WEBVTT 00:00:00.000 --> 00:00:11.040 align:middle line:90% 00:00:11.040 --> 00:00:13.190 align:middle line:84% SARAH GREEN: Welcome to the HBR IdeaCast from Harvard 00:00:13.190 --> 00:00:14.390 align:middle line:90% Business Review. 00:00:14.390 --> 00:00:15.690 align:middle line:90% I'm Sarah Green. 00:00:15.690 --> 00:00:19.270 align:middle line:84% I'm here today with HBS Professor Amy Edmondson, who's 00:00:19.270 --> 00:00:22.730 align:middle line:84% the author of the April, 2011 article, Strategies for 00:00:22.730 --> 00:00:24.000 align:middle line:90% Learning from Failure. 00:00:24.000 --> 00:00:26.150 align:middle line:84% Amy, thanks so much for joining us today. 00:00:26.150 --> 00:00:27.780 align:middle line:84% AMY EDMONDSON: Delighted to be here. 00:00:27.780 --> 00:00:30.140 align:middle line:84% SARAH GREEN: In the article, you talk about a number of the 00:00:30.140 --> 00:00:32.880 align:middle line:84% obstacles that really prevent us from learning from failure. 00:00:32.880 --> 00:00:34.580 align:middle line:84% And a big one is the blame game. 00:00:34.580 --> 00:00:36.290 align:middle line:84% And unfortunately, I think that's a familiar 00:00:36.290 --> 00:00:37.330 align:middle line:90% one to a lot of us. 00:00:37.330 --> 00:00:40.520 align:middle line:84% Why is that still such a challenge? 00:00:40.520 --> 00:00:42.910 align:middle line:84% AMY EDMONDSON: The blame game is still such a challenge 00:00:42.910 --> 00:00:46.330 align:middle line:84% because I think it's something very deeply ingrained in our 00:00:46.330 --> 00:00:47.270 align:middle line:90% psychology. 00:00:47.270 --> 00:00:51.680 align:middle line:84% And as small children, we learn pretty early on that if 00:00:51.680 --> 00:00:55.760 align:middle line:84% we do something wrong, get blamed, it's an emotionally 00:00:55.760 --> 00:00:57.180 align:middle line:90% unpleasant feeling. 00:00:57.180 --> 00:00:59.550 align:middle line:84% We don't want to experience it. 00:00:59.550 --> 00:01:00.790 align:middle line:90% We try to push it away. 00:01:00.790 --> 00:01:06.480 align:middle line:84% That childhood feeling follows us through our developing 00:01:06.480 --> 00:01:07.640 align:middle line:90% years, our adult years. 00:01:07.640 --> 00:01:09.030 align:middle line:84% We bring it into the workplace. 00:01:09.030 --> 00:01:14.110 align:middle line:84% And again, for reasons that are both emotional and quite 00:01:14.110 --> 00:01:16.530 align:middle line:84% logical, connected with promotions and other 00:01:16.530 --> 00:01:18.910 align:middle line:84% activities, we don't want to be blamed for 00:01:18.910 --> 00:01:20.740 align:middle line:90% things going wrong. 00:01:20.740 --> 00:01:22.890 align:middle line:84% SARAH GREEN: You mention that there is a sort of false 00:01:22.890 --> 00:01:25.690 align:middle line:84% dichotomy here that's also at work. 00:01:25.690 --> 00:01:28.330 align:middle line:84% Can you talk a little bit more about how we get stuck in that 00:01:28.330 --> 00:01:30.150 align:middle line:90% and what that exactly is? 00:01:30.150 --> 00:01:33.150 align:middle line:84% AMY EDMONDSON: Here's the dilemma, most managers 00:01:33.150 --> 00:01:35.710 align:middle line:84% understand logically that it's important to learn from the 00:01:35.710 --> 00:01:37.430 align:middle line:90% failures that do occur. 00:01:37.430 --> 00:01:41.250 align:middle line:84% At the same time, they're worried that if they are open 00:01:41.250 --> 00:01:43.930 align:middle line:84% and accepting of failures, they'll create a sort of 00:01:43.930 --> 00:01:46.460 align:middle line:84% anything goes atmosphere, where there's not high 00:01:46.460 --> 00:01:50.040 align:middle line:84% standards and people can pretty much perform at 00:01:50.040 --> 00:01:52.000 align:middle line:84% whatever level they want and that'll be fine. 00:01:52.000 --> 00:01:54.420 align:middle line:90% That's a false dichotomy. 