Software startup R&D process
Recently, I attended an excellent ‘Ask Me Anything’ event for founders looking for investment at Stone & Chalk in Melbourne. The focus was on (pre) seed funding and the panellists were: Simon Beckett, Tom Smalley and Andrew Cicutto. Many thanks to Graham McCorkill and Eike Zeller for running this event.
While Xilytix is not looking for funding (yet), it is well worth attending these events to gain an understanding of what investors are looking for in startup companies. After all, they have experience and skill in identifying company attributes that are indicative of growth and success. One of the attributes they value is the ability to ‘move fast’. Set ambitious goals and reach them quickly. With that in mind, how should a software startup carry out R&D?
The above diagram overviews how we carry this out in Xilytix. There is no specification. Instead it is driven by ideas and vision that lead directly to design and code. (That is why it is R&D.) The process is very incremental and iterative with frequent meetings to review ideas. Very agile! Our team members have the ability to quickly estimate whether we can implement ideas with our limited resources.
AusIndustry subsidises R&D to assist Australia’s economy. However to be eligible, companies need use ‘scientific methodology’. This methodology is slower, less flexible (more waterfall-ish) and significantly more costly. In addition, the outcome has to be ’new knowledge’ - not creative destruction (eg productivity gains) which is what the economy really needs. ‘New Knowledge’ is great for universities and science based companies but not so useful for engineering/software based companies undertaking R&D.
Software startups are an important part of driving productivity in our economy. Subsidies should be structured to encourage their use of R&D with methodologies which are agile, low cost and make sense to the team. The article below explores this further: