July 2018

Never under-estimate the value in quickly getting out a rough draft or prototype

David Griffin, Senior Consultant


Numerous personality types exist within organisations, fulfilling many different roles.  Often when a new product is developed, innovation is required but many key stakeholders are not, by their nature, innovators.

They may be analytical, detail-orientated, or results-focused, but not creative. How they engage with the process has a great impact on the outcome.

It’s said that a designer can imagine something that doesn’t exist yet, but conversely non-designers find it difficult to discuss or even conceptualise abstract future systems.

If an analytical person is asked “what other features should we include?” they are being asked to innovate.  If, however, they are shown a list of possible features and asked which ones they need, they’re not.

It’s much easier to edit or critique a document than to write it from scratch.  Often the best and fastest way to get input is to write a first draft and then get it out there for criticism.

Some requirements will immediately be shot down at first review.  That’s fine, because:

  1. The person who shoots them down has to replace them with something better.
  2. That person is now engaged with the process.


Create something to be knocked down

This first document is often known as an “Aunt Sally” (something for everyone to knock down, as seen in old fashioned fairgrounds).  The sooner this is done, the better, as it shows where designers and end users have differences of understanding.

The number of people who can write a good requirements specification is small, but the number who can spot improvements to a draft is much greater.  If you want to solicit input from as many stakeholders as possible, give them something to critique early.

Generally, engineers focus on de-risking some key performance or functionality challenge early.  This often results in a technology demonstration that looks nothing like a prototype, and to the lay stakeholder may appear unimpressive.   Yet buy-in at this stage is important, so there is real value in taking the extra step to make something look a bit more like the real thing.

This may involve adding a non-functional GUI, or making a 3D-printed case.  This may not strictly represent technical de-risking, but showing stakeholders something tangible can often bring hidden requirements or constraints out of the woodwork, and allows non-creatives to give input throughout the process.

This is, for example, where someone volunteers the information (that he or she assumed everyone knew) that the device has to be operated one-handed or must stand up on a table, or asks where the USB port (that wasn’t specified) will be located.


Fail fast, fail often

There’s a popular expression “fail fast, fail often”, meaning “test your ideas in the marketplace as soon as possible, learn lessons as quickly as practicable, and be prepared to throw away what doesn’t work”.

Though currently fashionable, it’s not how every organisation operates.  Some prefer extensive market research, focus groups and internal product testing before taking the idea public, and in some circumstances this is appropriate, especially when commercially sensitive.

But there are certainly times when the information you learn from getting something (even a non-functional model) in front of a customer quickly saves far greater expense down the line.

This may require making a working version of the minimum viable product that incorporates decisions you may not feel sufficiently informed to make yet.  In this case, it’s necessary to choose something (your best informed guess) and accept that being wrong is not the end of the world.

R & D is about learning, and every day you sit behind closed doors refining a product that hasn’t seen the light of day yet, you’re not learning as fast as you could be.

It’s been said that in the old world, the large ate the small.  In today’s world, the fast eat the slow.

There are a couple of key requirements for being able to “fail fast, fail often”, apart from the obvious change in thinking.  You must be able to generate a passable working prototype fast, and you must be able to do it sufficiently cost-effectively to be able to live with throwing away your failures.

So that first prototype may not be made the same way as the final product, perhaps using standard components, off-the-shelf modules and rapid prototyping.


Getting the organisation right is important too

Knowing vendors who can turn models into parts overnight using additive manufacturing is only one part of what’s required.  It’s also necessary to support this with appropriate internal systems and infrastructure.  Build an organisation that facilitates failing fast without employees needing to resort to “skunkworks” behaviour.

Finally, there’s another good reason for Aunt Sally – it gets the process started.

Most writers know that the first draft may be thrown away but it’s better to write something than to stare at an empty page.  Better to build something than keep refining the specification.

There comes a time to stop bouncing the ball and serve…