00:01:54.420 --> 00:01:59.650 align:middle line:84% So that assumes that would be a true worry in a world where, 00:01:59.650 --> 00:02:01.780 align:middle line:84% in advance, we could know exactly what to do 00:02:01.780 --> 00:02:02.770 align:middle line:90% and how to do it. 00:02:02.770 --> 00:02:05.450 align:middle line:84% And those who didn't do it should certainly be held 00:02:05.450 --> 00:02:08.580 align:middle line:84% accountable for not doing it precisely right. 00:02:08.580 --> 00:02:12.360 align:middle line:84% In reality, and in the workplace that we now face, a 00:02:12.360 --> 00:02:14.490 align:middle line:84% minority of work falls into that category. 00:02:14.490 --> 00:02:16.640 align:middle line:84% Most work is of the kind where there are certain 00:02:16.640 --> 00:02:17.800 align:middle line:90% uncertainties. 00:02:17.800 --> 00:02:18.920 align:middle line:90% There's lots of novelty. 00:02:18.920 --> 00:02:21.570 align:middle line:84% There's lots of need for creativity and new ideas. 00:02:21.570 --> 00:02:25.500 align:middle line:84% And it's a given that some portion of that new 00:02:25.500 --> 00:02:27.490 align:middle line:90% work will go wrong. 00:02:27.490 --> 00:02:29.620 align:middle line:84% And that's a given that we want to learn from it. 00:02:29.620 --> 00:02:32.360 align:middle line:84% It's not a given that we do learn from it. 00:02:32.360 --> 00:02:35.380 align:middle line:84% SARAH GREEN: So when we are trying to learn from it, in 00:02:35.380 --> 00:02:37.340 align:middle line:84% addition to the blame game, what are some of the other 00:02:37.340 --> 00:02:39.510 align:middle line:90% things that get in the way? 00:02:39.510 --> 00:02:41.640 align:middle line:84% AMY EDMONDSON: One of the important things that gets in 00:02:41.640 --> 00:02:44.530 align:middle line:84% the way of learning from failure, is that we don't have 00:02:44.530 --> 00:02:48.660 align:middle line:84% or practice the skills for doing it very often. 00:02:48.660 --> 00:02:51.030 align:middle line:84% And by skills, maybe I should really say activities, the 00:02:51.030 --> 00:02:55.840 align:middle line:84% activities that are very basic and very obvious, and yet, 00:02:55.840 --> 00:02:57.330 align:middle line:90% often underutilized. 00:02:57.330 --> 00:03:01.330 align:middle line:84% And so the skills I talk about in the article are the skills, 00:03:01.330 --> 00:03:03.500 align:middle line:84% organizationally and individually, of detecting 00:03:03.500 --> 00:03:07.600 align:middle line:84% failure, analyzing failure-- really skillfully, fully, 00:03:07.600 --> 00:03:08.500 align:middle line:90% thoroughly-- 00:03:08.500 --> 00:03:11.170 align:middle line:90% and creating failure. 00:03:11.170 --> 00:03:12.280 align:middle line:84% Why would you want to create failure? 00:03:12.280 --> 00:03:15.690 align:middle line:84% We want to create failure because when you experiment-- 00:03:15.690 --> 00:03:19.080 align:middle line:84% which you need to do under conditions of uncertainty and 00:03:19.080 --> 00:03:20.750 align:middle line:90% novelty or both-- 00:03:20.750 --> 00:03:23.430 align:middle line:84% some of those experiments will produce failures. 00:03:23.430 --> 00:03:25.610 align:middle line:84% We hope they'll be intelligence failures. 00:03:25.610 --> 00:03:27.460 align:middle line:84% But we know that we'll get failures. 00:03:27.460 --> 00:03:33.100 align:middle line:84% So those three activities, detecting failure, analyzing 00:03:33.100 --> 00:03:37.450 align:middle line:84% failure, and producing intelligent failure are skills 00:03:37.450 --> 00:03:39.710 align:middle line:84% that we can learn, that we can practice, but often they are 00:03:39.710 --> 00:03:41.680 align:middle line:90% given short shrift. 00:03:41.680 --> 00:03:43.645 align:middle line:84% SARAH GREEN: I'd like to get into a sort of case 00:03:43.645 --> 00:03:44.440 align:middle line:90% study, if we could. 00:03:44.440 --> 00:03:48.190 align:middle line:84% I know that you did an extensive study of the 00:03:48.190 --> 00:03:50.720 align:middle line:90% Columbia spatial explosion. 