A long, long time ago, in a magazine far, far away, Steve once wrote an article called ‘Writing Shareware – A Guide For Software Authors,’ where rather than code tips, he talked about marketing tips and how to make sure you get the best from your product when you take it to market. And I thought it was time to dust down those points, and see if they can be dragged into the twenty first century.
The market has changed now - in those days almost everyone was a 'bedroom coder' and there were very few companies per-se actually working in the mobile scene. This has been flipped over now, with the bedroom coder still there, but as a much smaller part of the equation when compared to the publishing companies. Yet the same decisions and methods remain.
First of all, when marketing your application you need to decide if you are going to offer a trial version, either as a seperate cut-down original, or (more likely) as a full version with some features restricted through the code until the application 'unlocked'. Of course, there’s always the suspicion on the part of the shopper that if there's no trial version then the application is going to be rubbish.
No matter how you restrict your trial version, always start with your most restrictive version first. It’s easy enough to loosen what you offer, but if you start out offering a version that has no restrictions at all, you’ll never be able to recall it. But don’t go too far the other way. By all means program in restrictions, but make sure the application is still useable. You want to let people realise the application works and fits what they want it to do, and then hit them with something that makes them consider registering.
Games are a prime example. Offer too little and nobody is going to care about wanting to carry on – offer too much and they’ll have their satisfaction paid off, and see no need to bump up to the full version. For games, there’s a much smaller window when you’ll be able to sell your application. Which means...
Always test your application thoroughly, and have others (play) test it as well. They should be able to help out finding the point where you want to ‘restrict’ the demo, as well as exercise all the functionality of the full version.
Steve’s 1996 article says this about the 'Nag' screen… "DON'T over-nag. Nag screens shouldn't be intrusive, i.e. once at the start and end is enough. If the user uses your program regularly he's going to see the screen regularly anyway." With Symbian devices nowadays, and the 'everything for free' market as it is, I think you need a bit more nagging than previously – there’s a lot less momentum to register nowadays than in the days of ten thousand enthusiasts all supporting each other, as was the case in the days of Psion.
Do ensure that registration can be done by typing in a code etc, and that the user is not going to have to re-install the program. Make the process as quick and easy as possible.
There are a lot of ways you can accept registrations nowadays, from stores such as the Nokia Software Shop, Sony Ericsson world, the monolithic Handango and Motricity, and your own web site. Use them all – the more places your application can be found, the better. And don’t forget that your own web site will have traffic, so offer links to all the stores, and allow purchasing the app through PayPal directly, as options.
Make sure users can contact you – a contact URL in the application is good – even better would be an email, but at the very least make sure there’s a contact page that’s easy to find on the web site.
Be aware just how much you are asking for. About £8 is a minimum for a small program and £20.00 about the maximum for an all-singing all-dancing mega-application. Of course, the world is very international, and the price is going to have to contend with fluctuating exchange rates. Be aware where most of your sales will come from (and for Symbian that currently means UK and Europe) so price accordingly for those regions. You should also consider foreign language translations – assuming you have a resource file of all your application strings, one foreign beta tester should be able to do the required translation for you and test it all works.
Do make sure your app is generally useful before expecting massive sales. Unless you're an absolute vertical marketing wizard, a scuba-diving pressure calculator, for example, would be best off as a freeware program, because it's only really ever going to get a handful of people interested in it. Don’t be afraid to enter a genre that has a lot of competition, just make sure you are offering something that is either better value or has more functionality.
Do make sure any installation notes are as clear as possible, both on your web site and in embedded instructions and comments in your SIS file. Remember that many users will be novices and not very computer-literate. Write your installation instructions to help these kind of people and you won't go far wrong. Be aware that most people don't read documentation thoroughly (now there's a surprise!), so try to use built-in help screens to provide the actual program manual, for moments when they're out and about and need help.
Be prepared to take time to respond to user wish-lists, fix any bugs and generally maintain and support the program. One of the main advantages of one-person/hobbyist software over its commercial equivalent is the way it can be frequently updated and revised in the light of user requests. If your program is successful, you will be facing 30 to 40 emails a day, and each should be responded to quickly and efficiently. Answering e-mail from your customers (and potential customers) is part of your life now and being off-line for more than 48 hours is a no-no. If this is a problem, stick to freeware!
Of course, if your application needs to be Symbian Signed for functionality (or anti-piracy) reasons, you won’t be able to release as many updates because of the cost involved, which is why it is important to test rigorously. And then test again, and again, on as many different devices as possible, in real world configurations.
Do be honest! With the best will in the world, your program will have bugs in it. You can make a lot of friends (and registrations) by admitting instantly if there is a bug and notifying all existing users straight away, via email or your web site. You will always find a few people who expect programs to be perfect, but the vast majority understand that bugs will be present and are perfectly content if you acknowledge it and either provide a workaround or email them a 'fixed' program file back in a short time. Make sure you have a user-friendly error handler in the program. Users seem much happier if the program pops up a short message telling them what and where the problem is and to email the author about it.
Do advertise yourself. No-one is going to register your program unless they know about it. So get your web site linked up with others and take every opportunity to tell people about its existence (adding it to your forum signatures is a good start, and won't do your Google ranking any harm). There are opportunities for paid adverts of course, such as on AllAboutSymbian (email Rafe!), but there are also a multitude of opportunities for free 'plugs'. Use the blogs that specialise in your chosen platform, for instance.
Do remember that the first 20 seconds of the program's running are absolutely crucial. It must be smart, professional looking, totally intuitive and immediately impress the user, making him or her want to explore further. Any minor problem or something that isn't obvious and you've blown it!
Of course these are only hints, there are lots of things to consider, and as we've shown here, some choices are mutually exclusive. As long as you remember that the aim is (strangely) not to maximise how much the end user enjoys your product, but give them enough enjoyment and confidence so they will register or buy your product, you should have given your application the best possible start.
Ewan Spence, November 2006