WEBVTT 00:00:00.000 --> 00:00:03.160 Welcome to the deep dive. Today, we're tackling 00:00:03.160 --> 00:00:06.820 something that's, well, both incredibly basic 00:00:06.820 --> 00:00:09.839 and somehow massively misunderstood in organizational 00:00:09.839 --> 00:00:14.439 improvement. The Plando Check Act cycle. PDCA. 00:00:14.820 --> 00:00:16.960 PDCA. It's pretty much the bedrock, the core 00:00:16.960 --> 00:00:19.519 framework for, you know, everything from quality 00:00:19.519 --> 00:00:22.539 control to lean manufacturing. Exactly. But here's 00:00:22.539 --> 00:00:24.440 the strange thing we found in the research, a 00:00:24.440 --> 00:00:27.480 real paradox. Which is? You can't really run 00:00:27.480 --> 00:00:30.120 a successful improvement project without PDCA, 00:00:30.219 --> 00:00:32.799 yet most projects that say they're using it, 00:00:32.799 --> 00:00:35.899 well, they fail. They do. And that's our mission 00:00:35.899 --> 00:00:39.020 today, isn't it? To cut through that confusion 00:00:39.020 --> 00:00:41.640 and show you exactly why they fail. We're really 00:00:41.640 --> 00:00:43.659 going deep on the fundamentals that just keep 00:00:43.659 --> 00:00:45.920 getting missed. And the cost of missing them 00:00:45.920 --> 00:00:47.780 is huge. I mean, the documentation we looked 00:00:47.780 --> 00:00:49.780 at is pretty clear. Most big improvement projects 00:00:49.780 --> 00:00:52.079 don't fail because they lacked some fancy software. 00:00:52.119 --> 00:00:53.759 No, not at all. They fail because they didn't 00:00:53.759 --> 00:00:57.159 properly follow this simple four step loop. So 00:00:57.159 --> 00:00:59.600 we're going to unpack each step, but really zero 00:00:59.600 --> 00:01:01.820 in on where things tend to go off the rails. 00:01:02.259 --> 00:01:05.019 Absolutely. It's about moving past Just knowing 00:01:05.019 --> 00:01:09.819 the letters, PDCA, and understanding the rigor 00:01:09.819 --> 00:01:13.359 needed for each part. If you rush even one step, 00:01:13.599 --> 00:01:16.620 especially those later ones, your project's chances 00:01:16.620 --> 00:01:20.560 of actually sticking, of being sustainable, pretty 00:01:20.560 --> 00:01:23.120 much zero. Okay, let's get into it then. Starting 00:01:23.120 --> 00:01:26.000 with P Plan. Now when most people hear plan, 00:01:26.120 --> 00:01:28.370 they probably think, yeah. Got it. We know the 00:01:28.370 --> 00:01:30.109 problem. Let's just start doing stuff. Right. 00:01:30.209 --> 00:01:32.530 The action bias. But the source material suggests 00:01:32.530 --> 00:01:36.329 this plan phase. It's massive. So big, someone 00:01:36.329 --> 00:01:38.829 commented, it maybe needs more letters. Why is 00:01:38.829 --> 00:01:41.900 that? Because plan. when you do it right, can 00:01:41.900 --> 00:01:44.840 easily eat up, say, 50, maybe even 70 % of the 00:01:44.840 --> 00:01:47.439 whole effort. It's not just jotting down an idea. 00:01:47.579 --> 00:01:50.079 It's this whole iterative cycle of analysis. 00:01:50.260 --> 00:01:52.439 And if you skip parts of that, well, you've pretty 00:01:52.439 --> 00:01:54.140 much doomed your project before you even begin 00:01:54.140 --> 00:01:56.239 the due phase. And hang on. Don't most places 00:01:56.239 --> 00:01:59.019 think their regular reports and dashboards are 00:01:59.019 --> 00:02:02.599 enough? Why this really intensive upfront analysis 00:02:02.599 --> 00:02:06.079 in the plan phase? What makes it so vital? Ah. 00:02:06.349 --> 00:02:08.750 That's where the rigor comes in. See, the plan 00:02:08.750 --> 00:02:12.370 phase isn't just one thing. It needs like five 00:02:12.370 --> 00:02:14.909 essential sub -steps. First, you gotta define 00:02:14.909 --> 00:02:17.389 the scope. What exactly are you looking at? Which 00:02:17.389 --> 00:02:20.770 machine? Which process? Okay, there we go. Precisely. 