‘Fast failure’ and Work-Out: Organizational cultures that support learning from failure

Just picking up on Patricia’s post on Learning from failure

This is a very important idea that has received very little attention in evaluation, but that the management literature has been discussing since the 1990s. The real challenge is building an organizational culture where not only is experimentation encouraged, but it is not necessarily a career-limiting move to produce  failures, mistakes, and negative results.

  • One way of building support for experimentation is the deliberate encouragement of ‘fast failure’ (Peters, 1994).
  • As Jack Welch of General Electric argued, those firms that experiment with something new and reach failure before their competitors have the chance to be the first to learn about a new product, market, or technology.
  • This is consistent with Vollman’s (1996) assertion that “If change proceeds with no failures, the speed of change is probably too slow” (p. 243).
  • Tushman and O’Reilly (1997) found that a critical factor in the support of risk-taking at firms they studied (including DuPont, Herschey, Nordstrom, and Fedex) was public recognition of good attempts at innovation that failed.

All this is a far cry from the usual way that negative results and failures are responded to by organizations. The survival strategy is to either avoid evaluation like the plague, bury negative results, and/or – especially if you’re in government – cloak them in ‘bureaucratese’ so that the public and the media won’t be able decipher them.

My own experience is that learning from failures seems to happen more effectively in evaluations where staff are involved in gathering and making sense of the evidence – people tend to believe and take on board things that they discover for themselves (with the help of a good evaluation facilitator and coach, we hope!). But if the organizational culture makes this a dangerous and career-limiting activity, it’s incredibly hard for evaluation to help organizations learn from failure.

My own read on this is that it only works when led by example from the very top of the organization. Jack Welch at General Electric used to run a process called Work Out where wasteful, bureaucratic processes and other areas for improvement were identified, people from all sides of the process (design, marketing, production, sales, etc) were brought together and challenged to find creative, silo-spanning solutions. Yes/no decisions were made in this public forum before the Work Out meeting was closed, and people were empowered to implement the chosen solutions.

The other important message here is to steer clear of the ‘blame game’ as much as possible because in many cases, negative results weren’t any particular person’s fault. The Work Out process at General Electric began by focusing on long-entrenched bureaucratic practices that had become embedded in “the way we do things around here”. As strategy and management specialist Vijay Sathe points out, in many cases the things that need changing weren’t necessarily mistakes or bad decisions; they were good ideas at the time, but the context has changed. He gets organizations to ask themselves, “What did we do right to get into this mess?”

All this is much more difficult for government agencies, who are under constant public scrutiny, with the media circling like vultures for any scrap of ‘dirt’ they can use to make headlines. As one central government employee commented to me the other day, if the media discover that a program has been changed (e.g. after an evaluation identifies areas for improvement), they immediately seize on it and announce that “Agency X keeps chopping and changing and obviously doesn’t know what it’s doing.” For this reason, the learning process is often deliberately hidden from the public eye.

Once again we are reminded that it’s not just about what organizations and leaders do to support evaluation and learning; inappropriate reporting by the media is an important barrier to learning from failure – and to genuine evaluation.

4 comments to ‘Fast failure’ and Work-Out: Organizational cultures that support learning from failure

  • Jane,

    For me, one challenge around learning from failure has been in the naming of it. “FAILURE” is such a negatively powerful word. Calling something a failure labels those who worked on it failures by association. The social stigma of failure is so strong that it can perpetuate commitment to unsuccessful programs so as not to damage the successful and valuable people behind them.

    The reality is that it is rare that a program is completely a failure. We can find at a minimum lessons learned and almost always some segment of the population that found benefit. Thus, we remove the label of ‘failure’ and re-frame the program as successful for some.

    I’m trying to increase, for myself, the frequency with which I define not only success measures but also failure measures. One can then posit that there is success, there is failure, and there is a gray area within which the most restructuring of a program might take place. When we define only measures of success, then we create a dichotomy of success or failure and the difference between the two can be a very small margin wherein everything that is not success is either failure or the gray zone. Too often, I would argue, it is the latter – the ambiguous zone of ‘not quite success’ wherein we hesitate to label something as failure and admit the need to learn from the experience, cease the program, and examine alternative ways for pursuing the programmatic goals (more capable of course than we were had the failure not occurred).

    Please note that the above reflections are my own and do not necessarily reflect those of my employer, the American Evaluation Association.

  • Patricia Rogers

    Thanks, Susan, for reminding us of the importance of language in assisting or hindering us to learn from the half-full glasses that are most programs, projects and policies. The “F” word (failure) might be quite an impediment to recognizing and learning from mixed success.

    It reminds me of a process that some colleagues systematically used to debrief after every training day. We would as a group ask “What did we do? What went well? What went not so well? What do we need to do next?”.

    “Not so well” is a broad and very vague term, but it did create a space for people to identify gaps, deficiencies, mistakes, and areas for improvement.

  • ‘Languaging’ is a term coined by NZ evaluator Nan Wehipeihana, and relates to the need to find ways for difficult or complex ideas to make sense in different contexts. Failure is certainly a word that seems to raise people’s defenses, and as Jane has pointed out, even more so in public sector contexts. So we are constantly looking for ways of broaching the ‘not so well’ territory that Patricia describes. One recent example where we used positive ‘languaging’ so that people could engage with areas of ‘not so great’ performance was when we developed a rubric in the mental health sector with the following catergories: not doing well; at risk; stable/ supported; self sufficient; and thriving. This terminology was much easier for people to hear, and subsequently they were much more able to engage in conversations about less than great performance.

  • Jane Davidson

    Very nice example, Kate – thanks for sharing that one! And thanks, Susan, for raising this important point.

    I have a few more thoughts on this one – more than I expected, actually – so will start a new post!