What Software Makers Can Learn From Adidas

Software makers can learn a lot from companies like Adidas.

In this episode of the Fortune's Path podcast, Tom talks with quality and compliance expert Norbert Teston, formerly Director of Compliance Excellence at Adidas, the international athletic wear behemoth. He's now President of NT Creative and Consulting, where he helps COOs and Heads of Operations eliminate the fear and risk of product recalls.

Learn:

• How this French native started his career in Vietnam

• What triggered his start in the footwear business

• What it's like to be in a factory with 20,000 people

• The nuances of creating a company culture of quality excellence

But most important, how the quality processes in manufacturing might be applied to software development. Turns out, there’s a lot SaaS companies can learn from how to make a great shoe with minimal defects.

Transcript

Tom Noser (00:00):
What can software makers learn from Adidas? How does quality control work in the shoe business? What's a recall for a physical product like? These are some of the questions I ask quality and compliance expert Norbert Teston, former director of compliance excellence for Adidas, the international athletic wear behemoth. Norbert is now the president and principal consultant of NT Creative and Consulting, working with COOs and heads of operations to eliminate the fear and risk of product recalls. Norbert talks about how he got to Vietnam, how he started in the footwear business, what it's like to be in a factory with 20,000 people, and how you create a culture of quality excellence on this episode of the Fortune's Path Podcast. Norbert, it's so good to see you. Thank you so much for joining me today.

Norbert Teston (00:56):

No, thank you. Thank you. It's good to be here.

Tom Noser (00:59):

I have no experience in manufacturing and I know you have a lot, and I'm very interested in having a conversation about the overlap between quality processes in manufacturing and how those might be applied in software. So to get started, I want you to tell me a little bit about your history at Adidas and what you did there and how your career went.

Norbert Teston (01:28):
I actually started for the company back in '97 in the sourcing office over in Vietnam. So I was responsible back then to set up the testing laboratories over in that sourcing market. They were building up the sourcing organization over in Vietnam. And so I was responsible for building all of the lab facilities in the office, but also in the factories. So that's what I did for a few years. And then I moved on from there, went into the headquarter, was responsible for standards, responsible for test methods and bigger plans as well that we had on the QA side. And then got involved eventually into product safety, compliance of products, larger programs, market compliance that we had with China, for example. Up to the last five years where then my team and I were working on the entire framework of quality and compliance for the organization.

Tom Noser (02:35):
First of all, there's a lot in there. That's obviously a lot of-

Norbert Teston (02:38):
There is.

Tom Noser (02:40):
Yeah. Different stuff. So let me dig into the work you did in Vietnam. You said that you were in a laboratory, so I'm assuming ... Are the shoes being assembled in Vietnam?

Norbert Teston (02:48):
That's correct.

Norbert Teston (02:49):
Most of your raw material comes from the area. Vietnam. Some of the leathers would come from South America. Argentina was a big producer for example. But a lot of the local materials like the synthetics, all of the outsole and midsole material, the rubbers, the polymers, that was all coming locally either from Vietnam or China, Taiwan and Korea. Those were the big producing countries. And then the assembly itself was done in Vietnam.

Tom Noser (03:20):
When you get a ... I'm going to call it a part. That may not be the right term. But let's say you get a sole. Is your lab testing the quality of that sole or the material? What is happening in that lab?

Norbert Teston (03:33):
Correct. For every new material supplier that would be onboarded and considered for use, we would do a bunch of approval tests just to release that materials. But then at the material level, we would take every type of material, run it through the lab. We had a set of standards that we wanted to follow internally, so all these materials would be assessed and then officially released in the system for use by the developers in the factories. So yeah, everything would be pre-approved during the creation process, but also during production they would have inbound checks at the warehouse. At the factory warehouse itself.

Tom Noser (04:19):
So when I'm designing a shoe, can I blue sky it and imagine a material exists and then like, okay, I'm going to write a ... I don't know if a spec is the right term, but I'll write a spec for what this is supposed to do, and then it's up to procurement to find the material.

Norbert Teston (04:35):
We had different functions within the organization so we had an entire designer group. So they would come up with a ... Well, actually it started even before that. There was a business unit that was determining the need and the briefs. So they would describe what they were looking for for the next range, so the next season. So they would work season by season. We used to have four seasons a year, and then they switched to two seasons after a while. So the business unit would determine the need and put a brief together, and then the design would come in and start drafting ideas for designs and types of materials and textures and colors and things like that. And then your development team would go in and start looking at, okay, what are the materials we have in our library? Because they had it all predefined. What suppliers? Do I need to look for new suppliers for those types of materials? And then they would start the sampling rounds. So based on this, they would work with the factories over in Asia to start putting a couple of samples together. Then it became a cycle afterwards where they would have different rounds of reviews of those samples until it gets confirmed for production and for the seasons basically.

Tom Noser (05:56):
So from start to finish, how long does it take to make a shoe?

Norbert Teston (06:01):
The development phase is the longest one, and that's what companies have been chasing after the most. It could be 12 months to 16 months, and then if you have innovation, it's working like five years ahead of what's going to get in the market. But then in terms of the manufacturing, once everything is defined, your specs are all in order, you know exactly what you're going to do just to assemble the shoe, that could be done in a few days, but there's between 250 and 350 manual operations that need to happen to put one pair of footwear together. And we're talking sportswear, which tends to be a little more complex because there are more pieces to assemble, but it's a long process.

Tom Noser (06:46):
So from concept, you said there's ... I'm going to call it a marketing group, which may not be right, but there's a group that determines what we want to do. And I'm assuming they're not drawing anything, they're not making pictures. They're saying the Olympics are coming up, or we have an athlete who we're launching a footwear line with, and so we think the market is interested in these kinds of things. That goes to a designer, blah, blah, blah. So from that process to the shoe goes in the box is close to two years.