00:02:21.009 --> 00:02:24.889 Second, define the target. What specific measurable 00:02:24.889 --> 00:02:27.569 goal are you actually trying to hit? If you skip 00:02:27.569 --> 00:02:30.810 those two, you're basically just wandering around 00:02:30.810 --> 00:02:33.400 hoping for the best. And the third step. That 00:02:33.400 --> 00:02:35.300 seems like the one that really pushes back against 00:02:35.300 --> 00:02:37.319 how things usually work in an office, right? 00:02:37.699 --> 00:02:40.780 Sitting behind a desk. It absolutely does. Step 00:02:40.780 --> 00:02:43.919 three is analyzing the current situation. And 00:02:43.919 --> 00:02:47.080 this is an optional. It must involve Genchi Genbutsu. 00:02:47.280 --> 00:02:49.060 Go into the source. Go into the source. Yeah. 00:02:49.340 --> 00:02:51.080 To the shop floor, the call center, wherever 00:02:51.080 --> 00:02:53.159 the actual work happens. You have to watch. You 00:02:53.159 --> 00:02:55.180 have to talk to the people doing the job. Collect 00:02:55.180 --> 00:02:57.039 data right there. Get your boots on the ground. 00:02:57.139 --> 00:02:59.419 Don't just, you know, stare at spreadsheets someone 00:02:59.419 --> 00:03:01.900 else made. So if you skip Genchi Genbutsu, you 00:03:01.900 --> 00:03:04.120 might be solving a problem that looks real on 00:03:04.120 --> 00:03:06.500 paper. But doesn't actually exist that way in 00:03:06.500 --> 00:03:09.340 reality. Yeah. Or you miss the real problem entirely. 00:03:09.620 --> 00:03:12.300 Exactly. So once you have that deep observed 00:03:12.300 --> 00:03:15.199 understanding, Then you do the last two parts 00:03:15.199 --> 00:03:18.060 of plan, develop a few potential solutions, different 00:03:18.060 --> 00:03:21.460 ways to tackle it, and then critically select 00:03:21.460 --> 00:03:24.120 the best one. Usually that's the one offering 00:03:24.120 --> 00:03:27.379 the biggest bang for your buck. Best results 00:03:27.379 --> 00:03:29.939 for the effort involved. Okay, so that intense 00:03:29.939 --> 00:03:32.400 planning sets the stage. That's our blueprint. 00:03:33.080 --> 00:03:37.039 Now we move to D, the implementation. Yep, D 00:03:37.039 --> 00:03:39.479 is where the rubber meets the road. You actually 00:03:39.479 --> 00:03:42.180 implement the chosen solution. You change the 00:03:42.180 --> 00:03:44.780 process, maybe pilot the new equipment. Sounds 00:03:44.780 --> 00:03:46.979 straightforward enough. Well, the material really 00:03:46.979 --> 00:03:50.060 stresses something here. Problems will pop up 00:03:50.060 --> 00:03:52.560 during due. Things you didn't think of and plan. 00:03:53.180 --> 00:03:55.639 And that's okay. That's not failure. No, it's 00:03:55.639 --> 00:03:57.680 normal. The goal during due is just to tackle 00:03:57.680 --> 00:03:59.800 those unexpected little fires as they come up 00:03:59.800 --> 00:04:02.379 and keep the implementation moving forward. But 00:04:02.379 --> 00:04:04.460 the really crucial bit, it seems, isn't just 00:04:04.460 --> 00:04:07.379 making the change happen once. It's making it 00:04:07.379 --> 00:04:10.199 stick, right? How do you make sure the new way 00:04:10.199 --> 00:04:13.319 actually becomes the standard way? Ah, that's 00:04:13.319 --> 00:04:17.839 key. Through really rigorous standardization 00:04:17.839 --> 00:04:20.319 and training. Look, it's easy to get people to 00:04:20.319 --> 00:04:22.300 try something new for an hour when you're watching. 