About a million years ago, I got a Commodore 64 and in between playing some of my (still) favourite games (Uridium, Archon) of all time I learnt BASIC programming and created some small games. Fast forward many years and in University to learn ‘c’, I decided to reinvent an obscure board game called ‘September’ and logically called it ‘October’. There was some more dabbling with games but uni finished and I got a ‘real job’.
The ‘creative’ in me never quite died. I modded an xbox, pre-ordered Wii, downloaded homebrew sdk’s and even paid for the Torque Game Engine all with a view to creating and publishing a game. The iPhone came out and I went and bought my first Mac, signed up as a developer and thought very hard about making a game. 25 years and about 2000 brilliant ideas have gone past since I wrote that first BASIC game on the C64….
Interestingly, I have done a lot of things in those 25 years. I recently had to cull my resume as i was concerned for the trees whenever I applied for a job but it wasn’t until very recently that I was brought closer to the games industry. One christmas break I decided to put my head down and build a modern, 3D version of one of my favourite C64 game, Uridium. I fired up Google and decided it was time to revisit OpenGL and dove head first into some tutorials. About 15 minutes in, I decided that it might be a better use of my time to take a step back from OpenGL. Hello Unity3D and WOW!
It is no coincidence that Unity3D is the tool of choice for almost every games developer except for AAA publishers. That statistic is even stronger when talking about mobile platform games. After about 10 hours, I had a 3D spacecraft flying, shooting, flipping and exploding in space with a beautiful stella background. That 10 hours is total time from knowing almost nothing to working prototype. In those 25 years of development, I am confident in saying that Unity3D is the very best of all the software tools I have ever used. I truly hope they do not become a victim of their own success and keep it at that exceptional level of quality.
It was a bit of a case of serendipity that DTT and I found each other at that point. They needed someone with experience and versatility to do a proof of concept for what has become their Ditto Lite product. What I needed was the discipline of turning a bit of hacking into a commercial grade application, or in the case of the Ditto, a suite of applications and games. If you look at about 75% of the apps available in Google Play and Apple iTunes, this is the piece that is missing. Apple and Google have dramatically lowered the barriers to publishing, but in doing so, opened the doors to games and applications that lack the finishing touches that invested product owners demand.
2 Weeks == 6 Months
My daughter, Ella really loves books and reading them to her is a real pleasure. She also loves the ipod touch I bought to test the apps I created. It was about the middle of Ella’s Kindergarten year and the idea struck me to put together a little spelling game with simple 3 letter words. The title of the game would be ‘Three Letter Words’. I had a pretty good relationship with the teachers at Kids Academy and frequently showed them the prototypes. They thought it was cool and were very excited, but being Kindergarten teachers, they do get pretty excited about just about everything.
After about 2 weeks, I was nearly done and gave it to Ella to play. The first version was in landscape mode, had tiles that looked a bit like scrabble pieces and sat in a 2.5D bucket on the right hand side of the picture. Ella struggled to find the letters piled underneath one another. It was pretty ugly too. Back to the drawing board. There were a couple of key things wrong with the game.
1. The alphabet needed to be accessible and navigable. The game was difficult enough having all the letters to choose from without adding searching.
2. It needed bright colours
3. I needed a menu to lay out all the words.
4. It needed an incentive/reward system
I moved the tiles into a grid but firstly, 26 doesn’t divide very well by 4 or 5 and secondly, 26 little squares take up a lot of space on the screen, so I put 2 rows of 13 along the bottom of the screen, extending past the end with scrolling left-right. I also decided that the incentive should be that letters ‘explode’, so you need to be quick and that each letter will be worth a point. Problem with this layout was that when a letter exploded, it left a hole and the more holes there were, the sillier it looked.
I was now at about 6 weeks. Finally, it occurred to me that 25 divides by 5 really nicely and that I can put 5 rows of 5 tiles. I just needed to get rid of a letter. I couldn’t think of any 3 letter words that used ‘Q’, so bye bye Q. If Ella Speller takes over the world, there will be whole generation of children that think the english alphabet only has 25 letters. As it turned out, if I put the rows around a cylinder, you could fit the whole alphabet in a fraction of the screen, it looks cool and when the tile ‘explodes’ it doesn’t look too ugly.
10 weeks. I have a menu, using a paper motif. Decided that ‘stars’ are the reward of choice for children. Talked with Lloyd (artist that drew the pictures) and the idea of a rainbow was formed. This worked really well as I needed some way to put boundaries on the size of the game and 7 colours == 7 levels and I had about 70 good words to be drawn.
12 weeks. It is starting to look a bit like the ella speller you see today, however, each level had 9 words and the menu showed thumbnails of the 9 drawings and when you clicked on the page, it would randomly iterate through the words. Ella had favourites and kept trying to click on the ‘ANT’ but would get annoyed when a different picture appeared. It was quite clear at that point that the kids should be able to ‘choose their own adventure’ and the menu a set of pages that open and 9 selectable tiles was born.
It is about 16 weeks since inception now. One thing you should understand is that when you have a crystal clear idea of what you are producing, the original 2 week estimate would not have been so far off. But with trial and error, long gestation periods between ideas and the Mass Effect trilogy to complete, 2 weeks became 6 months.
