Thursday, March 15, 2012

Platform choice for startups should not just be about technology

Most startups typically build on the platform their founder(s) know best, I'm guilty of making choices on that basis too, but this can have undesirable consequences as the company grows.  Initially it might make sense, since it probably represents the shortest path to market, but platform bias has no place in building a successful business and there is no requirement that your founders be experts in the platform you finally choose.  Here are other criteria worth considering when making your platform choice.

Talent Availability

It is definitely worth surveying the local employment market to determine the availability of talent to staff your technical organization as it grows.  Most of the great programmers (and great programmers are needed if you want to build great software) are reluctant to work on a technology they deem is "on the way out".  Likewise if you choose an exotic technology you may not be able to find any available resources with the requisite skill set.  Keep in mind however most great programmers also love to learn new technologies much more than they prefer to work on dated technologies, so don't exclude talent just because they don't have prior experience.  If you do hire staff without the exact skill set you need, try to make sure you hire at least one expert to act as the keystone and mentor to the team.

You also need to recognize that not everyone is attracted to the risks associated with working for a startup, and this definitely seems to be linked to geography.  Don't try to force square pegs into round holes by underselling the risks, or expectations of a role in a startup.

Talent Cost

As a corollary to talent availability is talent cost.  Generally the more exotic your requirements are the higher the salary demand, but talent demand in excess of supply can produce the same effect even around relatively common skill sets.  For example here in Cincinnati the majority of the major financial services and banking institutions have made major investments in Java technologies and are therefore major competitors for talent with this skill set.

Vendor Roadmap

This can be pretty hard to assess.  But do the best you can to ascertain what the platform vendor has in mind for the technologies you plan to use.  Get with the local representatives of the vendor, go to technical conferences, talk to the Product and Program Mangers, Microsoft has ISV Evangelists whose job is to help software companies with technology and architecture choices.  Sometimes even with the best information at hand you might not get the true picture however.  At the time I chose Silverlight as the basis for ProspectStream it looked like Microsoft was fully committed to the platform and that there was a reasonable probability that it might be ported to Apple iOS.  Of course we now know long term Silverlight will be supplanted by Metro/WinRT.  However if you have done your research you will find these transitions will take several years to infiltrate the market and give you time to react.  If you choose to do no research expect to be surprised, and blindsided.

Operational Costs

Finally evaluate the operational costs of your platform choice.  Hardware, Software and Management costs, or if you choose to use a Platform as a Service provider their usage costs.  The more exotic your platform choice the more likely these costs are to be higher and the less likely there are to be well integrated management and monitoring tools.  Also try to avoid a conglomerate of support tools that themselves require disparate platforms.

No comments:

Post a Comment