Episode 261: Tech Summit: The Largest Gathering of Technologists in the Promo Industry

Welcome to Skucast, the podcast for entrepreneurs in the promotional products industry. Skucast shines a light on our industry's best work, features maverick personalities, and discusses what's really involved in running a modern promotional products business. Skucast is the official podcast of Common Skew.

About 2 weeks ago, the largest gathering of technologists in the promotional products industry happened at Promo Standards Tech Summit. Over 150 attendees from over 60 companies gathered together at the premier tech event. And today on the SkewCast, we're sharing with you a glimpse behind the scenes at Tech Summit, which includes a panel led by Common Skew CEO Catherine Graham and also a breakdown from Robert Cathro-Oliver, Commsky's Director of Software Development on the PROMO Standards Committee. But before we do, I invited Dave Schultz, Commsky's Vice President of Supplier Partnerships, who was also there to join us and answer this question first. What is Tech Summit?

Well, Tech Summit is an event that was actually orchestrated and planned by the PROMO Standards Organization. You know, historically, PPAI as an organization has had a technology summit in years past that was usually held in conjunction with NALC. And you know, since COVID it hasn't, they haven't relaunched Tech Summit as an event. So Promo Standards as an organization took it upon themselves to create an event focused around Promo Standards as a standard, bring together developers, IT leaders, business leaders in the industry to talk about what's going on and push the standard forward.

What kinds of folks attend Tech Summit?

It's kind of a mix. So, right, like Catherine, our CEO, was there, who serves on the Promo Standards Board. Robert, our lead developer, was there, who is very deep into Promo Standards integrations. And then myself, who kind of straddles, you know, both worlds with a foot in development, a foot in the business side of things. And I currently sit on the Service Provider Committee for Promo Standards as well. So it was a great opportunity for me to kind of see what's going on in the business world, who's adopting the standards, and then also learn a bit more about the more detailed side of things in terms of what standards are being developed and where that sits today.

So is everything basically revolving around promo standards per se, or there are other topics being discussed at Tech Summit?

It's kind of a mix. It's very heavily weighted on promo standards because that's what the majority of the people were there for.

Right.

But there were topics related to cybersecurity. You know, there's been a lot of cyberattacks over the last couple of years. So some really good presentation by Mike Pfeiffer with American Solutions for Business talking about an attack that they had against their servers and how they dealt with it. And then, you know, some sessions also related to the business benefits of adopting promo standards and what it could do for you as an organization as well.

Right.

Many of us hear the word tech summit and we think, okay, a bunch of technologists got together, they nerded out over tech. But this year was really kind of different. Like it was the largest gathering of tech focused professionals ever, which signals a lot of momentum around this topic. Why that momentum now?

Well, if you think back, you know, I've been doing this for 20 years and there was a major initiative to standardize data in this industry almost 20 years ago.

Yeah.

And, you know, unlike some other industries that are much quicker to adopt technology, this industry is not one of them. You know, you've got— I have a factory right down the street that's using 100-year-old Heidelberg presses. So if you have companies in the manufacturing side that are using 100-year-old technology to manufacture products, you can imagine what some of the IT infrastructures at these companies look like. So fast forward to 10 years now, I guess, when PromoStandards was launched by a number of distributors and suppliers that historically looked at themselves as competitors. They all got into a room together and sat down and said, we have to deal with this. It's ridiculous the different ways we're sending data back and forth. It kind of in fits and struggles initially to get it adopted. But a couple years ago, I think we were at 100 suppliers that had at least one service launched. Today we're at around 450, I think. So the momentum is definitely there. And you've seen another reason for the momentum in recent years is, you know, you've seen some competing standards launched out there. And we all know that sometimes competition breeds, you know, a little bit faster development track. So I've seen a lot of momentum from that respect from the Promo Standards Organization. And I think what they did in terms of being able to sell out an event like this and tie a hackathon into it was pretty amazing.

Let's back up just a minute because there may be some folks that are new that are listening. Can you explain exactly what Promo Standards are? And you said, you used the word service. I understand you mean whether it's a mentory or EPO. Can you explain what Promo Standards is? And then when you say they have 450 suppliers have a service connected, what does that mean?

Yeah, so Promo Standards is designed to be a body of data standards, so to speak. And it's an analogy I used in a presentation once was, you know, if you look back to when the first commercial flights were launched, you know, and you started to get the first international flights into the US, you had pilots and air traffic controllers in other parts of the world speaking different languages, and it caused a lot of headaches. And so the governing body in aviation decided that English was going to be the standard language for all pilots, all air traffic controllers, controllers all the way around the planet. So if you kind of think of that, this promo standards as the language that we want to transact data with in this industry, everybody is now speaking the same language, the same data structure. So if your supplier X and distributor Y wants to consume your inventory data, meaning they want to be able to tap into your inventory and see what you have in stock without having to call your rep, log into your website, things like that, that's what a standard has allowed us to do. And there's— I'm going to get this number wrong because I know they launched one more last week at the Tech Summit, but there used to be 8 standards, everything from inventory to order status to shipping notifications.