00:03:50.720 --> 00:03:52.300 align:middle line:84% What did you find as a result of that? 00:03:52.300 --> 00:03:55.650 align:middle line:84% I mean, could it have been averted, do you think? 00:03:55.650 --> 00:03:57.440 align:middle line:84% AMY EDMONDSON: It absolutely could have been averted. 00:03:57.440 --> 00:04:01.050 align:middle line:84% So there's two levels of analysis on 00:04:01.050 --> 00:04:01.890 align:middle line:90% the Columbia shuttle. 00:04:01.890 --> 00:04:06.350 align:middle line:84% There's two ways we can make the claim that the tragedy 00:04:06.350 --> 00:04:07.870 align:middle line:90% could have been averted. 00:04:07.870 --> 00:04:12.530 align:middle line:84% One is, let's go back in time, and say that had NASA 00:04:12.530 --> 00:04:18.209 align:middle line:84% responded to the many, many very small and seemingly 00:04:18.209 --> 00:04:23.320 align:middle line:84% unimportant foam strike failures. 00:04:23.320 --> 00:04:27.580 align:middle line:84% A foam strike failure is a time that a bit of foam comes 00:04:27.580 --> 00:04:31.660 align:middle line:84% off the solid rocket booster and hits something. 00:04:31.660 --> 00:04:34.110 align:middle line:84% Generally it might hit the ground. 00:04:34.110 --> 00:04:37.030 align:middle line:84% In this case of the Columbia shuttle, it hit the orbiter, 00:04:37.030 --> 00:04:38.040 align:middle line:90% the shuttle. 00:04:38.040 --> 00:04:42.250 align:middle line:84% And we now know, left a big hole, which meant that when 00:04:42.250 --> 00:04:45.990 align:middle line:84% the shuttle came back into the atmosphere, it burnt 00:04:45.990 --> 00:04:47.960 align:middle line:90% immediately on reentry. 00:04:47.960 --> 00:04:51.080 align:middle line:84% So two ways in which that could have been averted, one 00:04:51.080 --> 00:04:54.460 align:middle line:84% is that throughout the many years of the shuttle program, 00:04:54.460 --> 00:04:58.330 align:middle line:84% when these small failures were being essentially ignored-- 00:04:58.330 --> 00:05:01.380 align:middle line:84% they were discussed a little bit, but held at bay-- 00:05:01.380 --> 00:05:06.180 align:middle line:84% they probably won't amount to anything, probably won't turn 00:05:06.180 --> 00:05:07.190 align:middle line:90% to a major failure. 00:05:07.190 --> 00:05:09.390 align:middle line:84% And we have too many other things to think about. 00:05:09.390 --> 00:05:12.400 align:middle line:84% Most managers can identify with that statement. 00:05:12.400 --> 00:05:19.620 align:middle line:84% So had those early warnings of a problematic phenomenon been 00:05:19.620 --> 00:05:24.110 align:middle line:84% really dug into, they could have found a way to not have 00:05:24.110 --> 00:05:27.070 align:middle line:84% it keep happening, and then therefore, not happen in the 00:05:27.070 --> 00:05:31.800 align:middle line:90% Columbia 2003 shuttle launch. 00:05:31.800 --> 00:05:33.480 align:middle line:90% So that's one level. 00:05:33.480 --> 00:05:34.890 align:middle line:90% It's long over time. 00:05:34.890 --> 00:05:38.550 align:middle line:84% It's the kind of small things that organizations ignore for 00:05:38.550 --> 00:05:39.230 align:middle line:90% a long time. 00:05:39.230 --> 00:05:41.470 align:middle line:84% They don't seem that important, don't seem worth 00:05:41.470 --> 00:05:44.200 align:middle line:84% spending the resources and time on fixing. 00:05:44.200 --> 00:05:52.080 align:middle line:84% The other is that upon launch in January of 2003, this is 00:05:52.080 --> 00:05:56.010 align:middle line:84% fairly large, in fact much larger, 200 times larger than 00:05:56.010 --> 00:06:01.410 align:middle line:84% any prior foam strike, came off and may 00:06:01.410 --> 00:06:02.690 align:middle line:90% have hit the shuttle. 