00:04:22.360 --> 00:04:24.779 Sure. It's way, way harder to get them to actually 00:04:24.779 --> 00:04:28.220 adopt it permanently. So you have to create the 00:04:28.220 --> 00:04:31.500 new standard operating procedure, the SOP. train 00:04:31.500 --> 00:04:33.819 people on it thoroughly, and then this is critical. 00:04:34.060 --> 00:04:36.379 You have to go back and confirm they're still 00:04:36.379 --> 00:04:39.300 following that standard a few days later, not 00:04:39.300 --> 00:04:40.579 just while you're standing over their shoulder. 00:04:40.759 --> 00:04:43.420 And checking if they're still doing it. That 00:04:43.420 --> 00:04:46.259 leads us straight into C -Check. And this is 00:04:46.259 --> 00:04:48.860 where, well, the documentation practically screams 00:04:48.860 --> 00:04:50.899 warnings at us. It really does. It seems to be 00:04:50.899 --> 00:04:53.459 the moment most projects just sabotage themselves. 00:04:53.680 --> 00:04:55.399 It is. It's the most frequently skipped step, 00:04:55.620 --> 00:04:57.819 definitely. And I'd argue probably the most important 00:04:57.819 --> 00:05:00.000 one. If we're looking at the big picture, check 00:05:00.000 --> 00:05:02.579 is, well, it's your reality check moment. OK. 00:05:03.079 --> 00:05:04.740 The fundamental question you have to ask is, 00:05:05.740 --> 00:05:08.500 did the solution we implemented actually work? 00:05:09.000 --> 00:05:11.279 Did it hit the targets we set back in the plan 00:05:11.279 --> 00:05:14.500 phase? And crucially, did it sustain those results? 00:05:14.839 --> 00:05:17.879 But if managers are supposedly so committed to 00:05:17.879 --> 00:05:19.899 improvement, why would they skip the step where 00:05:19.899 --> 00:05:22.240 they actually check if it worked? I mean, what 00:05:22.240 --> 00:05:24.180 makes them think a project is successful even 00:05:24.180 --> 00:05:27.060 if maybe it isn't really delivering? Ah, well, 00:05:27.459 --> 00:05:29.560 sometimes the mere attention from management 00:05:29.560 --> 00:05:33.779 creates this illusion of success. This brings 00:05:33.779 --> 00:05:35.660 us to something called the Hawthorne effect trap. 00:05:35.920 --> 00:05:39.120 The Hawthorne effect, right. What exactly is 00:05:39.120 --> 00:05:41.439 that and why is it flagged as such a danger here? 00:05:41.819 --> 00:05:45.040 So the Hawthorne effect. It goes way back, observed, 00:05:45.079 --> 00:05:47.079 I think, in the 1930s at the Hawthorne Works, 00:05:47.279 --> 00:05:49.519 a Western electric plant. Basically, it showed 00:05:49.519 --> 00:05:52.660 that just the act of management paying attention 00:05:52.660 --> 00:05:55.379 to a group of workers, implementing any kind 00:05:55.379 --> 00:05:57.459 of change, really, whether it was good or bad, 00:05:58.339 --> 00:06:00.439 temporarily caused productivity to go up. Just 00:06:00.439 --> 00:06:02.120 because they were being watched? Pretty much. 00:06:02.519 --> 00:06:05.180 So imagine the VP is touring the area every day, 00:06:05.420 --> 00:06:08.000 asking for updates on the new process. For that 00:06:08.000 --> 00:06:11.209 week, guess what? The KPIs, the metrics, they 00:06:11.209 --> 00:06:13.029 probably look fantastic. OK, I see the trap. 00:06:13.290 --> 00:06:15.329 Right. But that improvement might have nothing 00:06:15.329 --> 00:06:17.550 to do with how good the new process actually 00:06:17.550 --> 00:06:20.189 is. It's purely down to that temporary spotlight 00:06:20.189 --> 00:06:22.910 effect. Which means, as soon as the VP moves 00:06:22.910 --> 00:06:25.629 on to the next big thing, the next crisis. Exactly. 