Okay.

There's several that revolve around product data and pricing, and then you have purchase orders and invoices as well.

Okay.

And is some of the momentum behind the fact that we've been calling the new rush for suppliers getting into promo standards and getting calibrated with promo standards, the new 24-hour rush sort of phenomenon that's happening that suppliers— there's a little bit of competitive edge going on with suppliers because back when we launched PCNA in the fall, as you know better than anybody else, they achieved 50% EPO integration. I mean, which means 50% of the POs that were going through common SKU customers' hands was going through via EPO straight to their factory floor. Without manual touches, which is a huge—

it was so fast.

Do you think some of this momentum was around that competitive edge?

For sure. And, you know, it's in demand by the distributors, too. You know, if I'm a distributor now and I place an order with you, I don't, I don't want to have to call my rep or go back to your website to see if you have something in stock or see what the order status is. And you're seeing more and more distributors now, whether they even know the term promo standards or not, you know, they're thinking in terms of an overall general connected workflow. Like, I want to be able to see inventory in my workflow that I'm in all day long. Yeah. When I'm creating orders. And that's been a lot of the push for it as well. We're seeing our supplier partners get calls from distributors saying, will you guys please do this? Like, I don't want to have to call you and ask if you have an item in stock anymore. Yeah, that's been a lot of the push as well.

Yeah, I'm glad you mentioned that about in the integrated workflow. That's a huge key. What do you feel like is— I'm always interested in your perspective specifically, Dave, because you have a very unique perspective.

Active in the business.

There's very few people like you who have been involved in all aspects of the business. What do you feel is the biggest—

that was a compliment—

what do you feel is the biggest takeaway from Tech Summit for you specifically?

I, I seeing the momentum and the number of people that are involved in it now, there were companies there that I've actually never heard of before, and that was a surprise. You know, historically I've known all the people in this industry involved in promo standards, so to see newer, smaller distributors and suppliers getting involved and really starting to push this forward was, was huge. And I think, you know, as we see the momentum coming out of that and Robert can talk a bit more to the, to the hackathon side of this.

Yeah.

You know, to see a number of teams come together across companies and across countries in some cases and in one week build out solutions based on Promo Standard Services that could solve a lot of industry problems was, was pretty impressive.

What's the number one benefit for a CommsCue customer listening that's not involved in it, but they're going to be thrilled that Tech Summit's going on?

I just, I mean, it helps our team see what's coming and move on. You know, they announced a new standard last week called the order management standard that will now allow people to not only submit POs electronically to the suppliers, but modify them. So if there's something, change in a ship date, things like that, that promo standards has never been able to address before, there is a new standards that the committees have been working on hard to make those things realized. So, you know, just in terms of reducing the friction, making your workflow that much more streamlined, it's— now's the best time to check it out.

Awesome. Dave, thanks for joining, my friend.

Yeah, absolutely. Thank you.

Robert Cathro-Oliver is CommonsQ's Director of Software Development and has been instrumental in the growth of CommonsQ through his work on the platform. Robert serves on the Promo Standards Committee, and I wanted to get an understanding of what the Promo Standards Committee is. Here's Robert.

I have been at CommonsQ for about, I think, 11 years.

Wow.

More years than I have fingers anyway. So, and I am a software developer there. I think technically as of today, I'm a director of software development.

What does it help us understand that if you know what? For those of us in the promo world, right, we're not in your world where you're a developer all day long.

What's your day look like?

Give us a glimpse in your world, like what's on your agenda today to work on or this week?

Today is an atypical day because we are interviewing. So I have two interviews and this interview here with you. So I'm talking a whole lot more than I might on a typical day. Right. Yeah, we had a dev team meeting. Normally, normally I just at my computer, head down. I do a lot of reading, um, blogs, tech stuff, whatever. Yeah. Um, I'm on Slack, so because people are asking me questions, I'm responding to support requests, usually working on some larger feature.

Dave told us a little bit about Tech Summit. Um, tell us more about your role at Tech Summit, like on behalf of Common Ski, why were you there?

I was there to meet people, mostly people, see people in person who I mostly correspond with by email. I'm also on the Standards Committee, which was having a meeting just right after Tech Summit, which is another reason I was there. And as a member of the Standards Committee, I felt I should be there. There's a lot of discussion about people's desire for new standards, improvements to standards. Problems with current standards.

Yeah.

That sort of thing. That's very useful to me.

Tell us what, what is the standards committee?

So the standards committee is a committee within the Pro Standards, uh, organization, and they're charged with creating new technical standards to allow distributors and suppliers to interoperate and do data exchange.

Right.

Give us an idea of who's like the type of companies on that committee. Like who, who is all sort of organized together to make sure, to sort of give feedback and make sure things work well together.

