Fail fast? Just don′t do it.
Fail fast? Just don't do it (your MVP must be viable).
?======
The expression "FAIL FAST" results from the mistaken process of simplifying a concept that arose in the context of innovation by experimentation. Nobody wants to fail. Neither we want to fail fast. After all, if you fail fast, you're fu**ing fired!
In its original context, the innovation experimentation process must be prepared to quickly identify the initial signs that our experiment is following a different path from what was initially planned. This context is associated with four factors: planning, objectives, monitoring and adjustment.
1. Planning
Good planning is essential for the experiment to work; so that it is feasible. Your product (or prototype or process) has to be viable, even if it contains only a restricted (minimal) part of attributes (features). The concept of MVP (Minimum Viable Product) is based upon viability as a premise; and as a strategy, to be minimal. There's no point in waiting for something viable to fail fast.
Remember - at the end of the day, we need a VIABLE PRODUCT, with a minimum number of features; not a MINIMUM PRODUCT, eventually viable.
2. Objectives
The experiment should be supported by objectives and metrics. Otherwise, without the North of your compass, it will not be possible to evaluate the experiment's effectiveness. Even in divergent experiments structured for the production of data and insights, which often generate new questions to be addressed, it is essential to set goals in advance. Notwithstanding, one must understand that the characteristics of the goals of divergent experiments are different in nature from the goals applied in convergent experiments (associated with much more specific objectives).
领英推荐
3. Monitoring
Particularly useful to divergent experiments, the presence of controls capable of monitoring execution and alerting undesirable behaviour deviations is essential. Such controls should identify the first signs that something could get out of control – or notice a recent failure. For example, if there is a failure in the experiment's architecture, the people involved must be alerted of the consequences.
Note that we don't want to "Fail Fast". We don't want to fail at all! When failure is inevitable, the later, the better – so we can validate much of what is tested. Additionally, we want to recognize the need for change as soon as possible. Either at the beginning or end of the experiment. So we want to LEARN FAST.
4. Adjustment
That's what we want: to LEARN FAST. As we learn, it is possible to proceed with the original plan / experiment design, or change it (to pivot ...).
To learn fast, we must plan to monitor each step and capture the signals generated throughout the execution of the experiment. It′s all about planning. All signs are potentially helpful: those aligned with our goals, those that challenge our goals, and the unexpected signs which make us reflect on unconsidered opportunities.
Therefore, we should not confuse "Learn Fast" with "Fail Fast".
After all, if you fail fast, you ?re fu**ing fired!
?
The FAIL FAST concept was taught to me by a previous employer. They encouraged us to try new things (as part of continuous improvement), but preferred that we spent a few hours to prove/disprove something, rather than spend weeks/months on something that we had no evidence wouldnt work. You weren't going to be FIRED for failing, afterall failing is part of personal and professional growth, because you learn how to do things that do work. On that basis, I quite like the FAIL FAST concept.... I teach it to my current colleagues. To me, its aligned to the TEST & LEARN, having an interactive/continual testing approach, a mentality about questioning things and looking for the evidence that something will actually fix a problem, rather than just delivering a tick-in-the-box solution that resolved nothing. I encourage my colleagues to build simple proof-of-concept or prototype solutions, based on mvp, and test them, with stakeholders.... capture the evidence, see if the solution would work before spending much longer developing a full working version.
Technical Operations Sr Director - México at The Coca-Cola Company
2 年????????
CSCO | Diretora Sr Supply Chain | Opera??es | Compras | CSCP
2 年Agree! Learn fast and share learnings within the organization, so that everybody is on the same page and similar mistakes are avoided in the future.
Spare Parts Logistics Manager (Lean) @ Siemens Healthineers | Master's in Logistics
2 年#PDCA ??