Norbert Teston (07:21):
If you go look at it, 16 months, two years. Yeah. There are companies right now that are trying to really turn this around and say six or seven months. But for large organizations where the range is really complex, you have thousands of different styles of products, it's much heavier, so it takes a little longer.

Tom Noser (07:46):
You mentioned a materials library. If I'm a designer and I'm limited to only using things in the preexisting materials library, I'm only using materials that have been sourced and approved, does that cut down the time at all or not really?

Norbert Teston (08:05):
It tends to do. The challenge that organizations have been facing and Adidas was facing is that those material libraries tended to be very complex and there was just a lot of entries. So it was a constant work to try to clean it up and consolidate because very often ... When you reach a certain size, especially, and you're making footwear products like this, the business units don't necessarily talk to each other in terms of what they want to use in terms of materials so you end up with tons of duplicates or materials that are looking very similar but are sourced in different areas in different suppliers. And so the materials team was constantly trying to maintain that library of materials to make it as effective as possible. And the idea is if you consolidate all of that upfront, then obviously you're going to be able to negotiate your prices down the road with those suppliers. You have key suppliers. There's more of a strategy that can be established. But that's always been a clear challenge. And from a quality side for us, for my group, it was also very important because it's easier to manage 500 suppliers rather than 6,000.

Tom Noser (09:21):
I'm really interested in this, so I'm going to just stop for a moment. There's some analogies related to software. So very large software system has multiple ... They're a little bit like suppliers and producers because writing code all over that complex system and there are libraries, but it's possible that essentially the same thing gets written more than once or something very similar gets written more than once and in two different ways by two different teams. And there's debate I think within software about whether that's good or bad because the speed to delivery in software is so important. We're constantly updating. And a lot of the times software systems will sacrifice quality for speed. Like, well, we'll fix it later. I feel like in manufacturing you don't really have the opportunity to say, if we get it wrong, we'll come back and fix it later. Is that accurate?

Norbert Teston (10:27):
That is pretty accurate. Once your product is made, it's made. It's in the boat. And your only opportunity ... You might be able to repair some of it, but it's very limited cases where you can actually get into the product. The times when you have to repair it, the cost of the repair is ginormous. Simply because you have to go in the container itself, open all the shoe boxes, unpack everything, set up a line to clean or do whatever is needed to be repaired and then repack everything. And so just the sheer amount of administration or small manipulation that needs to be done is hyper expensive. So there's a big philosophy around right first time in that industry as well. Trying to educate all these people that it's better to spend one more second to actually do the right thing rather than ship it down the line as quickly as possible and having to repair and correct everything after that.

Tom Noser (11:36):
Talk to me a little bit about the establishment of a quality process. It sounds like ... Anyway, I'll just stop there. Tell me a little bit about how you establish a quality process.

Norbert Teston (11:52):
The first thing is probably looking at what you as a company, what you're shooting for in terms of vision, mission, purpose. What do you want to stand for? Because every standard that you're going to set up in the organization later on needs to service that mission and vision and purpose. Because you want your product to do certain things towards your consumer. You want your product to be fulfilling that consumer expectation. So that's a very big starting point that companies tend to not look at. So that gives you your first step and then the idea then is to look at your entire process and to say, where are the toll gates that I need to check throughout that process to make sure that I keep it as consistent and sustainable as possible? So you want eliminate variation. That's the idea. You want to create a process, whether it's during the development phase or during your production phase, your manufacturing phase, you want to have as small of variations as possible. So it could be if you're manufacturing process 10 steps. Okay, great. That's fairly easy. But things can go wrong in any of those 10 steps. So the question is understanding where all those variation potentially happening and then putting systems in place to minimize the variation. So that that's the basic foundation of it if you want.

Tom Noser (13:27):
That's very helpful. You mentioned at the beginning that an organization often skips the connection between quality and mission. Why is that?

Norbert Teston (13:41):
I think because very often quality and compliance as product are just afterthoughts. People tend to think in terms of, oh, I want to create this product, I want to make this, I want to sell that. But the question is every one of those product is a reflection of who you are as an organization and what you want to portray, and you need to make sure that you define those functions of that product and you ensure that they're going to consistently be there. And the only way to do that is to have a strong quality system in place in your organization.

Tom Noser (14:19):
So in manufacturing, do people check their own work or do you tend to have somebody else check what somebody did?

Norbert Teston (14:27):
It's a bit of a combination. In the apparel and footwear industry primarily, there's also a lot of training that happens around self awareness of quality systems. So making sure that as a person on the line who does this little operation within these 250 steps, you're aware of what's acceptable and not acceptable. But then also there's usually a supervisor or some sort of audit that happens within the organization. And then sometimes there are layers of audits above this as well with external organization that may come in and audit your work. And then there's steps at the end. So for example, you may have during the production line stitching operations that gets checked by the employee, then the line supervisor may do some checks on it, then at the end of the line you may have another check, and then you have another check at final inspection. So there could be some redundancy. The objective here is to try to make that process as light as possible and to empower the employees who are actually doing the work to make sure that they control and they do things right the first time.

Tom Noser (15:57):
So if I understand that right, it's like the culture is built around the person being responsible for their own work, but there are checks along the line. Not just one, but maybe five or six checks along the line. And then if a defect is discovered, who's responsible for fixing it?

Norbert Teston (16:15):
It would go back to ... There are processes. So if you actually walk down a production line ... So just imagine being in one of those factories and they're as big as towns. Some of those factors are 20,000 people. So you walk down those lines that are gigantic. Your main line runs your production and then coming out of the line, sometimes product is being taken out and then sent back to be repaired or reprocessed for specific operations because a defect has been found or there's a little stain or the stitch is not right, or it's just taken completely out of the production because it cannot be repaired as well. So there's a rework loop that happens along each of the production and that's what actually triggers your cost as well so that's what you really want to minimize, because visually you can see what's the cost of your employees and your teams to not do a quality product right from the first time.

