to see that it might have cost well in excess of $0.5bn to make Apple Pay happen. I don’t think that is the sort of money that banks are typically investing in their innovation programmes.

Apple Pay is the fruition of a very ambitious goal, the creation of new innovations like Tokenisation and TouchID across several industries and a very complex set of coordinated deliveries over several years. I think it’s fair to say that Apple isn’t playing at this.

For those that ask “why could the banks not have created their own successful mobile payment solution?”, the answer is simple. They can’t. Without the ability to influence the way the mobile OS and hardware interact with payments, it’s not possible to engineer the required user experience. This is why we’ve seen a proliferation of small-scale “experiments” that people like me sneered at, rather than anything significant.

Companies like Apple are also able to justify huge investments that banks would really struggle with. They can do that because they see mobile payments not as an end in itself, but as just another jigsaw-piece in the success of their wider mobile platform. And a company like Apple sees this on a global scale, whereas nearly all banks are regional players when it comes to retail payments.

The cost for traditional industries who wish to become serious digital players is higher than I think many acknowledge. Take Facebook’s urgent need to make the transition from desktop browser to mobile app - acquiring Instagram for $1bn and WhatsApp for $19bn. Or BMW, Mercedes and Volkswagon purchasing Nokia’s HERE mapping division for $2.8bn. Or Adidas purchasing runtastic for $240m. Whether purchasing mature and successful businesses (Instagram, WhatsApp, HERE) or raw technology (Authentec), the costs and business risks involved are often significant.

I find Apple’s ability to control not just its components of the solution, but also to influence and lead the required wider consortium, perhaps the most interesting lesson here. Vast arrays of retailers and banks, together with Mastercard and Visa, all needed to be influenced and coordinated to make Apple Pay happen. Banks would typically go about this kind of thing with armies of highly paid consultants and a heavy-weight “project office”. I don’t think that is how Apple did it - I’ve certainly not heard of those consultant armies. Intriguing.

Overall, it seems that creating success in mobile payments needs high ambition, deep pockets and an ability to influence all parts of the ecosystem involved - with that in mind, Apple Pay is a rather remarkable achievement. I’m not sure I see how banks can do this kind of thing on their own - which is an extraordinary statement about the role of the banking and technology industries.

Might "chat" become the universal User Interface?

Google’s home page has a single box: “what are you looking for?”. We type our questions, and receive answers.

In contrast, most other home pages are cluttered with images, adverts, complex menus and general clutter. We have to hunt for what we’re looking for amongst the trivia, the distractions and the irrelevant. Often times our answer is buried several pages into a site - so we have the added distraction of working out how to navigate to what we want.

The web was a brilliant invention for publication, but it’s a frustrating one when we just want an answer or to get something done.

Google originally became popular not because its search was especially better than others, but because its clean and simple home page loaded quickly, in an era when many of us were still using dial-up modems.

Fast, clean and simple, the “search box as universal home page” launched a $430bn company.

If search as a UI works for Google, why should it not be extended for a bank, an insurance company or a retailer? Of course our needs with those organisations are frequently more complex than simple search. So the simple search box needs to be extended to allow a dialogue. We ask questions, we get answers without distractions.

Your retailer might ask “What are you looking for?” and we might type “I’m looking for men’s casual shirts”. And Boom!, there you are, straight to the men’s casual shirt page.

Or your insurance company might ask “How can I help you?” and we might reply with “I’d like a car insurance quote”. The chat would then enter a natural language dialogue where we are asked a series of further questions to gather the required information. All of this can be done through a single chat dialogue - just in the way that we message our friends and relatives in iMessage, Facebook or WhatsApp.

Rather than searching in vein for the “how to change your password” and navigating several pages of a website, we would just type “how do I change my password” and get straight to the information we need.

Of course chat needs to be more than simple text. A chat dialogue can have embedded pictures, videos, maps, links and can cause an app to do things like display a popup. It’s not that there’s only chat, but that chat might be the centre of the user experience.

The dominance of social media apps in the modern age means that a “chat interface” is now our most frequently used interface with a computer. All of us, from teenager to grandparent, have a common and universal understanding of how chat works.

Indeed evidence from China is that chat interfaces for brands of all sizes are now expected and commonplace. The Chinese appear comfortable adopting this style of interaction not just with friends, but also businesses.

Of course there is one drawback to using “chat as the universal UI” – and that’s that it needs a human on the other end. It’s all very well us chatting to our bank or our local builder’s merchant, but that needs an army of trained people to respond…

That is where cognitive technologies like IBM Watson come in. Because Watson understands natural language and is able to hold a conversation, it can automate these chat-based interfaces. Instead of chatting with a human operator at a business, that business trains Watson to handle the chat on its behalf.

We’re already seeing these automated “virtual assistants” in our phones. Siri, Google Now and Cortana prove that no smartphone worth its salt can now come to the market without one. Some may sniff at the idea of chatting to a computer in our phones, but more than half of US teens now use “voice search” daily.

Riding on the popular familiarity of assistants like Siri, we’re beginning to see a new wave of “conversational” apps where the primary interface is a chat one.

• Native is a travel concierge app, with a chat interface
• Vida is life and health coach app, with a chat interface
• x.ai is a personal assistant who schedules meetings for you, through a chat interface
• Luka is an app that finds and books places to eat, through a chat interface

One of the most popular calendar apps on iOS is Fantastical. A big reason for its popularity is because you can use natural language to interact with it. Type “I want a meeting with Fred at 10pm tomorrow” and it creates an appropriate calendar entry for you.

Simplicity sells and chat is the simplest UI imaginable – instantly recognisable to all and more efficient than complex menus and mice. I think we’re seeing the start of a big trend with chat interfaces because of that simplicity.