Sam Iles's Blog

‘If you are not embarrassed by the first version of your product, you’ve launched too late.’ – Reid Hoffman, Founder of LinkedIn

I read this quotation recently and it really resonated with me. I broadly agree with him. Here’s why.

Feedback is critical to any product. Especially on the web. Websites are so easy to tweak that making a mistake doesn’t matter. But, if you spend a year writing code and taking your product in one direction which isn’t actually what people want then you’ve wasted your time. Feedback from users telling you what’s good, what isn’t and what they’d like to see is so important. Even if you have a handful of users, feedback is still invaluable. Even if “feedback” isn’t words, but just usage metrics it’s still important.

I always try to launch early. I figure that the sooner I have code in people’s hands the sooner I can build a user base and the sooner I can get tangible feedback. Sure, some users will be dissuaded from using your service if they can’t handle the pre-perfectly-designed version, or need some features or can’t deal with a bug or two. But, in the long run, developing a product with constant feedback will always make a better product.

This incidentally, what I intend to do this week. I started making a website on Monday and I am going to get it out this week. It will have the basic features and the basic design and we can go from there. If I can get half a dozen people to use it and offer the occasional comment then I can direct development far better and hopefully make happier users as they see their ideas and thoughts being realised.

So there we have it. A short post to add my thoughts to Reid’s: launch early, iterate often, make a better product.

I don’t know about you, but I took a ton of those job matching tests. I hear they’re called “Jobbos” in the USA, which is a funny name. Every time I did one I got a list of jobs I would obstensibly be suited to. Every time – hand to God – software programmer was on the list.

Mind you, it was often joined by such exciting career paths as “librarian” and “lawyer”.

But, software programmer was the one which stood out. I should say, I was programming before I took these tests. I started making websites when I was around eight years old. I had my first website on the line at around ten years old. I started writing the real software when I was around thirteen. Etc.

So, I was set by around age thirteen on software programmer. And, that’s what I do now and plan to do forever. In that, that’s where any money I have comes from.

And, I love it.

I remember a conversation in school re software programming. The general response was “wow, that sounds boring”. And, I guess it does. Writing line after line of words which don’t seem to do much for one’s fun metre…

Nevertheless, it does for mine. I literally love writing software.

Let me explain how it works. First, you have an idea for something to make. I have around fifty ideas written down in various places. This is the most exciting part. This all consuming tidal wave of thoughts which just won’t stop. Sometimes, all I can think about for days is that latest software idea. It’s a real rush.

Next, comes the general design and detailed planning. Drawing pictures, making notes, working everything out. Damn this is fun. Then, comes the actual planning of code. I love this part too. The experts call it ’software architecture’. It’s the bit where you design high the building will stand up and exactly how everything in it will work. Software architecture is about designing the code and exactly what you need to write. What database software will you use? How will things interact and happen?

Then, you have to write it. I love this bit too (surprisingly). The solving of problems is fantastic. Writing seemingly meaningless words on a screen and then watching as it does something. The incredible frustration when something doesn’t work and the inexplicable joy when you finally crack it. I have literally screamed and cheered in joy when something finally works after hours of fixing it.

I also love the sheer possibilities of software. There really is no limit to what you can put on the screen. Hardware is physical and limited. But, software isn’t. You could write a system with perfect, human level speech recognition. You could write a system which talks to humans and understands what you want perfectly. There are no limits: only the current version number of your software. At the risk of sounding like a unicorn infatuated six year old: if you can dream it, you can build it.

And that’s why I love writing software. I genuinely really enjoy it and it’s pretty much all I want to do forever.

So, I just got an email from Microsoft about making apps for Windows and – in particular – Windows Phone 7 (which they’ve been paying people to develop for). They badly need apps because they think their couple of thousand looks bad up to Apple’s hundreds of thousands.

Anyway, this email contained the headline “build an app for that”.

I just thought it was worth reflecting on the difference between their’s and Apple’s app stores.

Apple’s tagline: there’s an app for that. Microsoft’s tagline: build an app for that.

I just thought it was funny and 140 characters wasn’t enough to express it.

The Apple App Store for iPhone and iPad has been a resounding success for Apple. They are forever talking about huge figures regarding total developers and apps and how much money it makes. There is also no doubt that the App Store is a major driving factor in hardware sells for Apple: as both Bill Gates and Steve Jobs have said: it’s software that sells hardware.

Apps are cheap, too. Most tend to be the lowest band that Apple offers of just 99 cents. Pretty cheap. Some apps go up the to the next band if they get good word of mouth going and fewer still hit the third. It’s rare that you see an app getting into the tens of dollars. But, on my Mac, I routinely buy software which costs around $50. In fact, I (and my company) spend hundreds of dollars a year on Mac software. But, it would be hard to spend that much on iPhone apps.