00:06:02.690 --> 00:06:05.580 align:middle line:84% So the engineers were watching the videotape of the launch 00:06:05.580 --> 00:06:07.020 align:middle line:90% that the next day. 00:06:07.020 --> 00:06:11.220 align:middle line:84% And the videotape was grainy, and inexact, and they couldn't 00:06:11.220 --> 00:06:12.200 align:middle line:90% really tell. 00:06:12.200 --> 00:06:14.650 align:middle line:84% But they were worried, particularly Rodney Rocha, who 00:06:14.650 --> 00:06:17.585 align:middle line:84% was an expert in the foam insulation. 00:06:17.585 --> 00:06:22.500 align:middle line:84% And he watched the video over, and over, and over, and didn't 00:06:22.500 --> 00:06:23.950 align:middle line:90% feel good about it. 00:06:23.950 --> 00:06:29.570 align:middle line:84% He then tried to get more senior manager attention to 00:06:29.570 --> 00:06:30.320 align:middle line:90% the problem. 00:06:30.320 --> 00:06:35.180 align:middle line:84% He wanted a full-out, deep-dive, let's find out 00:06:35.180 --> 00:06:39.340 align:middle line:84% whether or not this created damage to the shuttle. 00:06:39.340 --> 00:06:43.050 align:middle line:84% And if so, then we'll figure out what to do about it. 00:06:43.050 --> 00:06:45.540 align:middle line:90% He was unable to get-- 00:06:45.540 --> 00:06:48.570 align:middle line:84% and his strategies were not good enough-- but he was 00:06:48.570 --> 00:06:55.650 align:middle line:84% unable to get attention to that potential, not necessary 00:06:55.650 --> 00:06:58.020 align:middle line:84% failure, but that potential failure. 00:06:58.020 --> 00:07:00.910 align:middle line:84% Had he gotten attention to it, the Columbia Accident 00:07:00.910 --> 00:07:04.030 align:middle line:84% Investigation Board report that came out a few months 00:07:04.030 --> 00:07:07.560 align:middle line:84% later suggested that there were two possible rescue 00:07:07.560 --> 00:07:11.260 align:middle line:90% scenarios for the crew. 00:07:11.260 --> 00:07:15.370 align:middle line:84% Neither one was an obvious slam dunk, 00:07:15.370 --> 00:07:16.960 align:middle line:90% it'll work, no worries. 00:07:16.960 --> 00:07:21.180 align:middle line:84% But they were both feasible and worthy of effort, was 00:07:21.180 --> 00:07:24.170 align:middle line:90% within the CAIB concluded. 00:07:24.170 --> 00:07:27.540 align:middle line:84% SARAH GREEN: So in a case like a space shuttle explosion, 00:07:27.540 --> 00:07:29.780 align:middle line:84% which is, as you point out, not just a failure but a 00:07:29.780 --> 00:07:32.920 align:middle line:84% tragedy involving a loss of human life. 00:07:32.920 --> 00:07:37.290 align:middle line:84% In any failure like that when lives are on the line, is 00:07:37.290 --> 00:07:40.830 align:middle line:84% there a qualitative difference in sort of how we respond to 00:07:40.830 --> 00:07:42.320 align:middle line:90% it or learn from it? 00:07:42.320 --> 00:07:46.320 align:middle line:84% Or is it just sort of an order of magnitude difference from a 00:07:46.320 --> 00:07:49.320 align:middle line:84% sort of typical business failure? 00:07:49.320 --> 00:07:52.990 align:middle line:84% AMY EDMONDSON: I think there are mostly order of magnitude 00:07:52.990 --> 00:07:57.290 align:middle line:84% differences in terms of the scale, or size, or magnitude 00:07:57.290 --> 00:08:00.230 align:middle line:90% of harm of a failure. 00:08:00.230 --> 00:08:03.520 align:middle line:84% Learning from failures requires essentially the same 00:08:03.520 --> 00:08:06.430 align:middle line:84% kinds of practices and processes. 00:08:06.430 --> 00:08:10.630 align:middle line:84% Obviously we're more concerned about failures that have 00:08:10.630 --> 00:08:14.170 align:middle line:84% bigger consequences, and particularly when they lead to 00:08:14.170 --> 00:08:16.400 align:middle line:90% loss of human life. 00:08:16.400 --> 00:08:22.240 align:middle line:84% However, the strategy and the desire to learn from them are 00:08:22.240 --> 00:08:23.860 align:middle line:90% identical, I would argue. 