Yeah. So, so there's, there's an array of distributors and suppliers and service providers.

So a good cross-section of folks that represent both distributors, suppliers, service providers to make sure that everybody's working together and improving on the standards that you have is what I'm assuming a lot of this is around. What's your role like as you bring Commonsku to the table? Is it— is to make sure Commonsku is working better than anyone else with all of these providers, right?

Yeah, we, we want to have like strong integrations with all our supplier partners and to be able to surface the data that these suppliers have to our customers, whether that be product data or order status data or inventory data. The easier it is for suppliers to implement these, these, these standards and the less work there is for them to do it and the more guidance there is, the more likely it is that the integration will happen. Correctly and quickly and right, the least amount of effort.

Give us an idea of the kinds of actionable to-dos you walked away with from the, the standards committee itself. Like, what were the kinds of things you were encouraged by? What are the kinds of things you walked away going, oh, we need to do this? Give us some examples of that.

In from talking to people who are implementing these standards. So people from suppliers and distributors and service providers, The main thing I came away with is that we need to provide more documentation and stricter semantics of how the standards should be interpreted. Right now there's a lot of syntax, so we've got a data format where we can transfer data, but interpreting that data at both ends, um, there's a lot of ambiguity for each, like, pair of people connecting to like hash out amongst themselves. And if that was more uniform and you could just assume that people were doing it the same way, that would reduce a lot of effort.

I see. So let me translate that because Robert, I know how big your brain is. Let me translate that for, for myself and our audience. In other words, there was too much generalities in it. People were taking liberties as opposed to being specific, and it was causing far more problems and friction. And if there's more specificity, more details, more documentation, and there'll be less of that wrangling and running around.

Yeah. Yeah. It's basically the ambiguity around the standards is— okay. It's— it makes it difficult to be right in standard.

Yeah.

Yeah.

This was the biggest gathering of technologists in the industry ever. Why do you think the momentum is behind this now?

The industry is modernizing, I guess you could say. I feel like traditionally hasn't really been at the forefront of technology and now it's playing catch-up and it's— you can play catch-up a lot faster than you can.

Yeah.

Be in the lead. So we're hoping to get our industry to a point where it's near the forefront of, of these new technologies, which are hopefully more efficient, better to work with, easier to implement.

Yeah, it is great to see the momentum finally, though. Let's talk while you're here, Robert. Let's talk a little bit about AI just because you're here and we haven't had a chance to talk about this. I know you study this stuff all the time. And since it's so rare to have you on a podcast, what's your take on AI? I know this is a big question in general.

It's a big question. And obviously it's in like the forefront of everybody's thought. I am by no means an expert on AI, right? And I in fact have very little technical expertise in that area. AI is, yeah, like statistics on steroids at this point. It's kind of like the holy grail of the whole computer science field, honestly. It certainly has.

What do you mean by that?

It's got the potential to disrupt everything, like society as a whole, if everything can be offloaded to computers that we're doing with our daily lives. But also has like a whole bunch of potential to generate a whole bunch of crap on the internet that you have to sift through to find, to find what you really want. But it also has the potential to help you find the stuff that you really want to find even better. Right. My main concern about AI is that it kind of like entrenches, because it's so capital intensive to train new AI models at this point, it kind of entrenches existing, like social disparities. And it's— I don't know, I, I— there's a lot of social disruption potential.

Really good point. Yeah. Was there any AI discussion at the Tech Summit?

Um, there's a lot of data science, uh, discussion.

And what do you mean by data science?

Data science is kind of like statistics, which is kind of at the— it's like you take your data and you want to draw insights from it that you can act upon to improve your business, basically. And yeah, the keynote speaker was talking about developing processes and maturing as an organization to collect data better, process it better, clean it better, and use it to gain insight into how to improve.

It seems so encouraging to me that we have so much momentum behind what's going on at Tech Summit, and finally that there are folks working together. Seems like it has been people pulling each other, and it— this feels more like there's a concentrated energy where people are heading the same direction, at least. Is that your overall sense?

Yeah, it feels very collaborative. Yeah, yeah. And, and, and it's technologists who like technology and want to be working with fun, cutting-edge, new things that have real impact. There are obviously business-level concerns, but, but, but yeah, technologists like technology.

Thank you, my friend, for joining us. This has been, it's been great and enlightening. Thanks for sharing a little bit of what's happening behind the scenes at Tech Summit and on the standards committee.

Okay, well, thank you. Thanks for inviting me.