00:06:25.769 --> 00:06:28.170 The spotlight moves, the old habits creep back 00:06:28.170 --> 00:06:30.910 in, things get comfortable again, and bam, those 00:06:30.910 --> 00:06:33.269 amazing KPIs, they often slide right back to 00:06:33.269 --> 00:06:37.089 where they started. Or worse. Ouch. Yeah. So, 00:06:37.389 --> 00:06:40.189 a real sustained improvement has to keep working 00:06:40.189 --> 00:06:43.209 long after that initial attention fades. And 00:06:43.209 --> 00:06:46.589 that's why a rigorous, deep, and maybe even long 00:06:46.589 --> 00:06:49.189 -term check phase is absolutely non -negotiable. 00:06:49.410 --> 00:06:51.009 You have to measure the results, the output, 00:06:51.189 --> 00:06:53.110 over enough time to prove the solution holds 00:06:53.110 --> 00:06:55.569 up on its own. Okay, so the check gives us the 00:06:55.569 --> 00:06:57.550 hard truth, which finally brings us to the last 00:06:57.550 --> 00:07:00.170 step, a act. This is the decision point, right? 00:07:00.350 --> 00:07:02.449 Based purely on what that sustained check told 00:07:02.449 --> 00:07:04.810 us. Precisely. Act is where you decide what happens 00:07:04.810 --> 00:07:09.180 next. If your check phase revealed, maybe uncomfortably, 00:07:09.560 --> 00:07:11.079 that the implementation didn't really meet the 00:07:11.079 --> 00:07:13.240 target sustainably, then you have to stop. You 00:07:13.240 --> 00:07:15.199 can't just ignore it. You have to dig in, find 00:07:15.199 --> 00:07:17.800 the root cause, why it didn't perform as expected. 00:07:18.220 --> 00:07:20.759 And that discovery, that immediately kicks off 00:07:20.759 --> 00:07:23.600 a new plan phase. The PDCA loop starts all over 00:07:23.600 --> 00:07:26.629 again, but this time you're smarter. OK, so failure 00:07:26.629 --> 00:07:29.129 leads back to planning. But what if check shows 00:07:29.129 --> 00:07:32.269 it was success? You've got proof, sustained results. 00:07:32.670 --> 00:07:34.810 The new standard works even when no one's looking 00:07:34.810 --> 00:07:37.050 over their shoulder. Well then, first things 00:07:37.050 --> 00:07:40.170 first. You stop everything and you celebrate 00:07:40.170 --> 00:07:45.639 the team. Yeah. Success needs recognition. Seriously, 00:07:45.839 --> 00:07:48.019 the research we read really hammered this point 00:07:48.019 --> 00:07:50.279 home. Celebrating success. Yeah, acknowledging 00:07:50.279 --> 00:07:52.639 it is critical for keeping people motivated. 00:07:52.699 --> 00:07:54.680 There's actually a great little anecdote in there. 00:07:55.100 --> 00:07:57.220 Something about how effective a simple three 00:07:57.220 --> 00:07:59.579 -pound bucket of gummy bears or similar could 00:07:59.579 --> 00:08:02.500 be for a shop floor team. Gummy bears? Really? 00:08:02.680 --> 00:08:04.319 Uh -huh. Yeah, it's not about the, you know, 00:08:04.459 --> 00:08:06.800 the monetary value. It's the immediate, tangible, 00:08:07.060 --> 00:08:09.120 hey, your hard work paid off. We checked. It's 00:08:09.120 --> 00:08:11.600 real well done kind of recognition. I like that, 00:08:11.699 --> 00:08:13.839 it's memorable. But okay, gummy bears consumed, 00:08:14.300 --> 00:08:17.079 celebration done. The work isn't over, is it? 00:08:17.420 --> 00:08:20.620 Continuous improvement means continuous. Right, 00:08:20.879 --> 00:08:23.439 so post -success, that newly proven standard, 00:08:23.779 --> 00:08:26.000 that becomes the new baseline, the new normal. 