Tom Noser (17:20):
I'm going to go back to software for a second. I want to give you a chance to think while I talk. I want to hear a story about what it's like being inside a factory with 20,000 people. That sounds just absolutely unworldly. And so in software, you can't really see what somebody else working on the system is doing. You accept the work that they do, it comes back to you as a chunk of code, and then you plug it into what you're doing. Or you don't even plug it in, you just access it remotely. And so the idea of, I got a shoe, I can look at a shoe and I can see do the eyelets align? And it's pretty-

Norbert Teston (18:07):
Right. It's a lot of visual checks.

Tom Noser (18:08):
A lot of visual checks. There are none in software. It is like, oh, it works on my box. There's the typical joke about, well, it worked for me when I ran it. I don't know why it doesn't work for you. So it feels like there are challenges of finding the invisible problems. I'm sure there are invisible problems in the manufacturing process too. How did you guys find those?

Norbert Teston (18:28):
There are. So if you think about sizing for example, so in footwear, this is something that you can't see along the line. If there's a small deviation in your pattern making, in your stitching, in your assembly, et cetera, those can all accumulate and then create a problem from a fitting perspective. Where your volume in the toolbox becomes suddenly a lot smaller, your materials have been rotating, you don't have the stretch anymore. There's tons of reasons that can accumulate and create a fit issue. And you're not necessarily going to see this until the product's on the market, unless you have a quick fit check that's being done before the product actually leaves the factory. So there are some invisible problems. What I recommend in those situations is you've got to build a learning process over ...

Norbert Teston (19:52):
What I would recommend organizations is to implement some sort of feedback loop on a regular basis. Any experience that you may have going forward where there's been a defect that hasn't been perceived during the manufacturing phase, then think of what can I put in place early enough that would catch that. And so we went through so many iterations with Adidas over the 25 years that I was there, that every time there was a major issue, we would think in terms of, okay, how do we set up something throughout the entire process that would help us at least highlight the potentiality of an issue and then take care of it in the right time so we don't have to end up in the same situation.

Tom Noser (20:45):
That's really interesting. So if I understand that right, it's a little bit like a risk management process in manufacturing. Is that reasonable? To think about we're trying to identify what all the possible risks are and what their likelihood of them happening?

Norbert Teston (20:58):
Correct. Yeah.

Tom Noser (21:00):
How do you do that? That feels almost like a philosophical exercise of trying to understand what are all the possible things that could go wrong.

Norbert Teston (21:10):
I think a lot of it is experience. It's interesting. My boss was a expert in shoe manufacturing. They were shoe dogs. That's what they were called. But an entire generation of people who had been trained to be extremely precise about shoe manufacturing. And then we had pattern experts, we had mold experts. It's a whole list of expertise internally. And so a lot of the work was about trying to understand the root cause of all of these situations because you see the problem at the end and it can come from a whole bunch of different root reasons. So doing a root cause analysis, asking why, reaching out to different people and evaluating scenarios. And then the teams would start to get really good at doing full investigations of those situations. And once we knew exactly where that root cause is, then we would go back and try to set up the system to make sure that we would catch that going forward.

Tom Noser (22:25):
So that feels like it requires a culture that's willing to stop. To say, we've discovered a significant flaw in our manufacturing process, we have to stop manufacturing until we get this fixed. Could Adidas ever do that? I would think the sky would fall if somebody said, we need to turn off the line.

Norbert Teston (22:47):
Believe it or not, the production managers and senior managers ... We had Adidas people and Nike has the same, all these brands tend to have the same. They have dedicated Adidas responsible personnel that is on the line in the factory and is responsible for managing production. And one of their authorities was to stop the line. They were able to say, "No, no, no, no. What's going through right now is not acceptable. We need to stop the line, fix the problem, and rerun it." So there was a first round of that catching of issues that would happen right there on the line. And I would not see it from a headquarter perspective because they'd be caught so early in the process. But that authority existed. Yes. Definitely. It's a scary thing to do because the factories were all ... They were not Adidas owned factory, they were subcontractors. And so they would flip out every time something like this was happening because there's so much money that goes out the window when you have to hold the line up.

Tom Noser (23:54):
You got a bunch of hourly employees standing around. Oh, gosh.

Norbert Teston (23:57):
Yes, absolutely.

Tom Noser (24:01):
When you got to Adidas, and between the 25 years that you were there, did you see their culture of quality change at all or did you come into a situation where quality was very important?

Norbert Teston (24:18):
The size of the organization just changed so much that we ... It's the approach around quality that was different in a way that when I started, it was more related to the individual, the idea of craftsmanship. It was a great inheritance from the founder himself, Adolf Dassler, that has carried itself over years after year. I think the more the company started to grow, the more we had to make quality an institution within the organization. We had quality values that were defined and quality was always on the board of high level projects as well for the organization. So it shifted in terms of that. It wasn't necessarily considered as a craftsmanship concept, but it was still very present. But it had to be present in a very different way just because the organization was a complete different size.

Norbert Teston (25:37):
Does that make sense?

Tom Noser (25:39):
It makes total sense. So it also sounds like when Adolf started the business, there weren't that many people. He was sourcing and he was manufacturing in a limited area. And then now you've got factories in Vietnam and factories in South America and suddenly it's a little hard to have the exact same culture in both places.

Norbert Teston (25:59):
Yeah.

Tom Noser (25:59):
Talk to me a little bit about those challenges of international business. If you were advising a company that was thinking about expanding into another country. I know that's not exactly your area of expertise, but based upon your experience, what would you tell them?

