Proofs of Concept (PoC) are great. They allow one to test a small component and then apply success (or failure) to future endeavours. Certainly the all time champion of the PoC are the Myth busters. Adam and Jamie would start each myth with a small-scale test before going big (and with the obligatory BIG explosion).
To Hack or to Formalize a PoC
PoCs come in many sizes. At one end is the developer who experiments and comes up with a workaround or a more elegant way to achieve a result (aka a good ‘hack’). On the other end is an organization that incrementally works toward a final objective. For example sending a series of Apollo missions into space with each one adding on to the knowledge and experience of its successor. This blog considers more than a midnight pizza fueled hack-a-thon but much less than sending humans into the unknown.
The Scientific Process (sort of) to the Rescue
One of human’s greatest achievements was the development of the Scientific Method which involves (courtesy of dictionary.com):
The following Six PoC Questions for Success is loosely based on the above method. The intent is to help an organization understand why a PoC is a good idea and the result. At the same time, this is ‘just-enough’ formalization. After all, it is important to let the brilliant folks develop ‘elegant-hacks‘ without too much paper work.
1. What was the Business problem being addressed?
Why was a PoC identified? Generally this is to address a specific business problem. Pure research is okay as an objective for a PoC. That is developing technologies or techniques with no immediate application but future potential value for an organization.
2. How is the problem currently being solved?
The answer to this question is that it is often not solved, done through intuition or completed via a manual/semi-automated process. This question helps the organization understand what to do the with the results of a PoC. If the manual process is only slightly more costly then a fully automated variety, why bother with the complexities of automation?
3. The Question
In effect this is the hypothesis portion of the scientific method. Ideally this question should be a simple Yes/No. If the nature of the question changes through the PoC process, that is okay – but the evolution of the question should be included as part of the final report. Thus we may have started asking question X but we ended up answering question Y. The reason is that X was too big/small/wrong and Y was answerable.
Defining the question is important so your PoC team does lose its way and they have a touchstone to come back to. A bit of formalization around how they can change, extend, shrink or otherwise amend the question is important.
4. What were the results at the end of the project?
This question should have two parts, a) and b). Part a) is the predicted result. By including a prediction, the PoC can stay focused on the intended result. This is not to discount secondary benefits or chance discoveries but it does help to ensure that a PoC does not become its own self-sustaining cottage industry. Consider keeping part a) secret from the PoC team if you want the benefits of the double blind effect.
Part b) is what happened, what were the results? This should support the response to the question answered above. Ideally the result is Yes or No but it might be Maybe. Of course everyone wants a momentous discovery every time. However failure should be seen as a positive result – such a result may have saved an organizations considerable time, talent and treasure.
5. What are the next steps?
This should be a very practical listing of how to use these results. Examples of next steps may include refining a subsequent PoC, engaging in a larger scale test or moving the resulting solution to production.
6. What is the Future Vision, What is Possible?
Question five focuses on the practical and immediate application of the PoC results. Question six let’s the team blue sky a bit and extrapolate findings to larger contexts. This is part of the fun and value of the PoC – the larger application of something new.
No Explosion – Using the Six Questions
Sorry, unlike the Mythbusters, there is no end of blog explosion. Instead, these questions are a handy reminder of the things to consider when a PoC is being suggested. Let me know your thoughts on the six questions. Would you add a question or take away one or more of them?