00:08:26.160 --> 00:08:29.300 And then you look around, prioritize the next 00:08:29.300 --> 00:08:31.459 most important problem, usually the one that 00:08:31.459 --> 00:08:33.259 promises the best outcome for the effort needed, 00:08:33.460 --> 00:08:36.220 and you kick off a brand new PDCA cycle for that. 00:08:36.320 --> 00:08:38.399 On to the next improvement. It all sounds so 00:08:38.399 --> 00:08:41.159 logical step by step, yet you mentioned earlier 00:08:41.159 --> 00:08:44.679 that check and act often get skipped. Why? Beyond 00:08:44.679 --> 00:08:47.080 just wanting to stick to the plan and do phases, 00:08:47.480 --> 00:08:49.759 what makes companies actively resist those last 00:08:49.759 --> 00:08:53.559 two crucial steps? Well, let's be honest. Check 00:08:53.559 --> 00:08:56.399 and act can be... politically tricky, sometimes 00:08:56.399 --> 00:08:58.659 even dangerous for people's careers. How so? 00:08:58.720 --> 00:09:00.740 Because they force the organization to confront 00:09:00.740 --> 00:09:04.279 potential failure. Uncomfortable truths. Imagine 00:09:04.279 --> 00:09:07.440 you spent, say, millions on some fancy new equipment 00:09:07.440 --> 00:09:09.639 based on a plan that looked amazing on paper. 00:09:09.899 --> 00:09:11.980 And then a really thorough, honest check reveals, 00:09:12.279 --> 00:09:14.639 oops, that expensive gear isn't actually delivering 00:09:14.639 --> 00:09:17.000 the promised results. Maybe it's even causing 00:09:17.000 --> 00:09:19.240 new problems. Does management really want that 00:09:19.240 --> 00:09:23.690 news getting out sometimes? Ah. So the goal can 00:09:23.690 --> 00:09:26.370 subtly shift from genuine, sustained improvement 00:09:26.370 --> 00:09:29.789 to just the appearance of improvement to avoid 00:09:29.789 --> 00:09:32.490 blame for costly mistakes. That's a very common 00:09:32.490 --> 00:09:34.929 trap, unfortunately. And this whole struggle, 00:09:35.129 --> 00:09:37.549 this resistance to frankly simple discipline 00:09:37.549 --> 00:09:40.409 of PDCA has led people to look for alternatives, 00:09:41.230 --> 00:09:43.169 competing frameworks, which brings up that whole 00:09:43.169 --> 00:09:47.289 debate around things like DMAIC. Right, DMAIC. 00:09:47.730 --> 00:09:50.090 Define, measure, analyze, improve, control. You 00:09:50.090 --> 00:09:51.990 hear that one a lot, often pitched as maybe more 00:09:51.990 --> 00:09:54.049 modern, more structured than PDCA. Sometimes, 00:09:54.250 --> 00:09:56.169 yes. And the critics of PDCA might say, well, 00:09:56.269 --> 00:09:58.549 plan is too vague. And do an act sound almost 00:09:58.549 --> 00:10:01.570 the same? Why waste letters? Isn't DMAAC's control 00:10:01.570 --> 00:10:03.629 step a better way to think about holding the 00:10:03.629 --> 00:10:06.190 gains than PDCA's act? That's definitely the 00:10:06.190 --> 00:10:07.970 argument some consultants make. They'll pitch 00:10:07.970 --> 00:10:10.950 DMAAC as more logical, maybe easier to teach 00:10:10.950 --> 00:10:13.779 in a workshop. However, the author... of the 00:10:13.779 --> 00:10:17.259 main source we dug into was, let's just say, 00:10:17.399 --> 00:10:19.919 absolutely not impressed with DMAAC. Really? 00:10:19.919 --> 00:10:22.860 Why not? They found the sequence wasn't as stringent, 00:10:22.940 --> 00:10:25.980 not as rigorous. Specifically, they argued that 00:10:25.980 --> 00:10:29.720 DMAAC sort of bundles the entire PDCA cycle into 00:10:29.720 --> 00:10:31.860 its single improved step, which kind of muddies 00:10:31.860 --> 00:10:34.360 the water. OK. But the real deal breaker, the 00:10:34.360 --> 00:10:36.139 biggest criticism, and this brings us right back 00:10:36.139 --> 00:10:39.480 to why projects fail, was that DMAAC's concept 00:10:39.480 --> 00:10:41.600 of checking if the improvement actually worked 00:10:41.600 --> 00:10:45.419 and stuck the control part, was described as, 00:10:45.419 --> 00:10:47.960 and I quote, extremely weak or completely absent. 