Norbert Teston (26:19):
I'll tell them, get your butt over there. That'd be my first recommendation is like, don't ever think that you're never going to have to fly over to see and interact with people. I was lucky enough to ... I was actually already in Vietnam when they were looking for people and hiring people on site and so I got hired directly from Vietnam. But what I understood very quickly is that I had to show my face. There's something very human and very powerful about connecting with people face-to-face and being there to tackle conversations in difficult situations as well. And you cannot do proper sourcing if you don't know who you're dealing with. And there's an extremely high risk of having things get completely wrong. And I actually was on a conversation two weeks ago with a lady and she was working with a small brand and they had done their sourcing, they haven't checked their POs before it shipped out. Their partner wasn't very solid and they ended up with all of their product being lower quality. And so now they have tons of returns and there's additional cost and it might actually sink the brand. So you cannot underestimate the power of just going physically there, staying for a couple of weeks to really understand who your partners are, and then also making sure that whatever comes off the line is what you expect it to be.

(27:58)
Because culturally where is such a gap, especially with Asia, of understanding of what products quality is supposed to be, and even understanding your own philosophy of product quality. What do you expect them to be following? And those can all be clarified once you have face-to-face conversations. And so that's what I was trying to do. When I was in Vietnam the very first years I would be out. They would send us to the factory all the time. We would get out to talk to the factories, be there with them, help them with their machine. So on the lab side, I would go and work with the lab managers of each of those factories, help them with the test methods, the machinery, et cetera, talk to their management, go through the line, see what's happening on the line just to get familiar and then they get familiar with us as well. Because one of the things you absolutely want is if there is an issue, you want them to call you. You want them to be comfortable enough with you that they'll pick up the phone and say, "We're having an issue here. We need to talk to you."

Tom Noser (29:09):
You have to build equity into that relationship. You're going to have a hard conversation at some point. There's just too many variables for everything to go perfectly the first time. And you need a relationship for them to be able to do that. So I have worked with offshore development teams, writing software offshore. And one of the things that was really difficult about working with those groups is you never had ... Or at least I'll speak for myself. I never had any idea what they were doing. I knew what I had assigned them, I knew what we asked them to do, but it was very hard to tell what they were actually doing. And I always had to take the word of the onshore developers about what the offshore developers were doing. And when something went wrong, whose fault was it, how did it happen, et cetera. And it was always the offshore guys. They were always dummies.

Norbert Teston (29:59):
Of course.

Tom Noser (30:02):
I feel like in manufacturing there's opportunity for the same he said, she said situation. As somebody in quality, how would you find the truth?

Norbert Teston (30:14):
Sometimes you don't find it. Sometimes you suspect it. Sometimes you have to close a chapter being like, "Okay. Well 50/50 this time, but we'll keep an eye on you." That happened as well, unfortunately. Again, I think it went back ... There's generally a pattern. That's also something that's important. Your best partners in sourcing are really ... Again, they'll have a different attitude around the problem. They'll reach out to you early enough. You'll have the opportunity to gather information and really understand what is going on, what do they know, what do they not know. The worst business, the worst sourcing partners, I would say, they'll try to keep stuff away from you. They'll try to hide things away. Again, you'll see it as a pattern. It's going to come up again and again and again.

(31:23)
Adidas has a clear management of KPIs for their suppliers as well, just to understand who's repetitively having issues that we are not able to solve, and where do we need to focus our attention a little bit more, either to strengthen the relationship or to address it in a way that maybe that relationship is not going to last very long because we can't really rely on each other. We don't seem to align and communicate the right way.

Tom Noser (31:56):
And so those KPIs, are they similar across all partners? Is there a core set across all ... I guess you couldn't do all partners but across partners in a similar function?

Norbert Teston (32:08):
Yes. Used to be very similar. The main KPIs that the company or the footwear based industry is looking at is delivery, quality support with innovation, investments, volumes and things like that. But it's a lot of time, cost and quality. A lot of it revolves around those three pillars.

Tom Noser (32:36):
When you're measuring quality, how do you do that in a way that avoids opinion?

Norbert Teston (32:47):
A lot of it is gathering products. That's the good thing about working on consumer goods as well is you can actually gather a product and clearly show yes or no, this is a problem, and then start documenting it so then it becomes more of an educational tool. So we would have ... And that's also something I recommend companies to start with is what are the problems you're facing with your consumers right now? What are the problematic products that you've sent to the market? And then start a root cause analysis of these so that you can understand where that problem is coming from.

Tom Noser (33:30):
Did Adidas, in your time there, ever have a product that went out and then you started to get complaints from consumers about something pretty bad was happening?

Norbert Teston (33:41):
Yes.

Tom Noser (33:41):
How'd you handle it?

Tell me that story if you can from start to finish.

Norbert Teston (33:48):
These were not the fun times for sure. Because it's an escalation. It's a bit of a trauma every time and you're like, oh Lord, I hope this is not a recall because this is going to hurt. And of course management just, again, it just runs around. And very often what we were trying to do is to tame down the people first to be like, let's not panic. We don't know yet what's going on. Something came back to us through a certain source. We know there's a problem. Let us investigate. Is this a serious issue? How many products have been returned for this reasons? What kind of reasons are there? Is there a product safety concern related to it? Are we eventually considering a recall? Let's gather our purchase order list. How many have we produced? Where have they been produced? What factories? Do we have? Any former examples that relate to the same type of issue?

(34:53)
So there was a whole investigation phase really that would allow us to evaluate the level of issue that we were dealing with. The lowest would be, okay, this is something easy. Maybe the consumer's not ... It won't be perceived by the consumer, it doesn't have an impact. We can just let it go and fix the production going forward to no, this is a major problem. There's potentially a safety issue for the person wearing it. We need to recall that product completely so that means pull it back from all the shelves around the entire world. And that has happened too. And yes, those would be generally a month or so of just focusing on this. Yes.