But, are iOS apps really worth so much less? I don’t think so. There are numerous apps on the store which cost just a dollar or two which I think are worth far more than that. Now, I am about to simplify a lot, but, generally speaking, a Mac app and iOS app with roughly the same features takes roughly the same effort to code. Of course, there are major differences. A Mac UI needs to be thought out more and there are more APIs available for iOS than for Mac and an application like Scrivener is obviously going to take a lot of work. But, fundamentally, the work is the same.

There are of course apps available for iOS which are worth nowhere near their asking price. Some apps asks for a dollar or two when they really should be free. But, conversely, there are apps which should be much more. And also, I have yet to see any company trying to sell a Mac app for 99 cents. Why do Mac apps cost so much more?

I think the cause is mainly competition based. There are so many versions of nearly everything on the App store than prices are driven down a lot and then somebody comes along and makes a free SSH client and the other seven which cost a dollar are history. There is also the issue of consumer volume. Most people buy one or two (or no) Mac apps in a year while people will happily spend $1 in the iOS App store many times. It’s odd that there is such a discrepancy between apps for PCs and iOS.

It is also true that some of the numbers we see reported for iOS app revenue are truly abysmal. (Before I get into these figures, let me say they are easily worked out from figures which Apple themselves have provided and while they of course should be taken with a pinch of ±, they are the best and most accurate we have). The average income for paid apps is $700 per month which certainly sounds good but this is highly top heavy. In other words, it is the Angry Birds (or as I now call it, Dangerous Crows) and those who Need Speed that make up that revenue (consider the average of 1 million, 2 million, 1, 2 and 3 for an illustration of this point). Here is the big, scary, evil number: half of all developers earn less than $682 per year from their app. I think that, even for a hobbyist, this makes it all rather worthless.

When you consider that over half of all paid apps get less than 1,000 downloads you realise why app prices are so low. Of course, there are many apps on the store that are bad but I think only delusional people would say that over half of apps are bad are ergo get few downloads. It isn’t that many. If we really think about it, how do most people download apps? I think it’s mostly from seeing them in the Apple featured sections or the top 25 (instant high value, very small selection of apps requiring mostly luck or an incredible product) and from search (low volume, people looking for specific apps). Unless you have a really, really great idea I don’t think there is much money to be made from paid downloads. I think of how I get apps: I rarely pay for them unless I see it in the top 10 and I know I will use it indefinitely.

Another stat: apps with iAd make 10-30 cents per user per month on an iPhone App. So, 10,000 free downloads could make $24,000 in a year. I think 10,000 is a big number but I don’t think it’s unachievable with a free app. Providing you get a mention or two on The Twitter or some random blog and perhaps get into the top 50 of one of the lesser categories, you’re set. This approach assumes people keep using your app and don’t uninstall it. I think this is good as it encourages developers to make better quality apps. (On a brief side note, I wouldn’t put a paid app in the store ever – only completely free of ad supported, which is what I plan to do soon…)

So, regardless or whether you can make money on the app store or not I think we can see the reason that apps (including very good ones) are generally so cheap: huge volume, intense competition, reluctance to purchase apps, prevalence of terrible software and the rest.

And then Apple makes an announcement: App Store for Mac. Now, I haven’t written about all the pros and cons of the Mac App Store and I’m not going to do that here – I just want to focus on price.

If the above problems transfer to the Mac store then I think it could be bad for software developers. Imagine if Scrivener suddenly finds itself surrounded by terrible – but one dollar – word processors or imagine if MarsEdit gets lost in a virtual see of blog editing software that’s ad supported. Now, one could argue that this is just capitalism, this is competition, it’s good for consumers but, seriously now, how can Lit & Lat (just one man) make a living from selling Scrivener for $2.99? Perhaps I am presenting an apocalyptic vision of Mac software here, but I do really think it is a big cause for concern.

Previously, the idea of putting a Deverous Mac app in the store so people could do everything within a beautiful, stable, offline version of the software. I figured that could conceivably sell for $20 or so and people would buy that. Now, however, the more I think of it, the less appealing it seems. I don’t want to find myself making an app that nobody buys. Of course, I am over simplifying here as it would be users of the service that saw it on Deverous.com and shelled out not App store customers, but you see my point. I think now that if such an app ever did exist in the future, it would be better funded through a Pro subscription (with other features) to Deverous.

Will the Mac app store devalue the software industry as it has to iOS software? I hope not, but I think so.

Although I keep promising that Project0 will be launching soon, it’s not.

I thought I would do this quick post to explain why.

Although it has been ready for some time now, I keep thinking of cool new features and adding them, putting of tasks which must be done for launch.

This, if you did not know, is called ‘feature creep’.

I promise I won’t do that on any future projects. From now on I will always plan carefully, make it and then release it as beta. I will then add any features I think off during development later. Else I will never release anything.

I promise.