When you have a real problem yourself, there is a very good chance that others face the very same problem. And if the problem is big or serious enough, you can create a business around it. Having a real problem to solve is a way better approach to inventing an imaginary problem that your startup to solve. The reason many startups die is simple: they create products that no one wants.
Some might tell you that formal market research might give you insights into whether your future product might have a demand. I’ve lost a lot of time and money chasing down ideas based on market research that was done on a cohort that directly was my potential audience. 30 of them gave took the time to answer, very positively, some questions we framed about the problem we were trying to solve. None of them eventually bought the product they said was their “dream solution” in the questionnaire we had provided during the survey.
Ever been in a situation where you thought you wish you had the money to buy something–and when you had the money to buy it–you didn’t really bother? Someone who only assumes she’ll buy your product will then give you not-so-useful feedback since they have nothing to lose and no boss to answer to.
There are other ways to figure out if your idea might hold potential.
We’ve come to a stage where software that forms the building blocks of what we do as a developer community is in general open source. Of course, there are exceptions. Examples of this are:
Creating software around areas mentioned above require that you either have a quantum leap in mind or you plan to open source it yourself.
Several engineers who I’ve spoken in regards to their business plans tend to have very technical ideas for what I call “infrastructure software” falling under one of the above mentioned categories. These requirements are generally served by open source software and are difficult to monetize.
The scope of what I call “infrastructure software” generally grows beyond the categories mentioned above to include the current, popular technologies and paradigms. Recently, Google open sourced TensorFlow. As I write this, machine learning stuff is fast becoming “infrastructure software” that will be used more and more commonly by future applications.
Case in point: As IBM’s patents on virtualization expired, Intel and AMD started including virtualization technologies right into their chips. Around this time, in 2007, Xen was a popular hypervisor solution. But, I knew that something that would take advantage of the CPU’s new virtualization capabilities would work better. I had a startup then named Binary Karma and I remember discussing ideas of developing something along these lines with colleagues and co-founders. We decided to drop it, since we knew this would end up being “infrastructure software”. Qumranet, which developed KVM along these lines got acquired by Redhat and KVM was merged into the Linux kernel in 2007. KVM today is “infrastructure software” along with systems like Open Stack that manage virtual machines.
With the scope of “infrastructure software” extending into virtualization, chasing down this product idea with very limited resources wouldn’t have got us anywhere unless we had a dramatically better idea and were super-confident it’s scope and execution. It is important to predict, given current technology trends, which category of software is likely to become “infrastructure software” for the next generation of innovation. When you are trying to build a startup and you’re evolving a product idea, you always want to be in the cutting edge. And the cutting edge is a place full of “infrastructure software” pitfalls. So, you need to choose carefully.
Don’t get me wrong here. There is nothing wrong developing “infrastructure software”. It is just that monetizing it is tough. You can open source it, build a community around it and play the long-term game if you know what you are doing. It just seems like a tough thing to be successful with if you are a first-time entrepreneur with no open source experience.
When you have competition from open source software, you are fighting not on one, but two fronts: cost, which is free and free press and marketing that open source software gets. It is not easy to fight both zero cost and free marketing all that easily.
Open Source and free tools are the software equivalent of content marketing. Well written open source software and free tools can get you a lot of traffic. Open Sources 2.0 is a great resource for someone trying to understand how open source software can make money. It is a very different and counter-intuitive model, but there are some very successful companies who have all of their “IP” out there on Github to download.
Even if you don’t plan on making your core product open source, making portions of it will most certainly make you known. This is the reason why many companies create free tools and open source libraries for people to use. A good example is Qualys’s SSL Test which tests for strengths and weaknesses in a web server’s SSL configuration. While using the tool, it should make many a sysadmin curious about what Qualys does as a company. And since the company put out the tool, it is instantly in a position of authority. Compare that to putting out ads, that need to be vetted for trust.
If not for the tool making the company’s name known and website visited, it would have taken real advertisement dollars on Google AdWords or a similar system. Over a period of time, free tools, which really are a form of organic marketing, tend to give you a lot more value compared to inorganic advertisements.
Getting consumers to pay is very, very tough. Traditionally consumers are used to not paying for online services. Prosumers and professionals will pay. Businesses will pay relatively easily if you solve a problem for them and make the process of doing their business easier. Consumers aren’t really doing anything while they are not working. They are usually entertaining themselves. They might pay for entertainment–but they tend to pirate, see? They all have a big problem paying up.