Tom Noser (35:39):
I'm assuming, God willing, that happens in one type one shoe, not in a line of shoes. It happens in one shoe. And so manufacturing goes on around the rest of the organization for everything else. If you have a recall in a shoe, it doesn't mean you stop making sweatpants.

Norbert Teston (35:57):
That's correct.

Tom Noser (35:59):
But that particular area is totally focused on pulling it back and-

Norbert Teston (36:08):
So the footwear and apparel industry, your runs are fairly short. You're not making a car that's going to run on your production line for two or three years. There were a couple of products that were ... Like all the original stuff, like products that were here for a long time, those would be running for years at a time and they're the evergreen type of product. But most of the range was actually produced for a specific season for a campaign. There was color related to it, et cetera. So generally, by the time you actually deal with the issue, the production is done. Your purchase orders have all been fulfilled and the product is either on the boat or it's already in the market. Unless it was a brand new technology, for example. Something that had a specific function that needed to be fulfilled and that function was failing or the technology itself was failing, and then you knew that you were in big trouble because that means there's more products that are waiting to be produced and those need to be fixed as well.

Tom Noser (37:13):
Because they're not using that technology just in one product, in one shoe.

Norbert Teston (37:16):
Correct. So everything that would go across the range, those would be the major very worrying issues.

Tom Noser (37:24):
I can see how that would be scary. Do you do a limited run of shoes just to make sure that it's ... Let's just manufacture five shoes or whatever. One in every size to make sure that that they're okay? Or do you just hit the button and you're like, nope, a hundred thousand shoes coming down the line.

Norbert Teston (37:45):
It depends on the philosophy of the organization at that point and the group that was working on it. I've seen both happening. My preference and inclination and the recommendation was always to say, if this is something that's brand new, make a short run. Try it out. Because as soon as you scale it up in manufacturing, you're going to start facing other issues that you didn't know about. And so there's the idea of, yeah, you need to ramp up your manufacturing and make sure it's ready to support that type of volume and then understand where the variabilities might be during that scaling up of manufacturing. Volumes vary quite a bit actually between products as well.

Tom Noser (38:35):
When you're advising clients, do you typically tell them, do a small run. Just make it part of your standard processes. You always do a small run because there's going to be stuff that's different between that and the prototype that you can't tell until you're doing a run.

Norbert Teston (38:52):
That's correct. Especially everything that's innovation related. Back about 15 years ago, I worked on a product that was electronic based. So there was a little battery, it was called the Adidas One. It was a little engine inside the sole that was pulling in elastomer components and making the sole harder or softer and it would adjust itself based on your running style as well. And so you could make it harder or softer, but then it would by itself adjust itself. And so there we go. We've been making regular footwear ... Regular footwear. Sportswear for the past 15, 20 years maybe. And then suddenly we're introducing electronics inside a piece of footwear that's going to be worn out in the rain, the cold, the heat, et cetera and with no knowledge whatsoever internally of how do we handle electronics. So I actually flew to China and spent two weeks over there to break down every single process of that electronic components to make sure we had everything upfront checked so that when production and mass manufacturing comes in, then we'd be able to at least ensure that whatever we're getting is consistent with what we had taken during development and run all the tests over. And then we introduced an additional set of testing, et cetera.

(40:24)
But the complexity of just putting this together, it was a first for all of us. And actually to set up a new tool that we then started using for new innovations that were coming up afterwards.

Tom Noser (40:38):
Was the tool a process tool? Tell me a little bit about the tool.

Norbert Teston (40:41):
It was a process tool. A lot of it was just Excel files that we had put together with a specific format to really align the organization as well. Make sure that ... Okay, define for ourself what's an acceptable ... If we get a small component that goes into that big mechanics, what are the approved and defective measurements that we're going to do on that component? How do we say it's okay to go in or it's not okay to go in? And so defining all of this for each of those pieces and then providing that to the inbound inspection teams as well.

Tom Noser (41:22):
So electronics in a shoe. As someone who knows nothing about shoe manufacturing, that seems like a really bad idea. I'm sorry.

Norbert Teston (41:32):
Thank you. I appreciate it.

Tom Noser (41:36):
I'm sorry to rain on that innovation, but how did it turn out?

Norbert Teston (41:41):
There was two generations of it, I believe. Two or three generations that happened. The problem with mechanics is that they abrade and they don't last forever. That's the challenge is as soon as you have mechanical part acting against each other, even if it's metallic, it's going to start abrading. You have corrosion issues. So the concept was probably not necessarily a bad idea. I think what we missed afterwards is the patient, the time and the investment to really move it towards something that's a lot more sustainable. Like the electric car industry, it's like you start where you are. You run at a loss for a while and then until you actually push it to a point where, okay, now it's becoming something very solid. And I don't think we got to that point unfortunately with that technology.

Tom Noser (42:42):
So when you guys would investigate a market, someone's thinking, we need a shoe that adjusts to the running style of the runner. We think that'll be a breakthrough technology for us. Do you have information that says how people decide what sorts of shoes to wear? I would think so much of that would just be like, oh, these look cool. I like the people who wear Adidas. It feels like it's not a rational decision.