00:08:23.860 --> 00:08:26.770 align:middle line:84% And the reason is you want to develop those skills, those 00:08:26.770 --> 00:08:29.870 align:middle line:84% practices, so that it becomes second nature, it becomes 00:08:29.870 --> 00:08:33.789 align:middle line:84% habitual to learn and learn thoroughly from the small and 00:08:33.789 --> 00:08:35.900 align:middle line:90% large failures that do occur. 00:08:35.900 --> 00:08:38.070 align:middle line:84% SARAH GREEN: So it sounds like one of the key things is 00:08:38.070 --> 00:08:39.690 align:middle line:90% getting people to speak up. 00:08:39.690 --> 00:08:42.750 align:middle line:84% How can leaders really make it an environment where people 00:08:42.750 --> 00:08:45.960 align:middle line:84% can speak and where they're heard. 00:08:45.960 --> 00:08:49.900 align:middle line:84% AMY EDMONDSON: It is a crucial question, how can leaders 00:08:49.900 --> 00:08:52.870 align:middle line:84% create an environment where people feel free and 00:08:52.870 --> 00:08:58.520 align:middle line:84% comfortable to speak up with not just certain disaster, but 00:08:58.520 --> 00:09:01.710 align:middle line:84% their worries, their concerns, their maybe this isn't quite 00:09:01.710 --> 00:09:04.050 align:middle line:90% right kinds of ideas. 00:09:04.050 --> 00:09:07.930 align:middle line:84% And in the article, I suggest a set of strategies that I 00:09:07.930 --> 00:09:09.410 align:middle line:90% really do consider, from my 00:09:09.410 --> 00:09:11.430 align:middle line:90% research, as the best practices. 00:09:11.430 --> 00:09:16.580 align:middle line:84% And they start, interestingly, with framing the work 00:09:16.580 --> 00:09:17.760 align:middle line:90% accurately. 00:09:17.760 --> 00:09:20.650 align:middle line:84% And by that, I mean leaders are in a position where they 00:09:20.650 --> 00:09:25.520 align:middle line:84% can explain the meaning and nature of the work that we do. 00:09:25.520 --> 00:09:30.960 align:middle line:84% If we do high volume manufacturing of automobiles, 00:09:30.960 --> 00:09:33.990 align:middle line:84% and we want them to be exactly the same and perfect every 00:09:33.990 --> 00:09:37.650 align:middle line:84% time, that suggests one kind of work environment. 00:09:37.650 --> 00:09:40.000 align:middle line:84% That suggests an environment where it's got to be 00:09:40.000 --> 00:09:43.240 align:middle line:84% absolutely safe for people to speak up when they see 00:09:43.240 --> 00:09:46.580 align:middle line:84% something that looks just a little off, a little wrong, 00:09:46.580 --> 00:09:49.800 align:middle line:84% and to sort of stop and immediately figure out how to 00:09:49.800 --> 00:09:50.930 align:middle line:90% get to the bottom of it. 00:09:50.930 --> 00:09:54.320 align:middle line:84% If we're doing work that's in a scientific laboratory, and 00:09:54.320 --> 00:09:58.790 align:middle line:84% we can assume that 70% of everything we try will go 00:09:58.790 --> 00:10:02.880 align:middle line:84% wrong, will fail, then the leader's job is to make sure 00:10:02.880 --> 00:10:05.720 align:middle line:84% everybody realizes that so that they are in a position to 00:10:05.720 --> 00:10:09.100 align:middle line:84% make more failures more quickly than the competition 00:10:09.100 --> 00:10:10.560 align:middle line:90% and have discovery faster. 00:10:10.560 --> 00:10:14.280 align:middle line:84% So job one for the leader is to start with frame the work 00:10:14.280 --> 00:10:15.020 align:middle line:90% accurately. 00:10:15.020 --> 00:10:20.310 align:middle line:84% Tell the story of what kind of work we do and how failures 00:10:20.310 --> 00:10:21.190 align:middle line:90% fit into that. 00:10:21.190 --> 00:10:24.140 align:middle line:84% The second thing is embrace messengers. 00:10:24.140 --> 00:10:25.530 align:middle line:84% Don't shoot them, embrace them. 00:10:25.530 --> 00:10:29.520 align:middle line:84% It will never feel like a happy day when someone comes 00:10:29.520 --> 00:10:31.460 align:middle line:84% to you and says something's wrong. 