Save to My Latticework unsave

Experiment and Fail Fast to Succeed
Experiment and Fail Fast to Succeed
save0 saved view3.2K views
(2)
Share this with your network
Share this with your network
Overview

"Only those who attempt the absurd will achieve the impossible."

— M. C. Escher.

In our complex world, where we are constantly being required to make sense and take action through ambiguity, you have a key method of your disposal that can change everything. 

Experimentation. 

Creating mini-experiments or 'small bets' as some have come to describe them, allows you to learn more through action. In particular, this approach allows you to fail fast — which means testing and dismissing many ideas efficiently with minimal time, money or risk, so you can find the unexpected gem. 

So what are you waiting for? Start experimenting with this powerful collection of mental models to empower you and your organisation to fail fast and innovate ;) 

This Playbook includes the following sections, select a heading to jump to that section.
PREPARE
You'll need some basics in place to get started experimenting.
Cynefin Framework
Before you dive into experimenting — ask, is it really necessary? Use the Cynefin framework to explore whether you're in a 'clear' domain that is relatively straightforward.
The Scientific Method
This is the foundational model behind so many, defining an approach to better understand and act in our world. Develop your hypothesis, then use experiments to test and learn.
Curiosity Zone
If you want to experiment, what better place to start than to tap into your curiosity. Use this model to ensure that you have the optimal knowledge to create an information-gap to help drive your curious investigation.
Psychological Safety
At an organisational or team-based level, you can talk all you want about experimenting and 'safe to fail' approaches, but unless you have Psychological Safety, it will all remain as hype. Start here, then experiment.
CYCLES OF EXPERIMENTATION
Apply these models to test, experiment and learn.
Return On Failure
Use this simple model to help reframe the way you think about failure including reducing your investment into the experiment and ensuring you get maximum lessons from it.
Risk Matrix
lock

This one is to help you talk about the F word without creating panic in management or executive circles. And, by the F word, of course we mean 'Failure.' You're not only failing fast and cheaply, you're failing safely. So consider the risks, consider the implications, and consider how your experiment can test an idea in the safest way possible.
Agile Methodology
Agile methodology is really about bringing the Scientific Method to project teams. Or at least, that's the vision...
Split Testing & A/B Testing
How much impact does your experiment really have? How will you know unless you run a Split Test and compare it to a control group? If the experiment shifts the dial, this will also arm you with a compelling business case moving forward.
Minimum Viable Product
Move beyond minor tests to delivering the smallest thing possible, that has value, to your audience. Click into this model to see variations on MVPs and how we've gone with the 'cup cake model' to preview an experience.
Lean Startup
lock

This influential model in the startup world has leveraged MVPs and broader Lean Methodology by advocating for quick cycles of releases and pivoting to find the place where your business best delivers value.
Riskiest Assumption Test
An alternative of sorts to MVP — focusing on the potential fail points upfront rather than experimenting with value.
Prototypes
Prototyping can be fast and deliver tangible results. Click into this model to explore the variety of options you have at hand — from wireframes to storyboards and more.
CHECK YOURSELF
Use these models to be skeptical about your results and open about alternatives.
Confirmation Heuristic
And, in all of your experiments, beware of the Confirmation Heuristic. It's a truly powerful model that biases everyone, especially when you really want an idea to work. Just assume that you've got it (because you do), and organise your experiment and process to counter it. Click into the model for some tips.
Correlation vs Causation
When understanding and interpreting the data from your experiments, be sure that you are not mistaking Correlation for Causation. You might require a few follow-up experiments or additional data to confirm a link.
Categories:

Premium content

Please do login or sign up to see premium contect

Subscription expired!

Please renew your subscription to access this feature.

note My Notes

    Nothing here yet. Join ModelThinkers and login to be the first to comment.