Norbert Teston (43:08):
Part of it was every business unit, so every marketing group was organized by sports category. So you had a soccer group and you had an indoor group, you had an outdoor group. And so those would be sports specific and very focused on understanding who that consumer base is to understand their needs as well. What type of product do we want to create to help that athlete achieve better performance? So the running group would have connections with running professionals and athletes and college athletes. And they would be very connected to that community to a point where they would understand much better what type of product people were looking for. We had a year actually ... It was a very interesting experience. But back in 2008 I believe we got approached by our development team and they're like, we are making racing flats. And racing flats are basically running shoes that you want to run one or two marathons with. They're super light, so they feel like you have nothing on your feet. But because of the fact that you're taking off the weight, you're also taking out the durability of the product. So after two races, that product is good to throw away.

(44:38)
And this was going against all the standards that we had established because we always wanted to build shoes that would last forever. And so we had to have those very interesting conversations to say, well, what's your client base? What is it marketed as? What kind of standard are we really trying to build here for that specific product? How do we isolate that product in the range so that suddenly not everybody starts using the same constructions and then retail as an actual running shoe that's going to last forever. So there's a lot of prevention that needed to happen when we worked on that product, but it turned out great. There's a great team collaboration between the designer team, the development team, the marketing team, us as a quality team to really try to understand that niche market and really create a product that would work for it knowing that there are some downfalls, but those downfalls were also being communicated properly to the consumer.

Tom Noser (45:41):
Love that. To me, what's so fascinating about that story is that it goes all the way back up to what you said at the beginning of the discussion about values. And so the QA manufacturing process, the quality process to support that shoe had to be different than the quality process to support other shoes because that shoe had a different mission, which was, it was a disposable shoe. But because customers said, we want a disposable shoe, we like a shoe that feels great for two races, and then we want that new shoe feeling all the time. And that runs counter. If you'd said that to Adolf on his bench making a shoe, he'd have been like, you're an idiot. So I think it causes you to reconsider your mission.

Norbert Teston (46:35):
Yeah. So this was really interesting because actually Adolf on his bench, because he was so much in control of everything ... Because he was the one talking to the athlete. He knew how to put shoes together. He was the one in contact with his production and he was the one back with the athletes when the shoe was being worn and tested. And so you have tons of picture of out of Adolf Dassler sitting together with athletes and really trying to understand what does that athlete need, how do we build it? And then once he had the prototype, trying it out with them to really confirm that the product was doing what it was supposed to do. And that type of loop, that really end to end understanding of the process and the product and integrating the customer into that loop is critical I think.

Tom Noser (47:28):
I love that. I had made an assumption, which just turns out it was false, is that he didn't start out in the athletic shoe business, but it sounds like he did.

Norbert Teston (47:38):
He did. Yes. He was an athlete actually, all around athlete. He was racing, he was doing high jumps, he was playing soccer. He was doing it all and fascinated by sports. And so he started developing products for all these categories of sports as well, and then went into Olympics and ... Yeah.

Tom Noser (48:00):
Wow. So that story about the disposable track shoe or marathon shoe goes back to the very origins of the company, about that collaboration between the designer and the athlete very closely. The only thing right now ... Back then it was just the organization was Adolf. But right now there's a gigantic multinational in between the athlete and the manufacturing process that you in that case had to go around because I'm assuming there were certain beliefs about what shoes were supposed to be that were pretty ingrained into the organization. And so I love what you talked about, how you kept them separate. Talk to me a little bit about what are those ... So this is another software problem of how do you manage risk with new development, new releases? How do you manage risk as it relates to the data that you're collecting and maintaining privacy of that data and maintaining the good separation of that data, et cetera. So the idea of, okay, for this purpose we need a totally separate manufacturing process, talk to me a little bit about how that happened.

Norbert Teston (49:12):
We would really try to look at it in the end to end, meaning we would start looking at how do we control the supplier, the raw material as it's being produced? So who are those suppliers? Do we need to talk to them? Do we need to set very specific guidelines for them? And then we would look at the manufacturer, the assembly piece of it. So with our tier ones, what we call our tier one factories, so the manufacturers themselves and do the same exercise. Do we need a special line for these kind of products? Do we need to isolate it? How do we isolate? How do we secure it? Make sure that it's packed the right way and isolate on the floor too. But then what we would do as well is once it gets to the market, we would track all of the potential returns.

(50:02)
And so we would do weekly monthly reports on understanding if the rate of returns for that product based on the type of defects that it was coming back for was actually higher than normal. So we would track the performance of the product once it's on the market as well, to really make sure that we can pull the trigger if we needed to say, hey, we need to hold back, there's a problem. There's something we haven't anticipated, we need to look into it. So we would be under surveillance mode for at least six months to a year to really make sure that whatever we put in place in advance upfront in the manufacturing phase would be efficient as much as possible.

Tom Noser (50:45):
That's really interesting. If I understand what you're saying, it's like, okay, the disposable shoe and the shoe that lasts forever, they still have a lot of things in common. So you want to understand what are those quality processes for the shoe that's supposed to last forever? We already have lots of measures around those. We want to make sure that the disposable shoe is going to meet all of the appropriate measures. Is that a reasonable way of looking at it?

Norbert Teston (51:13):
Yeah. There's a common foundation that would work for both in terms of standards and processes and everything that's on the line. So we would try to assess if whether that common foundation would be able to be applied for both products, and then we would look at ... Don't call it disposable shoe.

Norbert Teston (51:38):
We would look at that ... Let's call it technology.

Tom Noser (51:45):
The technology.

Norbert Teston (51:45):
We would look at that technology and then see what is very specific about this technology that we need to tack on to the existing process or the existing targets or controls that we may have in the existing process. So we would handle it a little bit as an exception, but really would make the upfront work of understanding what those exceptions are about and then making sure that we have system that capture those exceptions. And sometimes the electronic shoe, we put this entire plan together that was just for that electronic shoe that was related to the electronic components. The rest of the standards were still valid. They still had to be fulfilled. But then we shut that down once the innovation was scratched from the range. We were like, okay, well we'll keep that on file in case it comes back, but for now, nobody actually needs it.