Technology, automation, and digital transformation are allowing supplier companies to scale their production and shop floors and warehouses in the promotional product supply chain. Katherine Graham, CEO of CommsCue, led a panel discussion with 3 industry production leaders and technology experts to investigate the impact and challenges of these changes. The panel includes John Norris, COO of StarLine, Phil Gergen, the SVP of Technology at Koozie Group, and Dono Deutsch, VP of IT Systems and Operations at Alpha Broader. The panel discusses some really fascinating topics. If you're a distributor, I encourage you to listen closely as you hear the complicated processes and challenges your key partners often face in their transformation journey. Topics like relaying real-time information on orders in production, robotics and how robotics is changing our industry, and little things we take for granted like breaking case quantities and what that does to production. And do you know what cross-input items are? I didn't. Or how human skillset and machine skillset come together to manufacture your order. Or maybe how blockchain can possibly one day impact transparency traceability, machine capacity versus labor capacity, and a whole lot more. Today's episode is brought to you by Commenscue, the work from anywhere platform that powers your connected workflow, enabling you to process more orders and dramatically grow your sales. To learn more, visit commenscue.com. And we're jumping right into the panel as Catherine asks a question about how to give process updates on orders as they are in production. Here's Catherine and Phil Gergen, SVP of Technology at Koozie Group, to kick it off.

Phil, you've spoken about giving greater visibility to customers during the production process and what that can do in terms of establishing greater trust. What do you— if you, you know, the equivalent of like Phil makes your pizza in terms of is going into the oven, it's, you know, all the visibility along the way. What are the challenges you see with implementing a solution like that?

Yeah.

So today, even within formal standards, we have a lot of order status, but then it comes in production. Right? It goes into this black hole that can take anywhere from, depending on the priority of your order, a few hours to days to weeks, depending on the imprint method, the processes. So bringing that trust into what the process is, is there places that we can, you know, the technology is there to easily take pictures in the process, but then what does that take from putting it on the shop floor, making sure it's there always, giving the users all technology to take those pictures, to do those things. The more status updates we can give, it's the better, right? Because this industry is a lot of hand-holding of orders, right? Everyone wants to hold it, hold it, hold it, check on it, make sure it's good. We want to help you build that trust, but we don't want you just to trust us. We want to show you what the process is, make sure that the times you do need a handhold, you can identify it quickly, understand that. And so giving out the processes without making it looking like you're building sausage, right? No one wants to see how the sausage is made. They want to know the end product. But in this industry, a lot of times they kind of want to see how the sausage is made. So building that out on the shop floor, giving those tools from processes that are digital that can be easily done to the hand-holding ones and have a consistent model that you guys can depend on as distributors in the industry is really hard as we do so many different ways to do things. So that's one thing that we consider all the time from the process, from the handling process to the digitization process. How does that look consistent, feel that you can trust what those statuses are and that you're going to get what you need when you need it?

Yeah. Shifting gears to software, John, Starline was one of the earliest drivers of promo standards. You know, what software changes has Starline been able to implement that's helped with automation efficiency as a result of data standardization?

All right, that's a good one. It's a loaded question. So actually, so Peter's back there and one of the early things that we had saw In a part of our processes on the floor, every single order that's starting gets a picture of it, right? Mainly we were using this for our own QC purposes. The graphics team is constantly reviewing real time, you know, just looking for oddities of print, looking for misprints. And then we kind of use it internally as a, well, this could, if something goes wrong, we have proof of what we actually printed, then we don't have to try to get all the stuff back. So we'd had this data or these pictures and this library of them for years and years. We had no way to communicate that. We had no use for it other than internal purposes. And then I had a conversation with Peter the one day. He's like, well, I want those pictures. He's like, I want to give those, I want to feed those up to my customer. And we had no mechanism to externally expose those. And we had hijacked into the first order status 1.0. We put it in one of the link, one of the random comment fields, and Peter was taking those. I know what Eric is going to talk about tomorrow, where status 2.0 now has these mechanisms to give some of that real-time data and images we're collecting and feed them back up the supply chain, create better customer experience, other software automations. So the biggest thing, and I learned this with packing swag bags in the last few days, and I've— you've heard me rant about counting. The biggest challenge we have in this industry is counting. My biggest thing when we hire somebody is like, if you can count to 6, 12, 24, 25, you'll be a hero. You know, we could probably I had to count out 7 pens for every table this morning. I think half the tables have 6, right? So counting is a big challenge. And in our organization, us knowing the quantity we actually shipped at the end of the day was one of the largest problems. Like, what did we ship, right? So I know a customer ordered 100, I pulled 101, I might have scrapped one. And now it's trying to bring a lot of that automation and the inputs for data back to the production floor. No longer, you know, trying to eliminate the accounting, eliminate shipping, you know, bringing all of that, putting on the factory floor. So more label printing at the press, more, you know, bringing it real time from a— it's more of a process enabler for us. So yes, having your product data game in order probably limits some of us from doing some of the dumb things we shouldn't be doing. So you take the flexibility out, which helps drive the automation. The more that we can stay on track and not customize, we customize every order, but not create custom solutions for everything. I think just an overall mindset of data standardization, process standardization has really been massively beneficial for us. And growing through all of this, growing hides a lot of sins, right? So as we've been growing, doubling our growth over the last few years, you know, a lot of things get swept to the side, but at the core we stay down. Product data is our number one focus. I mean, Stuart, he's a big believer in our product data, but if we can get that right, a lot of things just fall into place afterwards.

