The positive feedback loop generator

It currently takes Apple ten days to review and then (hopefully) approve an app for release. Ten days. This isn’t ten days of them running the app through various rigorous testing procedures. For nine and three-quarter days your app is waiting for review. Then it moves into review and usually less than six hours later the app has been approved.

Which means this is probably an issue of resourcing. It seems Apple has not enough of the technology or staff it needs to review apps. The queue to review is nine and three-quarter days. The time someone spends reviewing it is six hours. Is it too simplistic to imagine that more people/technology doing the reviews the shorter the queue would be?

The shorter the time from submit to release the more innovative we can be. If we can try things out and fix them quickly we’ll be more willing to experiment. As it is, we need to make sure the app is just right before submitting, which in my experience means a drawn-out cycle of minor changes. God help us if we get it wrong and the one star reviews come flooding in (to which we can’t reply) with a ten day wait before we can put things right again.

If this is an issue of resources then Apple is being cheap. The app store is a highly profitable business that deserves investment. We know review times can be shorter because they’ve been shorter before. Better apps are going to come if they’re shorter again.

I’ve built a stop smoking app that’s being downloaded a fair bit (over 10k times a month at present). Given most of this traffic comes from the UK and only 10% from the US it’s reasonable to think there’s more people we could reach.

I won’t deny that vanity is one of the factors driving this desire, but it’s not the main one. Beneath the app is an experiment that seeks to test the efficacy of different behaviour change techniques on this, one of the most challenging of habits to break. More users means more data and more data allows us to determine not just which behaviour change techniques work but for whom in particular and when.

So why isn’t the app being downloaded in the US? The reviews in the UK store are great so its seems people like the thing. It also appears high in the search results for terms such as ‘stop smoking’, ‘smoking cessation’, ‘quit smoking’, ‘give up smoking’ and ‘smoke free’. Or at least it does when I search.

I don’t know much about what makes people choose a particular app but I do know that search is key. Apple restricts descriptive keywords to 100 characters and also weighs terms in the title more heavily. Therefore, it’s important to choose the right keywords carefully.

A tool I’ve been using to help is SensorTower (others are available). I’m on the free service and get the ability to check where my app appears for five sets of keywords in one country’s iTunes Store. I also get a sense of the traffic those keywords receive, their difficulty of appearing high in the results and how many other apps are returned when they’re entered.

There are other features (I believe many of them free), though I’ve only really used the search ranking one. It did work though. I saw that US smokers use the term ‘quit smoking’ more frequently and so changed the title of the app. Now when it’s entered mine appears in position 6, up from 54. This position can change of course but you get notifications whenever that happens (an email I simultaneously look forward to and dread).

My US users aren’t growing through the roof, but they are growing and I think the longer the app stays in those kinds of positions the more the traffic should increase.

I should point out that one of the reasons I’m writing this blog is because SensorTower are offering access to their paid-for services in return for a blog about them. However, they were clear they wanted honest reviews and I’ve not felt obligated to praise them. In that spirit I’d like to say I think their starter prices too high for small developers and would urge them to do more to accommodate the kinds of people who could most benefit from their service. I’ve just started a PhD in technology and behaviour change and have found a large number of people working hard to create evidence-based health apps. There’s a huge need for apps that deliver good information but people searching for them have to wade through a load of apps that aren’t up to scratch. Problem is that most researchers operate under tight budgets and will probably find the fee just shy of $500 a year to be prohibitive.

Apart from that I really like their service and have genuinely found it useful. I also liked the email I got from one of the founders ten minutes after signing up. It could have been an auto-response but it seemed personal and it was probably the single reason I spent some time on this software rather than sticking it in the pile of things to do later.

I’ve been reading Jane McGonigal’s Reality is Broken with a somewhat skeptical mind. I don’t play games much anymore and the idea that people would immerse themselves so completely into a game is pretty alien to me.

But then I read with awe a post on Reddit describing in great detail what someone else called “one of the largest PvP battles in the game’s history“. LG03′s description of the universe he/she and their fellow players inhabit is so vivid it’s clear that for them there is a very definite realness to it. The situation might be fantastical but the feelings it engenders are real. They embody the characters they take on, revel in the role their character has got and respond in the way they believe they would.

One of the main criticisms of evolutionary psychology is that it’s all post-hoc rationalisation. We can’t test hypothesis about how our behaviour has evolved because we can only see how things have evolved by looking backwards. To be clear, Evolutionary Psychologists claim that this is false, that hypothesis can be tested and that they’ve created or have found studies that support their claims. However, as any student knows from a quick search on PsycNet, you can find studies to support more or less any claim you want and (to use an redundant phrase) more research is needed before EP will silence the majority of its critics.

I don’t know if there is a Massive Multiplayer Online Roleplaying game of Eve’s type for Evolutionary Psychology but perhaps if one were created we could test some of the hypothesis for how human behaviour has developed and what, if any, parts of it are universal. We could put players in the role of our Savannah-dwelling ancestors, give them the tools and skills they would have had, set up the situations they would have faced and see how they respond.

This wouldn’t answer any questions definitively of course, but given the sheer immersiveness of these worlds and the way players throw themselves into their character (and the sheer number and range of people playing) we could get a better sense of what behaviours are triggered in different circumstances. Is conflict a winning strategy or is it trumped by cooperation? What sexual selection behaviours do the male and female characters engage in? What compensatory strategies do people come up with to deal with the problems of group living?