Norbert Teston (52:38):
So we do ad hoc plans like this.

Tom Noser (52:47):
Tell me about a day that's fun. What's a day when you're doing what ... Because you obviously have a lot of interest and passion for what you do. What does a fun day look like for you?

Norbert Teston (53:01):
Tackling a big issue that no one knows what's going on. So it's a big puzzle. A lot of what I do is a lot of puzzling and is a lot of risk management. I'll give you examples of days that have really had fun in the past is ... I'm thinking of China requirements. So China came up with this entire matrix of standards that we had to fulfill and across our products. So there's codes that you have to pick to understand what standards need to be applied, and the code has to be on the label. And consumers in China are able to basically pick a product from the stores and send it to a third party lab, get it tested, and if it's not passing those requirements, they're entitled to get a more than a refund. They're actually-

Tom Noser (53:52):
Damages.

Norbert Teston (53:53):
Damages, yes.

Tom Noser (53:55):
Good lord. And the company gets fined for times the amount of products that have been actually distributed on the market. So there's gigantic incentives for a company to get it right. And this came up. I don't know why, out of the blue for us, and we hadn't heard about it, but suddenly someone was like, we need to address this issue. So we ended up flying to Shanghai. For a couple of years we did that on a regular basis to really just understand what this thing is about because we had no idea. And it was touching everything. That's the challenge. It was touching test standards, it was touching labeling, like I said. Factories. And what do we do during development? How do we assign the right codes? And what are those test methods? Do we need to reassess all of our standards? What's the current production? What's future production? What does it mean for innovation? How about if the standard doesn't exist?

(54:52)
Let's just go to China.

Norbert Teston (54:56):
I could see ourselves. We had this big group of people that we had gathered over in China office and flip charts everywhere on the wall. I'm like, what is going on here? And then until we started processing it, and it became clear in my head, I was like, okay, well, I think we've got some pieces of it. We're starting to cluster things a little bit and we're starting to understand what we need to do. And it was just the beginning of that long journey. But this was a major piece, and this is really what I enjoy doing, is when we had large challenges like this that would allow me to go into a complete different world. I had never been to Shanghai. I never connected much with the Shanghai team as well and suddenly I was in the middle of these people and then trying to figure out ... Talking to the lawyers, talking to the sourcing team locally, talking to the external labs to understand exactly what is going on with all these new regulations. Yeah, it was fascinating. I just really enjoy genuinely connecting with people and really understanding what's your world looking like that I don't understand that I need to understand if we want to solve this situation.

Tom Noser (56:11):
That's a great story. There's a ton of different expertise required to build a shoe, it sounds like. And obviously you don't have all that expertise.

Norbert Teston (56:22):
No.

Tom Noser (56:24):
I'm going to use the term which may be inappropriate, so feel free to tell me it's inappropriate. You are in some ways like an internal shoe regulator or manufacturing regulator. And that means you have to at least have enough understanding to know when people are bullshitting. How did you develop that? In that room it sounds like, as you say, we're starting to cluster some things, but there's people with very different expertise who probably don't understand anything about what's on the other guy's flip chart. How do you start to pull that together?

Norbert Teston (56:55):
A lot of it is for me to understand if it makes sense, and that sounds very simplistic. But what I tend to do in my own head is I tend to run through their process for myself. Be like, okay, I'm going to put myself in their position and I'm going to try to understand what their process is. And it's something I'm really good at generally, is to really run through a process steps and establish whether there's a hole or a gap or something that's missing there that could either break the process down where I'm like, how does that happen? What's the link between this and that? Who does what here when and how does that happen? And what system? So I tend to be very good at mapping this out in my head and then pulling out areas where I'm like, that sounds fishy. I'm not really sure about that. And actually the first time I did that ... And it was a great exercise. When I started with Adidas in 97, my boss who then became my boss for 16 years down the road after that, they decided to send me and a couple of other newcomers to the factory every day for six weeks.

(58:10)
And so there we were in Vietnam on our motorcycle, one hour motorcycle drive to the factory to sit down on the line with the workers in the factories back then. So we would have contacts and et cetera. And they would take us throughout the factory and make us work all of the different steps that are coming into to make a shoe. So this allowed me to at least map out in my own head, these are all the processes that need to happen and that are currently happening to make an entire product and that have been using this forever since I started. Because a manufacturing process tends to be fairly linear overall. But every one of those little step can go wrong. Something can happen there. There's always an important criteria to make sure that that single step goes right and is as efficient as possible too.

Tom Noser (59:05):
I love that. I'll say software desperately needs you. I can't remember who said this, but if we built buildings the way we build software or every building would fall down within a day.

Norbert Teston (59:24):
Oh no.

Tom Noser (59:24):
Yeah. There's all sorts of variables in building software that has a bazillion lines of code, just like there's 250 steps to make a shoe. There's gazillions of steps to compile software, and much of it is invisible. Much of that process is invisible. It's either done by machines or it's done in a ways where it's not really shared that well from one group to another. And so there's tolerances. What are your error tolerances? I also wanted to ask this about Adidas. So in software sometimes it's like, well, it's good enough. It's like, that's close. It's not really going to hurt anybody so ship it. Because if you spent all of your time fixing everything that was wrong, you'd never ship. When you touch one thing, something ... It's like a Jenga.

Norbert Teston (01:00:21):
Something else changes.

Tom Noser (01:00:22):
Something else changes, right. How do you make those decisions about quality in a manufacturing process when it's like, it's a little gray? Well, this isn't really what we wanted. Is it going to hurt the consumer? What's the criteria that you use?