So Filacuzzi, it not just doesn't just decorate, they also manufacture a number of things in-house. Can you speak to how having control kind of of the entire process, you know, from beginning to end enables greater flexibility or efficiency from a software perspective?

Yeah, it does. It eases the digital process that John hit on the previous point where if you're coming in, you're digitizing, you're getting right to the press, getting that information. If you know your materials, your product, you know, you can do quality checkpoints earlier than you can in other processes. We know the material of the koozie, we know what we're printing on it, we know the material of the pens that we're making, we know the quality. There's the variances get a lot easier. So that makes the digitization of that process easier. We don't have to work for the handles or even the handle being off or different around on that. We're able to, if we want to make a different size, we know exactly what size we're making is on our equipment, our machinery. That process becomes a lot easier to drive on anything that we manufacture ourselves. It also helps with the inventory and the supply chain and that process where we know what's happening. There is no train that's going to slow us down. There's no boat that's stuck somewhere. Sometimes there's still trucks, that are stuck, but it definitely helps us know that process, be consistent, and drive that consistency as we go through that.

Donat, robotics projects tend to focus on hardware as being the success story, but there's a significant software lift in implementing a large-scale robotics transformation. Can you speak to what that entailed in terms of implementation?

Absolutely. Yes, you're absolutely right. There's a huge software lift associated with the integration of the robotics and integration of and the enablement of the robotics. Solution as well as kind of changes necessary as a consequence of it, right? So, you know, we have 9 distribution centers in our network, right? Each with their own instance of WMS. So we had to build out 9 individual API endpoints, right? And that WMS is managed by a single code base, right? So the rollout took, you know, it's about 12 months, right? So during that period, we have DCs that are on robotics and not, right? That single code base had to support both. So all code deploys had to be forward and backward compatible, right? Some examples of the effort and challenges from a software perspective. So at the point of pick, we present as part of our process, we present two images to the picker. One's basically a pick location. Generic kind of monochromatic icon, a balloon, a car, a boat, that kind of thing, right? So he sees a balloon and he sees a balloon. It's a quick reference. The picker knows, oh, I'm at the right location. Then we present the product image, right? He sees a gray hoodie, he's picking, he's counting out gray hoodies. So now he's reasonably confident based on those two pieces of information he's picking the right product at the right location. Challenge is the robot can only display one image, right? So our developers had to figure out a way to programmatically merge the two images into one image file on the fly at the time of order drop. Right. So, you know, and that proved to be a bit challenging for them. Another example, you know, if you think about, you know, the amount of messaging that happens that goes back and forth between the robots and our WMS system. Right. And the messaging is asynchronous, which is a good thing. Right. The bot doesn't have to wait on the WMS to acknowledge the receipt of a message. Before it goes off and does this next thing. Unfortunately, in a single-threaded messaging architecture in high volumes, right, those messages can get delayed and problematically so. So we were finding situations where, you know, a picker would say, all right, pick complete. The bot would send off the pick complete message to WMS. It would kind of get hung up in the queue, right? And then it would go off to ship, right? And the shipper would say, all right, time to ship in the WMS. WMS says, no, you're not done picking. Right. So, you know, we had our developers had to figure out a way to develop a multi-threaded messaging architecture to kind of alleviate some of that. I think lastly, I'll wrap up on an example on how the robot, the robot solution forced changes from other parts of the operation that required software changes. So more specifically, our putaway process. So prior to the robots, I'll say prior to the robots again, The pickers drove pack mules around doing the picking and our put away folks driving man lifts did their work and they got along just fine. They just managed to avoid each other, get rid of the mules, put bots in. And we learned that the robots are very good at looking horizontally. They don't look up.

Right.

So we're finding situations and surprisingly often, right, where a man lift operator is up 25 feet in the way, in the air, right, putting away a skid. Of cases, right? While he's up there, a bot comes along, parks right under him, right? Waiting for a picker to come and complete a pick. He finishes his put away and comes right down to crush the bot, right? Like a lot, right? Like you think that's got to be a fluke, right?

How often can that happen?

So it happens.

It's us asserting our dominance over the robot.

Exactly right.

You start to wonder if that had to be intentional, right? Yeah, so we had to rework the entire putaway process in software in order to stop that from happening. Right. Because that'd be a pretty expensive thing, as you can imagine. So, so yeah. Yeah, for sure. You know, I think you're right, Kathryn. Robotics tend to be seen as hardware solutions, but they're also equal parts software as well. So, yeah.

So let's talk about software opportunities. John, orders that break case quantities create challenges for creating a more automated process around shipping and handling. Human decisions are made around master carton size. How could software help solve this kind of problem?

