Lots of people have talked about how the economics of start-ups have changed. As costs have come down thanks to tools like MySQL, Amazon AWS, and free frameworks like Rails, where it used to require millions of dollars to build an early stage product, now a product can be built for virtually nothing.
The bottleneck now appears to be access to engineering talent.
There are no hardware constraints. The only thing that stands between an entrepreneur’s idea and the realization of that idea is convincing technical talent to devote their time to it.
One could write a whole host of blog posts around this, but I want to talk about a few interesting side effects I have seen.
There are more tiny consulting companies than ever. I am talking about free agents (1 person) and small shops (2-10 people). I have noticed that many of these are managed poorly. I would say a common thread in many of these is that the people that start them either never worked at a large consulting shop or they were a person that hated all the process and machinations at their prior employer, did not value them, and quit to go their own way. These companies are easy to start – you can simply hang around on oDesk or something worst case and build relationships from there – because demand for technical talent is so high.
But an interesting side effect of this is that as the technical talent gets distributed to these tiny companies, the process and project overhead diminishes significantly. That sounds great to engineers, but do not doubt that there is some value in process – particularly as regards to the creation of visibility in the process (which usually is completely unvalued by the people doing the work). The result is that the burden is placed on the client to “manage” the project and make sure that they are getting bang for their buck. I have also seen a sharp transition away from the trend toward fixed pricing of projects and de-risking the project for the customer to a time and materials model and “scrum-ey” project methodologies. This is great for the developer and CAN BE great for the customer, but only if the customer has the infrastructure needed to manage the project well.
So as technology frameworks have simplified the development of technology, the trend in consultancies has been toward smaller and smaller organizations, but increasingly these consultants are providing only technology development, not the infrastructure that many have come to expect that results in complete delivery of the final project. Customers own project management, QA and other typically core engineering functions – the consultant is just manpower.