Norbert Teston (01:00:38):
So a lot of it is a risk analysis. Yes. A lot of it is understanding what could be the impact going forward. Especially if your product is already there, it is been packed, and it's a container. It's like, okay, where does that product go? Is this a large production? Let's talk to the marketing team to see is the change really relevant for them? Is this going to make it different to the consumer? Obviously, if there's a product safety issue, then that needs to be ... Hold it. Hold the shipment and let's do an evaluation. Let's evaluate what the risk could be, and then potentially that risk is major. And then yeah, those product doesn't leave the factory. So it's really about understanding the configuration of the issue itself and then where it's going and what could potentially go wrong or what could be the problem if it does make it out the door.

(01:01:36)
And a lot of it as well is building that history after that. Because every issue you face is a potential for a lesson learned that you then need to integrate back into your systems. And so we would have visual requirements, for example, in the manufacturing phase where if there are stains or if there are little variation of the stitches or a little defect here and there, it's like to what extent are we accepting defects? What's the tolerance? And that would generally come from having conversations in the past about things that have been returned or things that have been claimed because consumers didn't accept this because it's a lot of gray. It's a ton of gray. It's definitely not a black and white. People want it to be black and white, but it's not.

Tom Noser (01:02:32):
To me, again, it goes back to what you said. People skip a lot of the times in the beginning is what are the values of the organization? So while you and I are talking, there was the whole AI dust up over people leaving ... This is awful. I can't even remember the name of this company, but it's the big AI company that built ChatGPT. OpenAI. Excuse me.

Norbert Teston (01:02:54):
OpenAI.

Tom Noser (01:02:55):
Allegedly the dust up between the CEO and the board was over risk. And we don't know who's on what side but they disagreed about how to manage risk. And so they fired him and then blah, blah, blah. But I feel like that's an area where software is just awful, awful, awful is understanding risk. I think one of the reasons why is that we don't have a particularly good history, and when we do look back on history, we tend answer it by like, ah, technology's changed. That's not a problem anymore. That would only happen back then because the technology is new now. It can never happen again. Did you ever get that kind of thing in manufacturing where people would say, oh, well that's not an analogous case because of blah, blah, blah?

Norbert Teston (01:03:48):
I think that the difference probably here between the two industries is the fact that footwear hasn't changed that much actually. Because you still need ... And that's why it's all produced in Asia. If you were to produce a pair of shoes in the US or in Europe, you'd have to pay $2,000 for that pair shoes because it's so much human driven. And they've been to all these brands that have been trying to automate the process as much as possible, but basically the process remains very much the same as it was 20 years ago. There hasn't been any gigantic breakthrough. And I think that's different from the IT industry where your platforms, your programming languages change and then suddenly-

Tom Noser (01:04:36):
All the time.

Norbert Teston (01:04:36):
All the time right? And so it moves at such a pace in terms of the internal innovations that you don't see that in the footwear or apparel industry as much as much as that. Yes. I've heard more the argument, oh, consumer's not going to care, I'll sign off on this. And then we would have to go back and say, actually I think they will because we've had, this situation already happened a couple of days ago or a couple of years ago. So this was more the argument that was coming. Or our standards are too high. That is the magic argument that would come back. And then I would always go back and say, what standards are you actually talking ... Because we have 350 test methods. So which ones are you talking about and what standards? Because for each test methods, there are 10 or 15 different standards depending on the type of product. So when you say our standards are too high, you're throwing this against the wall to see if it's going to stick. What exactly are you talking about here? And I would usually be faced with a blank stare. I'm like, okay, we'll do some research first and then let's regroup and have that conversation. Because more than happy to lower the standard if we think it's not appropriate for that application. But generally it was the same. It's like, oh, well, let's get it signed off.

Tom Noser (01:06:04):
We're trying to get to market. As I wrap up, I need to ask you something that I probably should have asked at the very beginning. How did you end up in Vietnam? Because you got there before Adidas.

Norbert Teston (01:06:15):
I did get there before Adidas. I was one of the last generation that had to do their military service in France. Because I grew up in France and went to engineering school in France. And so I did not want to do my regular military service. So one of the alternative solutions was to find a cooperation position. And so what I did, I ended up finding a company and I created a position for myself, and then I applied as a military person to that same position. I got myself matched, and I eventually did about 10, 11 months for that company over in Vietnam helping them with silk manufacturing. And that's how I got there. So I was detached from the ministry of the Army to the ministry of the industry. There was all these programs that they would do for graduates back then to try to utilize the manpower and workforce of the younger generation, give them a first experience abroad. And so that's how I got there. And there's a bunch of us and a bunch of French guys that ended up over there at the same time. So made a ton of friends in the meantime.

Tom Noser (01:07:31):
That's amazing.

Norbert Teston (01:07:31):
And that's how I got there.

Tom Noser (01:07:31):
So instead of doing pushups and carrying a rifle over your head, you were manufacturing silk in Vietnam.

Norbert Teston (01:07:41):
Exactly. That's exactly what I was doing. And I was like, great. That was just perfectly because I did not want to be an army base. I just did not want that at all.

Tom Noser (01:07:52):

You've had a spirit of innovation your whole life, right?

Norbert Teston (01:07:59):
Yeah. I worked really hard to make sure that wouldn't happen.

Tom Noser (01:08:01):
Norbert, it was a lot of fun talking to you. I really appreciate you spending time with me today.

The Fortune's Path Podcast is a production of Fortune's Path where we work with technology leaders to address the root causes that prevent rapid growth, fractional product management, product leadership coaching, competitive intelligence. Find your genius with Fortune's Path. Special thanks to Norbert Teston for being our guest. Music and editing of the Fortune's Path podcast to my son Ted Noser. Look for the Fortunes Path book from Advantage Books on fortunespath.com. I'm Tom Noser. Thanks for listening, and I hope we meet along Fortune's Path.