Yeah, minimums and breaking cases. I could rant for hours on these, but, you know, essentially things come in, let's call it 24, and obviously we try to set our minimums at 24 and, you know, hope that, you know, 1 out of 10 times you actually order case quantities. But the most process labor inefficient thing we can do is to break cases. Right. So by the time we break cases, it's a disaster for inventory. It's more handling. That's part of the picking process. Like once we start breaking cases, that sucks. But one of the things that I guess we struggle as an industry is we have so many SKUs. I know I was at one of your panels that you were running with Alpha and the SKU reduction, right? Getting those SKUs down because we just have— I mean, Starline has 2,500 SKUs. So that means right now we have I think about 1,000 different box sizes. So our biggest challenge is to— we're looking at different pieces of software that we can tell the operators exactly how to pack it in what box. But I have 1,000 different box sizes. So we're trying to limit our, you know, from both from our factories and overseas or domestic factories and then just internally packaging. How do we get from 1,000 different carton sizes down to— hell, I would take 50. And then there's box automation software, which then if I take 12 different pieces, a mug, a flashlight, a bag, it will optimize the case pack, the best optimal way to pack it and what box to put it in. I think everybody has the experience with Amazon where you get a tiny little pack of batteries in a box this big, right? They ruined shipping for the rest of the world because of that. And so we're trying to explore those automation pieces to avoid that massive box, 3 things in it, because A, customer doesn't want to pay for shipping, just everything, you know, I don't want to pay for more boxes than I have to. So there's tools out there and it's not replacing the thought processes of the packer, it's more assisting them. You know, say there's go grab this exact box or I can bring them that exact box with the order and the pulling to try to, you know, optimize packing solution. So, you know, as we as an industry, unfortunately, the more we try to customize, the more it goes against these automated and semi-solutions. So it's a balance between SKU reduction, selection reduction, and trying to offset that with semi-automation or even software tools such as, you know, packing software tools to find a happy medium between customer selection and cost optimization.

You call that co-boxing? Co-boxing.

Yeah. So robots one day do this for us.

Phil, machine scheduling is a huge challenge, particularly with cross-input items, meaning items that can be decorated multiple ways. A lot of focus tends to be on inventory planning with less focus on decoration planning. How can software help solve this when there's limited visibility into what kind of demand is coming down the pipe for what kind of decoration?

Yeah, so we all know the time from order to shipment can be, you know, as little as 24 hours, right? So we don't have a long demand curve. And with different products, you know, we have to talk about SKU count, which is big for inventory, and then decoration count. We average, I think, 2.1 decoration methods per item, right across the board, somewhere 3 to 5. And then we have flexible machinery. So we have a machine that can be maybe 30 items can be printed on that. So, and you're looking at demand in a short period of time, trying to understand what's going to be there. So then you have machine flexibility and then you have operator flexibility. Some operators can work more than one machine, some can't. Understanding all those inputs and having a software that can take machine availability, human availability, human skill set, machine skill set, tying that all together to help us plan and organize will just continually make everything easier from having people there on time. You know, all the leftover temp work that we do because we're looking last minute, trying to decide what to do. The overtime. No one wants overtime. We want to have employee engagement. We want to have all the employees happy to be there, know when they're going to be there. That will help us drive everything within our shop floor to bring so many different things together. But there's a lot of inputs, a lot of data that still needs to be collected because in real time a person can look at you and say, I'm confident, or I can ask you, what can you do?

Right.

Hey, you've been on this machine before. I know this is an easy order, you know. Okay, go work this easy order on a machine that you're not as familiar with. You might need a more experienced one doing the exact same order. How do you calculate that and drive all those inputs and get them across so we can make good decisions going forward? That's really a really hard process, and we're really looking hard at what are all the inputs to collect? How do we make sure that's there? How do we get all the flexibility? And skills matrix in so a piece of software can do that because it's a math problem once you have all the right inputs and it's a relatively easy math problem, but the inputs are the hard part.

Has anybody fully automated their scheduling for production? Exactly.

Yeah.

Zero hands.

Anybody partially automated scheduling? I mean, other than itself. I mean, we obviously do board itself, but there's still a lot. It still fits the need after that, right? You get, okay, here's your scheduling report. Here's what you're doing. Okay, let's manipulate that a little bit more because what happened? You know, who's here? Who showed up today? What skill sets do they have? Two people not showing up, if they're your most flexible and experienced, those two can really drive problems for days to come, driving that back out to, to bring everything forward.

Switching gears to a more kind of future focus, data is one of the critical drivers in identifying opportunities for transformation. What are areas of opportunity for better instrumentation or ways in which we as technologists can be bringing forward data insight to strategic decisions? Product and marketing decisions can often be made based on trends and customer demand rather than taking a full product lifecycle perspective. So Phil, to your point, you know, what can we do versus what should we do? SKU proliferation and complexity of decoration options make standardization difficult. Can each of you speak to ways in which data has helped drive decision-making and ways in which you wish data would influence decision-making to sales and marketing?