It would obviously take a huge amount of development work to create a game that would answer these kinds of questions in a scientifically reliable way. But I’m sure it’s not beyond the ken of high-end game developers to create something playable and with lots of players and well thought-out hypothesis the answers would be fascinating.

32 different countries will compete for the 2014 World Cup and London is probably one of the very few places outside Brazil where you’ll be able to find a community from each one of them.

So instead of going to Brazil to watch the tournament why not a book about staying in London to watch the supporters? What are their pre-match rituals? Where do they go to watch the game? What kind of people watch? What kind of language do they use when watching, how do they respond to injustice, how do they celebrate?

Comparing and contrasting the way different people around the world watch football might be an interesting project for someone.

Wikipedia is an incredible resource. It has proved beyond doubt that crowdsourcing works, that intrinsic motivation delivers amazing results, that people want to contribute. I use it, well if not daily then five times a week at least.

It’s not always well-written though. Too many cooks perhaps, though there’s no easy solution to that. One or two editors would be ideal but how are they going to get appointed? How do you stop people correcting facts but not changing style?

So how about we combine Reddit and Wikipedia and add a bit of Quora. For every Wikipedia article people find and vote up links that best detail the content of that page. Instead of the Wikipedia page on Buddhism we go to a page that describes Buddhism in the way that people who know a lot about Buddhism think is good.

And this is the crucial difference. Because there would be a way for people accredited with knowledge on the subject to have extra voting power to you and me. We get one upvote per story. Mr professor-got-this-subject-by-the-balls, ten points per differential psychology story.

The user arrives the page on Buddhism or differential psychology or anything else and sees alongside a set of links to other stories, each one of which has got, and this is the Quora bit, details of who’s voted it up and why they’re knowledgeable on this subject.

Unlimited money man me (the one who’d buy Arsenal) would also buy the Olympic Stadium and turn it into the best damn venue in the world. Rent it out to huge artists for them to put on the biggest show imaginable.  Things you could do with those lights and that system in that space.

If I was running a site that allowed reviews of its products I’d add a field showing the time difference between purchase and review. Buy something in May, review it in August and the post would say “Owned for four months.” The review system would need to link accounts with the purchase ledger, but that must be possible.

Knowing how long someone has owned something before reviewing it tells us how much experience they have with the product. It also quite a lot reveals more about their feelings if the reviewer feels compelled to come back some time later. And could be more persuasive as a result.

A lot of the things I call ‘ideas’ seem very obvious to me. The Spotify one below for example. Surely there’s a matching engine for Spotify somewhere. Amazon’s got one, so has iTunes, Netflix and I’m sure countless others. Spotify is the ideal place for a matching engine given the rich data it can mine about collective musical preferences. Nothing unique there surely.

Another one of these is the thought that any company with technical support should have a team dedicated to proactively answering support issues on other people’s forums. These days, first thing I do when I encounter a problem is Google it. If I can find the solution on a forum I feel great. If I can’t, I call support, in what’s often a bad mood.

Not only am I now making the support line’s staff a misery, I’m also needlessly consuming company resources. Wouldn’t it be better all round if the company allocated four or five (or fifty if you’re Google) to looking on forums other than their own, identifying themselves as experts, and walking people through the solution?

The key is other people’s forums because many companies will have staff looking after their own. But not everyone goes to the own-brand forum, many have their own favourite or will go to the first they find. And some forums, yes yours Google, are terrible at allowing you to find the right answer.

A proactive approach to answering known issues in online spaces other than one’s own, would – I’m sure – save a significant amount in phone support costs. It’s so obvious I’m sure every big tech company is doing it. Right?

As an employer I pay HMRC the tax I’ve collect from staff and the National Insurance contributions my organisation makes on their behalf or about the 17th of each month. The month I’m paying for is the calendar month before. So on the 17th February I’m paying my HMRC dues from January, meaning I’ve had that money in my bank account gaining interest for 17 days.

Which, as anyone in credit with the bank will tell you, amounts to not very much at all. My £7k is too small to do anything very profitable with, something the measly interest rate the bank gives me proves. But what if there was a way to combine my £7k with the hundreds of thousands of other employers who also wait until the last minute before paying?

My idea would be for the HMRC to say “Pay us on the first of each month instead of the seventeenth and we’ll invest your money with everyone else who does the same. This rather large sum of money will gain far more interest than you could get on your own and to be fair, we’ll split the profit with you. Half the profit of what we earn from investing your money we’ll give you back. The other half we’ll put into a Sovereign Fund and we’ll use it to fund national projects.”

There’s no major change involved for the organisation. Most payroll is calculated electronically and my guess is it’s about as easy to pay in the 1st as it is any other day. If a company was having cash flow problems they could elect to pay later, all online.

Other organisations outside of HMRC could do the same to encourage early payment. It is, after all, what all businesses want. What consumers want is a fair deal. And the web makes mass aggregation easy.

There are a number of sites that give you access to Spotify playlists but what there doesn’t seem to be is a site that says – if you like that you’ll like this.

What I’d love is to be able to mark a track as a favourite (for some reason I see this different to starring it) and have a matching engine analyse the track and find others like it.  I could then mark the tracks it finds as on the money or off the mark and by so doing, further develop its ability to bring me the music I want to hear.

Tools that help us filter signal from noise are one of the great next developments of the information age. The problem now is too much choice and this is particularly acute when it comes to music. There is so much wonderful music out there I know I would love but I just don’t know how to find it. I know what I like and with Spotify I’ve got a way of listening to it, be great if someone could develop technology to point me to it.