00:10:31.460 --> 00:10:34.680 align:middle line:84% Resist the urge to get annoyed, frustrated, and take 00:10:34.680 --> 00:10:35.740 align:middle line:90% it out on them. 00:10:35.740 --> 00:10:39.110 align:middle line:84% Instead celebrate and thank their willingness, their 00:10:39.110 --> 00:10:42.000 align:middle line:90% courage to come forward. 00:10:42.000 --> 00:10:47.860 align:middle line:84% A third strategy is to set up structures and processes 00:10:47.860 --> 00:10:52.610 align:middle line:84% through which the failures that do occur are learned from 00:10:52.610 --> 00:10:58.060 align:middle line:84% carefully, multidisciplinary teams, systems, protocols that 00:10:58.060 --> 00:10:59.170 align:middle line:90% help people do this well. 00:10:59.170 --> 00:11:03.370 align:middle line:84% Because our inclination is to just do it quickly and get 00:11:03.370 --> 00:11:04.890 align:middle line:90% back to the real work. 00:11:04.890 --> 00:11:07.640 align:middle line:90% That is part of the real work. 00:11:07.640 --> 00:11:09.920 align:middle line:84% SARAH GREEN: You raise an interesting point. 00:11:09.920 --> 00:11:13.080 align:middle line:84% We're talking about what the leader can do. 00:11:13.080 --> 00:11:17.320 align:middle line:84% But actually, as you sort of imply, it can be hard for 00:11:17.320 --> 00:11:20.830 align:middle line:84% someone who's not a leader, or not even a manager, to bring 00:11:20.830 --> 00:11:23.550 align:middle line:84% up bad news without feeling like a gadfly. 00:11:23.550 --> 00:11:27.230 align:middle line:84% If you're just a sort of worker bee, how can you make 00:11:27.230 --> 00:11:31.040 align:middle line:84% your company culture more tolerant to failure? 00:11:31.040 --> 00:11:33.480 align:middle line:84% AMY EDMONDSON: I think you try it. 00:11:33.480 --> 00:11:34.810 align:middle line:90% You do what you can. 00:11:34.810 --> 00:11:36.790 align:middle line:90% As an employee in any large 00:11:36.790 --> 00:11:38.240 align:middle line:90% organization, it won't be easy. 00:11:38.240 --> 00:11:40.850 align:middle line:84% It will take a little bit of courage, in some cases, a lot 00:11:40.850 --> 00:11:41.880 align:middle line:90% of courage. 00:11:41.880 --> 00:11:46.340 align:middle line:84% But by coming forward, by raising the things that you 00:11:46.340 --> 00:11:49.530 align:middle line:84% see or are concerned about by asking questions, by asking 00:11:49.530 --> 00:11:54.000 align:middle line:84% for help, it can be, if you do it with enthusiasm and you do 00:11:54.000 --> 00:11:58.800 align:middle line:84% with the intent to make things better, which many if not most 00:11:58.800 --> 00:12:01.350 align:middle line:90% people will, it's infectious. 00:12:01.350 --> 00:12:02.940 align:middle line:90% And your peers will do it too. 00:12:02.940 --> 00:12:05.830 align:middle line:84% And certainly if you're in a position where you have some 00:12:05.830 --> 00:12:10.810 align:middle line:84% people reporting to you, it makes a very real impact on 00:12:10.810 --> 00:12:13.930 align:middle line:84% their behaviors and their assumptions. 00:12:13.930 --> 00:12:15.590 align:middle line:84% SARAH GREEN: Well Professor Edmondson, I think that's kind 00:12:15.590 --> 00:12:16.800 align:middle line:90% of nice note to leave it on. 00:12:16.800 --> 00:12:18.610 align:middle line:84% Thanks again so much for coming in. 00:12:18.610 --> 00:12:20.570 align:middle line:84% AMY EDMONDSON: Thank you for having me. 00:12:20.570 --> 00:12:21.870 align:middle line:84% SARAH GREEN: That's Harvard Business 00:12:21.870 --> 00:12:23.570 align:middle line:90% School's Amy Edmondson. 00:12:23.570 --> 00:12:26.820 align:middle line:84% And the article Strategies for Learning from Failure appears 00:12:26.820 --> 00:12:29.980 align:middle line:84% in the April, 2011 of Harvard Business Review. 00:12:29.980 --> 00:12:31.230 align:middle line:90% For more, go to hbr.org. 00:12:31.230 --> 00:12:37.994 align:middle line:90%