I was ready to go, I gave Lloyd the final list of words and told him to get cracking, I will be ready to publish in about 1 week. I grabbed the Prime31 plugin for In-App-Billing on Android and put the code in to lock down levels 2-7, that should be pretty straightforward…and it was. The early versions of Ella Speller required the kids to get a minimum score on each level to progress to the next so adding the freemium locks to subsequent levels meant that a level could be locked because you didn’t pay for it or it could be locked because you needed to collect more stars. I experimented with different ways to represent this, but in the end, I just had to abandon the minimum score requirement.
Go Go Go!
20 weeks. I have 3 levels worth of art (27 drawings) from Lloyd and they are beautiful. Originally, I hoped the drawings could speak for themselves, alas, there was still enough ambiguity to require the word to be said aloud. I sat Ella down at the computer, fired up Audacity and flicked through the pictures getting her to say each word. I hurriedly attached a button to the drawings and played the recorded audio whenever someone pokes it. It’s a bit of a last minute hack and it is not obvious. I am sure many of the users still haven’t figured out that the pictures speak Australian if you touch it.
PUBLISH!!! Google play is first to go, no reviews, In-App-Purchasing tested and functional. The excitement, the anticipation. Waiting. Hitting refresh on the Play Store app until ‘Ella Speller’ search returned a result. It’s LIVE. *crowd cheers*. I wonder how many downloads it will get in the first day, 10, 100, 1000? Twenty four hours later, after hitting refresh on the stats page about a thousand times, it turns out that it WAS refreshed.
My phone, Lloyd’s phone and my tablet. 1 week later, 8 installs. How could that be? I had a catchy name, screenshots, carefully crafted app description, even a video on youtube. There must be something wrong, I thought a brand new app would attract at least a little bit early attention from the ‘newest apps’ browsers. I decided to pretend to be a user and searched ‘spelli….’ google autocompletes it to ‘spelling games for kids’. Ok….not on page 1, that’s to be expected. Page 26, somewhere after ‘Harry potter’s magic spells’ and various other results that do not include the keywords spelling, games or kids.
In the meantime, I was wrestling with Apple’s developer program. I had signed up to the program as Sage Mobile Pty Ltd, a company I had registered a few years ago, but didn’t operate. I needed to give Apple proof of my ABN and GST registration before they would BEGIN reviewing Ella Speller. I quickly applied for an ABN thinking I would get it at the end of the application process, but I made a trivial error in the form and the application entered the ATO’s black hole where their service standard required them to take action within 28 days and the call centre could not take any action unless the 28 days was exceeded. It eventually arrived (after about 10 days, I think) and I desperately completed the remaining forms.
Rather than try and figure out the In-App-Purchasing for Apple and risk getting something wrong and having the app rejected, I decided to submit V1.0 as a single $0.99 purchase with all the levels unlocked. It took about 7 days but it got approved and went on sale in iTunes on the 10th of December, exactly 2 weeks after launching on Google Play. Being a little more prepared for disappointment, I was not so surprised that Ella Speller’s first week netted 27 purchases which roughly accounts for close friends and family with iOS devices and young kids.
According to the timestamp on the ThreeLetterWords folder, it was created on the 22nd of March, 2012. Ella Speller v1.0 went live on the 27th of November, 2012. It was originally going to be a little side-project for fun and to get my feet wet in the world of app publishing before taking on a more ambitious project. As it turns out, the app stores are so flooded with content, no matter how good something is, you need to draw attention to it. How do you do that? There are plenty of guides, marketing resources, review sites, etc. I took to Facebook and Twitter, dabbled in advertising. I have learnt some new respect for people in marketing.
As I write this (almost 2 months after launching), Ella Speller has 35 twitter followers, 36 likes on facebook and accumulated about 500 downloads across iOS and Android. On the AppStore, I have 3, 5-star reviews, by people I know, zero on android. Approximately 1 in 20 people on iOS upgrade to the full version, and 1 in 15 on the Amazon app store. 1 in 100 from Google Play.
$20 of targeted facebook advertising bought me about 12 ‘likes’ and I saw the best day of downloads immediately after that, but it is still about 1 or 2 orders of magnitude shy of where it should be to pay for itself. I released v1.1 with a couple of enhancements, level 4 added and a direct link to the rating and review page. Didn’t make much difference. I abandoned (or postponed) the idea of creating 7 levels. It is just not economically justifiable at the moment.
The game I originally started to make is still sitting in the background, about 5% complete. I started a new project, code named ‘Devious‘, because it is ‘Deviously simple’. I have learnt some lessons. It is about 4 weeks since inception and it is about 75% complete but I am looking to build an audience before it is released. The current plan is to release it exclusively on facebook, using the viral attributes of fb to build an identity. I am still not sure of the numbers, but it seems critical to get an app into one of the ‘top’ charts on mobile to achieve success. To do that, when it is launched on mobile or tablet, it needs perhaps thousands of downloads per day and without a big marketing budget, that is going to be tough.