To the point, we did a lot of work on what input methods were profitable and what were usable, and some were profitable, didn't happen enough to have the machinery there. We do that analysis very regularly with our SKUs. We started out as an organization, and I've been through lots of iterations over the 20 years I've been in this industry of what can we do, right? Because you get a machine and someone says, this is really cool, I can do these 17 things on this machine., and then someone says, okay, let's build it all into the system and build it into configuration orderable. But no one was really looking at that of what we actually ordered. We were looking, what's our SKU count, right? Just inventory-based information. Having all the information about what we can do, what we can do profitably, regularly, and consistently, those 3 things together then provide the experience back to everybody. And that's a huge one that we've done to help just make sure it's data that we had. To do that. There's still more data to— on the manufacturing side to help get that better and to get those decisions better as well. But there is a lot of data in this industry already, just making sure that we're using it. We talk about the technology, what we're capable of. You know, every time we put out something new, I tell my team what year we're at. I think we're like 2008 right now as far as how we do this, you know, what this industry is like. So we it's not technology problems always. The things other people have solved with technology, it's making sure that we have the data set in a way that we can do things that are out there.

And kind of to elaborate on that, and Tom touched on this morning, is so it's one thing to collect the data. It's like, well, now what do I do with it? Right? And that's the jump. So this was, you know, we like to fail often. And John remembers this nightmare working at Starline. We had the idea to equip all of our equipment with Raspberry Pis, right? We built a literal army of Raspberry Pis, mounted them to every piece of equipment, mounted accelerometers as one of the pieces on there. I wanted to track when my machines were actually printing versus sitting idle. So for an entire year, we had an army of 50+ Raspberry Pis constantly collecting data. We had dashboards on the wall when they were, you know, printing versus not printing. At the end of the day, we scrapped the whole thing and just came to the realization it was gee whiz data. So it's just trying to get in that, like it was cool as hell, right? We had Raspberry Pis everywhere, accelerometers. It was like the nerdiest thing we've done. It was great, but we didn't do anything with that data other than like we tried to use it as a production management tool of, you know, ultimately the goal was back to scheduling is how do we integrate this into a, if I use that data for real-time scheduling, but we never made that jump. So our cautionary thing is trying to separate the gee whiz data and the like the cool projects from what's meaningful data that we can help drive our decision and the decision-making process, maybe for production scheduling and more of just, you know, on versus off. Like, great. Like I can physically stand out there and tell you if the machine's on or off, but me knowing that it's 63.2% runtime, I don't know, we weren't at the scale to make that jump, to make that overly meaningful, to get a quick ROI. So that's always been our cautionary tales. We have tons and tons of data, but like, how do we translate that into actionable, meaningful information and like talk to your marketing people, right? They have tons of data and you can interpret it a thousand different ways. Like everybody has their website analytics and, you know, what's the big adage? Like 50% of my marketing budget is actually useful. I just don't know which 50% of it it is. So, you know, that's one thing we struggle at. And I think, you know, back to Tom's thing, just getting a hold of your data analytics and having a plan in place, I think goes a long way.

So yeah, matching those data points, right? Because that's probably a useful data point. But what do you match it up against? Because if you don't have anything to use it against and measure it with, having that measurable, what are you going to get when you collect it, right? What's the business goal of that data, right? And what decision would you make differently when you get that data? Sometimes you're like, oh, this is, this is data. But what, what we were going to actually do And like you said, you probably knew the answer before you put up those Raspberry Pis. It didn't change the way you did your business.

Yeah, I knew the machines weren't running, but I had to prove to myself what 50 Raspberry Pis is.

So machines are great. I'll take a little bit of a less practical and more kind of conceptual approach to that, right? So if we're kind of being future-focused and, and I think, you know, anytime you're talking about the future utility of data, I think blockchain needs to be a part of that conversation, right? Especially when you're talking about data exchange, you know, inter-organizational data exchange, right, with customers and suppliers, right? And I think the, you know, the potential enablement here is around ESG, right? ESG is huge, right? Huge nowadays. And it is a big priority for us. You know, customers and end consumers are mandating that we ensure our products are ethically and sustainably sourced, right? Which is, you know, is labor intensive, right? You're talking about products sourced and produced all over the world, right? So because at its core, blockchain is a distributed ledger, a distributed immutable ledger, it can provide kind of, you know, this permanent tamper-proof record of every transaction and every event in the supply chain, right? So if you make that data publicly available, right, it can provide a great amount of transparency and traceability that's needed in order to demonstrate your ESG compliance. So I think, you know, obviously this is far— it's not something we're going to be doing today, but I think it's something conceptually that we can be looking at, especially as the need, you know, ESG is just going to continue to grow.

Right.

So and I think there's other utility, just track and trace and just operational efficiencies from a supply chain perspective. Right.

