Few people launch software products expecting them to fail. But many products do fail. I don’t have any figures, but I think I can fairly confidently state that more commercial software products fail than succeed. You think your product isn’t going to be one of the failures. But so does everyone else. The only way to find out for sure is to launch. The sooner you launch, the sooner you will find out. I have banged the drum for releasing early before, so I won’t labour it here. But it begs the question – how do I launch fast? What do I leave out? Based on my experiences of launching 3 software products, this is what I would leave out.
Polish
As developers we (hopefully) all want to do great work that we can feel proud of. But, as entrepreneurs, we need to be careful not to spend lots of time polishing something that might be a turd. So ship v1.0 before it is polished. Early adopters tend to be fairly forgiving of a few rough edges, if they are interested in the direction you are taking. I spent 6 months (part-time) working on the first version of my AdWords keyword tool. It flopped. So I shipped the first version of my visual planning software within a few weeks of writing the first line of code. It was pretty bare-bones and a bit slow for plans with hundreds of cards, but it was enough to demonstrate the basic concept.
Designer website
You don’t need a beautiful, state-of-the-art website to launch your product. My own table planner software had a pretty ropey website (designed by me) for the first 10 years and it did fine. Just make sure the website clearly conveys what your product does.
Logo
You don’t need a professional logo for v1.0. The product name in coloured text using a font other than Arial will probably be fine. I did the initial logo for Hyper Plan in Microsoft Word Art in 10 minutes. Here it is in all it’s glory:
I only paid a designer to come up with something better once I was sure it was worth my while.
DRM/Payment processing
I shipped the first version of Hyper Plan without even setting up licensing or payment processing. Every time you ran it, it just put up a window saying that it would expire on a certain date and that a new release would be available by that date. After that date it just stopped working.
I only added licensing and payment processing once I had proved enough people were interested in the concept to make it worth my while. If you are going to take this approach, make sure you let people know that they will be expected to pay at some point.
Sophisticated pricing model
Ideally you want to segment your customers so you can charge more for the people who are prepared to pay more. But you probably don’t understand your market well enough to do this when you are starting out. So just pick a single price. I introduced segmented pricing for PerfectTablePlan in v4. Hyper Plan still has a single price.
Feature parity with your competitors
Trying to achieve feature parity with established competitors in v1.0 is a fool’s errand. Just pick one pain point that you think is not being well addressed and try to solve that. Make your lack of features a selling point by emphasizing how simple your product is to use.
Multi Platform
If it is going to take significant additional effort to release multi-platform, then just pick one platform to launch v1.0 on.
Extensive documentation
The first version of your product should be simple enough and well enough designed that it doesn’t need extensive documentation. My Hyper Plan software has been out for a year and it still only has a one page quick start guide.
Mailing list
Many people advocate building up a mailing list of interested people before you launch. It obviously helps a lot if you already have an audience in the market you are launching into. But, if you don’t, it takes significant time and effort to build that audience. I would rather put in that effort once I have something to show them.
Trademark
Why bother to spend time and money trademarking something if you don’t even know if anyone wants it?
Patent
I’m not a fan of software patents and I don’t have any patents after nearly 11 years in business. So I certainly wouldn’t waste time and money on a patent for v1.0.
Lawyers
If a bug in your software could kill someone or destroy their business, you should probably talk to a lawyer. Otherwise a boiler-plate end user licence agreement is probably fine for v1.0.
Company
I did create a limited company before I launched my first product to get a bit of extra legal protection. But its not strictly necessary (in the UK at least).
Trade-offs
It’s all a tradeoff. Obviously it is better to have a beautiful website than an ugly one. But is it worth spending lots of time and money on designing a beautiful website for an unproven product?
The best approach depends very much on your market and circumstances. If you are a big player with lots of money and reputation, then much of the above may not apply. If you are selling web design products, you had better have a pretty slick looking website for v1.0. If you are selling aircraft avionics systems then I hope v1.0 of your product is pretty polished.
Successful Software 
Interesting article. One thing though, wouldn’t it make more sense to have a Buy Now button of some sort that referred to your web site? This way you could see in the logs how many people were interested in buying.
I have also seen some people suggest A/B test pretend prices to this kind of page and having a message that explains that product is not available yet when the user clicks (plus the ever-present mailing list to be notified when it is available for purchase). Of course, it might be better to only go through all this trouble when you’ve already seen that there are enough people getting to this page.
(personally I add a web page open to my installs with some instructions and a video on how to use the program, but it is a different thing)
A fake ‘buy’ button is another approach. But I prefer to let people have it for free to start with and see how much engagement there is (emails, pings to my server when it automatically checks for updates).
Andy, do your users know your software is going to check for updates? None of us here particularly likes when a program suddenly wants to connect to the Internet.
It puts up a window telling them it is checking for updates, which also tells them how to turn off the check (in the ‘Preferences’ window). It only checks once per week by default. I don’t remember ever getting a complaint about it.
Everything makes perfect sense! And here’s a life hack for you and everyone reading your blog.
If you want to have a proof before a competitor steals and patents your cool feature, go to a local notary public. Print out a screenshot or write up a document and ask them to date it. It’s cheap and, in theory, you’ll be able to provide some kind of proof in a court of law. If you’re taken to court, that is.
May i ask in what languages / ide you developed perfecttableplan?? Thanks.
C++ and Qt.
Great article Andy, thank you for the tips.
Quick question:
Whether you prefer to publish immediately a commercial version or maybe at least the beta version for a month or two?
Davor,
As with most things related to marketing, it depends…
With Hyper Plan I made the first few releases completely free, but it was clear that the release would die on a certain date and you would have to pay to continue use after that. I found that was a good way to validate whether anyone actually wanted to use the product, without bothering about licensing and payment processing.