00:10:48.200 --> 00:10:50.759 Wow. Extremely weak or completely absent. That's 00:10:50.759 --> 00:10:52.899 a pretty strong accusation against a popular 00:10:52.899 --> 00:10:54.960 methodology. It is. And if that core component 00:10:54.960 --> 00:10:58.279 for long -term success, the rigorous check, the 00:10:58.279 --> 00:11:01.039 verification is missing or weak, then you're 00:11:01.039 --> 00:11:03.179 almost designing project failure right into the 00:11:03.179 --> 00:11:05.539 process from the start. So DMAIC might let you 00:11:05.539 --> 00:11:07.840 declare a victory too soon based on maybe just 00:11:07.840 --> 00:11:10.039 the Hawthorne effect? That seems to be the core 00:11:10.039 --> 00:11:14.139 critique. PDCA, by forcing that distinct check 00:11:14.139 --> 00:11:16.059 and then a conscious act based on the check, 00:11:16.399 --> 00:11:19.840 makes you confront the long -term reality. DMAC, 00:11:19.980 --> 00:11:22.179 in this view, can let projects wrap up based 00:11:22.179 --> 00:11:24.980 on those initial, possibly misleading gains. 00:11:25.519 --> 00:11:27.860 Leaves the door wide open to being fooled. Okay, 00:11:27.919 --> 00:11:30.240 so wrapping this up, what does this all mean 00:11:30.240 --> 00:11:32.559 for you, the listener? Whether you're trying 00:11:32.559 --> 00:11:34.860 to optimize a huge production line or just, you 00:11:34.860 --> 00:11:36.919 know, make your own work process a bit smoother. 00:11:37.399 --> 00:11:39.740 PDCA really is the fundamental building block. 00:11:39.799 --> 00:11:42.159 It sounds simple. Maybe even simplistic. But 00:11:42.159 --> 00:11:44.360 doing it right, consistently, that's what the 00:11:44.360 --> 00:11:47.100 real challenge lies. Absolutely. While plan and 00:11:47.100 --> 00:11:49.220 do often happen naturally, people like planning, 00:11:49.480 --> 00:11:51.720 people like doing stuff, getting that true lasting 00:11:51.720 --> 00:11:54.759 improvement. That demands real focus, real discipline 00:11:54.759 --> 00:11:57.340 on the whole cycle. Especially check and act. 00:11:57.480 --> 00:11:58.980 The steps that, let's face it, aren't always 00:11:58.980 --> 00:12:01.559 fun. Yeah, confronting reality isn't always fun. 00:12:02.100 --> 00:12:04.980 So the big takeaway seems to be genuine improvement 00:12:04.980 --> 00:12:07.720 isn't about that first shiny presentation showing 00:12:07.720 --> 00:12:10.879 success. It's about what's actually happening 00:12:10.879 --> 00:12:13.340 three months down the line when the spotlight's 00:12:13.340 --> 00:12:16.580 off and the gummy bears are long gone. Exactly. 00:12:17.039 --> 00:12:18.419 Which leaves us with the final thought for you 00:12:18.419 --> 00:12:21.059 to maybe mull over. Based on that observation, 00:12:21.259 --> 00:12:23.500 we discussed that sometimes management might 00:12:23.500 --> 00:12:25.620 not want to hear the bad news from a rigorous 00:12:25.620 --> 00:12:28.519 check. Right. What should you prioritize in your 00:12:28.519 --> 00:12:31.059 work? The immediate appearance of success may 00:12:31.059 --> 00:12:34.340 be closing the budget on time or the tougher, 00:12:34.639 --> 00:12:37.360 potentially unpopular, but sustained truth that 00:12:37.360 --> 00:12:39.460 only comes from really following through with 00:12:39.460 --> 00:12:42.340 check and act. Something to think about. That's 00:12:42.340 --> 00:12:44.240 our deep dive for today. We'll catch you on the 00:12:44.240 --> 00:12:44.600 next one.