Guerrilla Prototyping

How BMJ gets prototypes delivered.

In common with many organisations BMJ is keen to test new concepts, but when it comes down delivering the goods we have very little capacity available as we cannot justify using a whole sprint team when there is always masses of revenue generating work to deliver. Additionally prototyping is viewed by some as a large investment for low returns.

As Head of Digital Strategy I need to be able to try stuff so I and a few BMJ staff who also thought it was really important to get prototypes out in production and  test our assumptions and the possible value of ideas, decided to create prototypes using guerrilla tactics. This process is based on the old yet true axiom, it is easier to ask for forgiveness than permission.

Guerrilla definition: Referring to actions or activities performed in an impromptu way, often without authorization.

Our Guerrilla Prototyping process – 5 easy steps.

The Guerrilla prototyping process we developed is in fact very aligned to our agile development process, as it is focused on delivering small features to test their value in the real (production) environment.

Step 1. Identify one thing you really need to test.

Our Hypothesis: An app that delivers Open Access BMJ research content and supports users selecting topics of interest will be successful in exposing the BMJ brand to non BMJ users.

Find one or two people really who want to know this and no more.

Meet Rae and Kristy who really want to know this.

Raegel De Guzman, BMJ Systems Coordinator
Raegel De Guzman, BMJ Systems Coordinator
Kristy Ebanks-Kegnia, Assistant Editor Veterinary Record
Kristy Ebanks-Kegnia, Assistant Editor Veterinary Record







Don’t tell anyone else you are doing it as they will unwittingly add complexity by wanting to test more things or stop you doing it.

Step 2. Identify the one KPI that is the most important.

Do not try and measure any other KPI at this stage, it will add complexity.

Our KPI – Usage,  more than 100 people will download our app, rate it more than three stars and the comments on the app store will be favourable

Still tell no one…

Step 3. Find resources to help you.

Meet the “resources”.

Chris Wroe: Health Informatician, Michael O'Connor: UI Designer, Esther O'Sullivan: Head of Digital Strategy and Kimon Kotronis: Health Informatician.
Chris Wroe: Health Informatician, Michael O’Connor: UI Designer, Esther O’Sullivan: Head of Digital Strategy and Kimon Kotronis: Health Informatician.

In the team in London we have (left to right) Chris Wroe: Health Informatician,  Michael O’Connor: UI Designer, Esther O’Sullivan: Head of Digital Strategy and  Kimon Kotronis: Health Informatician.

Dan Amos: Digital Strategy Lead and Mobile Champion

In Cardiff, we have Dan Amos, Digital Strategy Lead and Mobile Champion.

Tell them, and only them about the idea.

Step 4. In the simplest way possible build the prototype.

There are two sorts of prototypes you can deliver using this method. The first is a throwaway prototype which is useful if you are coding from scratch as this means you can put speed of delivery over scalability, maintainability etc. If you want more information on throwaway prototyping there is a good overview on the Programmers Stack Exchange site here.

The second is use a ready made application framework and as much off the shelf plug and play software as you can and put BETA all over it, this is what we did using  the meteor framework and a feed from our content catalogue.

Step 5. Put it out there.

G PrototypeOur prototype is here, log in using Twitter or Facebook and have a play. It is at this point that you should probably tell the relevant stakeholders what you are testing.

We will update this post with the prototype results.


(Visited 93 times, 1 visits today)