So I want to open it up for questions. What questions do you have for these guys?

Our company deals with a lot of seasonal fluctuations in demand. Which I know everybody deals with. And, you know, one of the main drivers for our automated projects is like increasing capacity and flexible capacity. How much is flexible capacity a driver for automation in your facilities?

I can take that one. The great example for us is koozies, because the koozie is obviously a very summer product. I mean, we have them around, they come And like that, drinkware is obviously flexible, but that one, the more automation we can build on that because we build that from scratch to material, we're bringing in another machine this year to be able to increase that capacity. So it's not a labor capacity issue. It becomes a machine capacity issue, which we then can build on that, drive it up, automate around that. So that's an example where we are changing our processes and adding equipment instead of adding labor.

And one of the natures of this industry is having that spot capacity, right? I mean, The dumpster fire is I don't know what's coming in tomorrow. I could get a, you know, something that eats up 80% of my weekly production capacity could come in tomorrow. I knew nothing about it today, but now it's my problem. So having that spot capacity is huge. And then, yes, with the seasonal fluctuations is we unfortunately are handling that by stretching shifts. I think, you know, both of these guys mentioned that, you know, well, now I don't have the, you know, that overtime thing comes up a lot. Unfortunately, the human side has to work more hours in order to handle that fluctuation. Obviously, you see all shift scheduling, it helps as well. But back to the semi-automation, semi-automation works because there's a human semi-assisting this machine doing it. So then they work. Good for us as humans. We're not getting rid of us. We need that. But the biggest challenge I face and like running a production environment is the staffing and scheduling and basing it around that. It's like, I know my business is going to get 43% bigger in the fourth quarter. It's like, but I can't afford to carry 43% extra staff for 9 months. So I also— but I have to buy the equipment to be able to support that 43%. And how much do I get with extra equipment and automated equipment versus, you know, just running it longer?

Right.

And so it's a balancing act and I'm not sure there's a great answer.

But I mean, yeah, finding those decisions, right, where you actually invest. Right. Because investment gets very expensive, as you mentioned earlier, you know, half a million, $1 million for these machines. That give you that capacity, you have to be very specific in what you do and be very sure that you're not going to invest in something that demand will change, the industry will change, something will cause that demand to fall when you just made this huge investment into increasing your capabilities.

What's a robot go for these days?

Actually, so we lease it. So, so we were able to keep our capital investments down and, and it scales with, with business, right? So we can have robots parked over there. If we're not using them, we're not paying for them. So that's one of the great advantages of it. It scales almost in real time based on business demand.

Good time for one more question.

Yeah, usually for other industries, there's like open source technologies that can be applied to, let's say, multiple factories. So I'm curious, is there similarity or difference between like each decorator in how you do things?

The challenge is, so yes, there are, at the end of the day, there's, call it, 20 methods in the industry. Like there's, you know, That is one piece. But the problem is the SKU. I mean, you guys have 100,000 SKUs. I mean, there's the trying to get a flexible method and a flexible, you know, just even to get close to that solution. Unfortunately, what makes us unique as an industry is, yes, we have 800,000 different products and we can decorate millions of different ways. And that nuance is what makes this industry unique and I'd say probably successful. So it's that fighting against the we can do anything and put a logo and do it into any physically shaped product in the world and trying to merge that with automation or semi-automation. They don't get along well. They don't get along well at all. And, you know, as technologists, it's like we're in the middle. Can you use technology, use software, use standards, use protocols to at least make it cost effective to where, you know, we can still make everybody's sales and marketing departments happy by printing anything on everything. Versus the production side, which is like, I want to do one thing on one color of this one shaped widget, which in other industries, if I'm just making refrigerators, that's cool. I make one refrigerator. I don't make 150,000 SKUs of anything and everything.

Right.

And I think he's talking about a data standard too, around the naming and the structuring. Right. As soon as you get a data standard for what you call an imprint method, your marketing team will say, well, we want to call it this.

So we want to sell you something, but it's And distributors have no idea what this means.

Yeah, no idea. Right. You know, I think as Norwood and Big Graphic, we came up with the term BrightPix a while back, which is really just full color. But someone decided this is going to be our go-to-market for this new term. Right. But now everybody, do they put BrightPix as a method? Is it full color? It's really just a full color imprint. But in the structure of the order, in the place that we put it in our dataset, it went in the place of full color. So then you have to give a different designation of is it full color or is it not. So that's agreeing upon a data structure is going to be a hard one for a while, I believe, even though we have the transmission process today.

So all the challenges that we've just talked about, I'm expecting all of you to solve over lunch.

All right.

So with that, please thank me and join me, this crew.

Thanks so much for tuning in to this episode of SKU Cast. Be sure to keep up with our latest content by subscribing to SKU Cast on iTunes or to our blog at community.commsku.com. Until next time, friends, thanks so much for listening. Mm-hmm.