Failure to Launch. Quality or Quantity. Chicken or Egg.

Matthew McConaughey or Sarah Jessica Parker? Got you interested didn’t I?

Anyone seen the movie - Failure to Launch with this pair? Worth an entertaining watch. I have three teenagers in various states of launch around my home. Now you’re going to say, Venki this is LinkedIn! Why are we discussing movies here? So lets just discuss failure to launch of a different sort relevant to this space then. I have had a rather persistent and morally earth-shattering dilemma that has plagued my entire career. Not sure if I will ever fully resolve it. But here to share. As always, to cause you to think. And choose wisely.

So the central question is what comes first Quality or Throughput? I know, I know. The answer is obvious, and dare we waste more time on this page listening to some gyan. Hang with me for just a few. And let me pose the same questions with just a wee bit of context. And see if you’ve been in such a predicament.

You’re in the early stages of a launch and instead of defects per 100 or defects per thousand, you’re looking at double digit defects per product! Per product. Would have been defensible if you were at some prototype stage. Or the very first batch. But you’ve been at it now for well over 6 months, and you’re not even on the quality launch curve. Can’t get on it unless someone changes the scale for you. Got your attention, didn’t I? What do you do? 

Complain loudly about the unreasonable quality standards that sales signed us on to with no clue about our manufacturing capabilities? Or if they wanted “this” level of quality we should have gotten double the price, then get better machines and more people up front to get the job done? Sweet talk your customer into relaxing the quality standard for a few weeks, while you show him presentations on what you plan to do with the repetitive defects? Ask for a modified launch curve as the customer faces multiple requests for the same draconian delays across the supplier chain? Or increase your output with expensive armies checking quality so you have large quality losses but the customer quality is back on the curve? So yeah, now you remember. Do I make my count for the shift? Or do I stay on the launch quality curve with significantly diminished output. You have to pick, either the launch volume curve or the launch quality curve. Come on now. Think. Take a position.

Here is my position. Learned in blood, sweat and tears. Sharing with you in the hope of sparing if only one of you - the agony wrought by getting this wrong. Perhaps you, unlike me, can have a full head of hair at 55!

Let me first offer a simple timescale: Pre SOP, SOP and Post SOP. For the uninitiated SOP stands for Start of Production. This is when sellable parts are being made and finance is holding its breath hoping for us to get on the revenue curve. Oh yeah, the revenue and profitability curves!! Hold that thought, that’s definitely the topic of the next blog. Depending on the complexity of the product take it in one month, 3 month, 6 month or 1 year increments. For example lets say SOP is on January 1st 2020. Take Pre SOP from Oct-Dec 2019, SOP from Jan-Mar 2020 and Post SOP from Apr-Jun 2020. For simplicity’s sake and to draw a telling analogy, let us call each of these stages Stretch, Jog and Run. What you do in each stage is what I am about to prescribe. Ignore at your own risk.

Stretch: 

This is the most important stage in a launch. If the right steps and resources are not brought to bear here, the next two stages will be sheer agony. This is the stage where gross product and process problems have to be discovered and corrected conclusively. If it doesn’t get fixed here, its likely the most logical time to ask for a delay of SOP. 

Job number one. Get your people trained. And trained. And trained. On what the company means. What the launch means? And what does success mean. And what is in it for them.

This implies that enormous capital has to be expended right here where the process is put to the test with prototype parts to discover encountered issues: in part, process, equipment, facility, IT, packaging, material storage, material movement, the list goes on. The operative mode here is to quickly get into a “production” mind set with a clear daily production schedule. Not a schedule where you make several hundred parts in a shift. But maybe two. With very detailed observations, measurements and data gathering by trained team members clear about the overall goals of the launch. In most cases, spend half the shift simulating, then the other half getting cross functional teams to completely eliminate the root cause. Here is where you go from maybe a hundred defects per product to one defect per product. 

You might recall that in an earlier blog on problem solving I spoke of problem containment before correction. This is the one time where containment is suicidal. Only correction. The error, I have found in many cases is that companies budget inappropriately for this stage (low revenue, so low costs). They contain. They say, when we get more mature parts we will automatically get better. When the team members get more experience we will get better. This is where the lion’s share of company resources both corporate and local need to be deployed in recording and truly eliminating root causes of defects. Typical downfalls are when equipment gets installed too late to be tested or prototype parts are so primitive or too expensive to have enough of them to go around to run actual simulations. I have to say that I have seen the whole gamut of preparation across various products and industries. 

Honestly, you need to get into this stretch phase with a mature product and settling process. If the product is not mature yet, you’re dead. If the process is not fully functional yet, you’re double dead. Do not cross go. Stay in jail. Show me a botched launch, and more often than not, they blow it in this “Stretch” phase. A dollar saved here, for something absolutely necessary to discover and fix problems, is worth at least 10 dollars in containment cost in the subsequent stages. Pay me now, or pay me later. A lot more.

Jog: 

Ideally when you get into SOP, you should be entering with a handful (say 10-12) of gnarly and persistent problems from the product or process that simply cannot be fully contained or corrected. You simply ran out of time.

Round about now, you are going to encounter high levels of scrap as draconian quality gates will need to be in place to protect the customer. You have to budget the costs for these losses and invest in both quality resources as well as scrap losses. More importantly, your yield has to drive what you actually produce, and are capable of producing. This means that by this time your process capability has to be much further along. Like 6 months ahead in volume capability so you can still net enough good parts. 

Not only that, there has to be a razor sharp attention on the defects being found and walked back from the end to the beginning. As well as the causes for scrap starting at the customer walking all the way back to the beginning. Applying the same rigor of data collection, and correction in the first phase. 

Through containment, and only having a finite set of problems to solve, you are now able to mount and stay on the launch quality curve. And mostly meet the launch volume curve. Only, the extent of resources required to inspect, tabulate, solve and budget required to throw away bad parts needs to be allocated well in advance. The problem that I have seen often is that by the time the realisation comes for the need for resources in this stage, there are such hellfires going on that the same thin resources running the show try and solve these issues also in parallel. Wrong. The good news is that as problems truly get solved, the need for resources drops with it. Revenue starts climbing and costs start assuming normal spend rates.

Run

You think you are cruising along. With even a couple of weeks of making quality and volume targets. And bam! And Kabaam! Customer went up a sharp increase in volume. Didn’t simulate it or plan for it. Or a completely unexpected part shortage or material defect hits you in the solar plexus. 

This is again when a storm-trooper team has to be on standby to grab the situation head-on, protect the customer, fly parts on planes across the globe, rework parts through the night, machine a broken part from raw metal or whatever it takes to restore the line to launch performance. In my humble opinion, this sort of stormtrooper can only come from folks that have done this again and again for your company. This cannot be handled by rookies. This is where you have to station the big guns that go from launch to launch, alert and able to dive into action and results. 

All along, you have to bring your customer along also, transparently sharing your issues, containments, and corrections. The more transparent you are, and the early alarm you are able to pull, the more cooperative the customer is likely to be to help you solve and overcome the expected and unexpected surprises. 

So there, I hope I have explained where my learning on this dilemma is. Quality comes first. And you owe your customer the volume you promised him too. So crank up your repetitions well in advance so you get to protect quality without having to compromise quantity. 

And oh! Shall we say, that when someone mentions curves around me. I lose colour. Appetite. And Hair.