PING Achieves End-to-End Supply Chain Planning: A Case Study

Achieving end to end planning at Ping, a case study. For that, I'm joined by Scott Neiman. He is forecasting manager with Ping. Hi, Scott. Hi, Bob. Nice to be with you.

Good to have you. And Matt Hoffman. He is vice president of product and industry solutions with John Galt Solutions. Hello, Matt.

Thanks for having us.

Great to have both of you today with me. So let me start with you, Scott. Tell me about Ping. Who are you guys?

Absolutely. Ping is a golf equipment manufacturer. We are located in Phoenix, Arizona. We have global locations, elsewhere as well. We make golf clubs primarily, as well as an assortment of golf related, soft goods.

Okay, and a leading name in the field as well.

Certainly.

So what was the supply chain challenge that you faced that we're gonna be talking about here today?

The simplest way I can put it is that we needed to house demand and supply planning in one central location to be able to see data flow from a high level demand forecast through our various, what's now called scenarios, and eventually achieving the end result being component procurement.

You just lacked that visibility before?

We did.

Must have been tough getting by with that?

It was. We had some disjointed systems, processes, Excel workbooks that we hoped spoke to one another that usually did, sometimes did not. When they didn't, it might be tough to recognize that they didn't. In addition to that, all housed in one place, I would add that we wanted the ability for our demand planners in our marketing department to be able to see how their demand forecasts flowed all the way through eventually down to our supply planners resulting in those component buys and have input and sign off on what those buys would look like.

Okay. Good challenge, good initiative to undertake. Why did you pick John Galt Solutions to guide you in this initiative?

We started a search in, I believe it was twenty eighteen, talking with various industry partners. And what we wanted them to be able to show us was you can take a high level forecast of ours, didn't matter what the numbers were per se, but we wanted to be able to see, okay, let's take what we think that demand is going to look like, break it down into assembly planning because we do manufacture, we assemble in Phoenix for much of the globe as we do in the UK, Japan and Korea.

And then from there, from the model level, breaking it down into the component SKU level resulting again in that material plan that we talked about. So when we started talking with companies, we wanted them to be able to show us very clearly what it would look like. That we needed to be able to level load our production, get it down from the model level to the SKU level. And it became clear pretty quickly I think that in meeting with a number of companies that some couldn't exactly show us what we were looking for. Then once we found John Galt, I think we did maybe a virtual session to start and then a live demo. And it became pretty clear that they were the partner for us.

So Matt, you won the beauty contest. Congratulations there.

Exactly. And it was pretty quick by the time we got involved. And I think that went well.

Did you feel that this was kind of in your wheelhouse as the type of engagement that you had undertaken before with previous customers?

Absolutely, so we worked with a number of different companies across a wide variety of categories and channels. And so as we looked at what Ping was looking to do, we've got the flexibility of thinking not just at an item level or an item location, so we were able to model the model, pardon the pun, that Ping was looking for.

So that made it very easy. Then we're not a customized software. What we do is we build in the functionality and so we've got this flexibility to go through and meet the needs without having to go write custom code. So what they gave us a lot of confidence in doing was the ability to meet the needs, plan at these different levels, and then rely on having worked with a number of different companies that do make to order, configure to order with all of these variations that Ping saw.

Certainly, Scott, you make a customized product, so that adds to complexity in your supply chain quite a bit. Okay, so you pick these guys, you're off and running. What was involved in implementation of the solution, even as you had to continue to do business as a going company while you were putting in this new tool?

The first thing was figuring out what data we had even available. I mentioned our previously disjointed processes, a lot of homegrown. I mean, Ping was founded in nineteen fifty nine. And solutions, in house solutions had been developed over the decades, some of which existed four decades.

Because of that, we weren't always sure—Our IT department is great. They did as good a job with those homegrown systems that they could. But there were certainly, and they would echo this, there were a lot of pieces of information that were housed in different places.

Some more readily accessible than others. So I think the first challenge was just figuring out what we could get our hands on to feed the engine, as it were. And then cleaning it up from there and being able to do what we needed to do from a sense of high level model forecasting and getting that down into a component level.

He mentioned the customization piece, I would echo that. With all of our various component offerings, it's an extreme, I don't exaggerate to say it's an extreme undertaking to start with a high level model forecast. We have some really, really talented demand and supply planners that are able to break out what those forecasts might look like at a percentage basis, and Atlas helps us to do that.

Matt, what did you bring to the table in terms of support and resources to get this thing implemented?

We brought in our own services team. So one of the things that we're very proud of is we work with a number of different implementation partners, but we've got our own services team. And so our services team has a lot of familiarity with making Atlas work without having to do customization and works very well with our R and D teams and our support.

So a big element that we were able to bring to the table was talking supply chain, but also then making sure that everything that we designed fit within the way the platform works and would be something that our support team would be able to cover as well. Because what we didn't want to do was build this bespoke solution that just never worked and wasn't reliable. It was something that really had to be bulletproof.

Yeah, would add to that, I'd echo everything that he said in addition to that. It was the best solution that we saw of being able to, and we saw more of this as the further we got along, but to be able to tailor views and uses for our end users to those end users. So however they wanted to see their particular roles and information, Atlas said, no problem. We have the ability right now to tailor it to how you wanna see it.

Were there any surprises along the way? I mean, you learn that you had capabilities you didn't expect to have? Or for that matter, were there any lessons learned that you didn't anticipate in the implementation process?

The first question is tougher for me. I'll need to think on that for a minute. But I can quickly touch on the second. I mentioned the data challenges. I also think that we were, and this is not a poor reflection on John Galt in any way. I think we had a bit of a misconception of how quickly we could truly go live once we undertook this project.

And they were upfront with us. I think maybe we were a bit ambitious. But once we realized that we needed to take a scenario by scenario approach, right? So we started with the model forecast and then we might have implemented some elements of MRP before we really got to the bones of component mix. So I think that we just need to learn that you truly do to walk before you can run. And that took us some time to realize.

Another challenge or lesson learned that I would say is when we start forecasting a new product, we're on a rolling twenty four month basis. And we have an idea of what those products are going to look like years down the road. When we're first starting the planning process for those, we don't know what that name is going to be. So something as simple as a product name, it took us a while to understand between the two of us that that's going to change and that has to be okay. That's non negotiable because we started thinking that the name would be a key field. Then as soon as we changed that name, we're going to have to basically start from scratch with the product in Atlas.

Which ended up being the case?

It did not. Correct.

So we ended up, it boiled up pretty early on and then we created a solution to make it work. And so as mentioned, Scott and the team did a great job. We had a lot of, I'm going to call it sprint sessions, so there's a lot of interaction between here's how something works, does this meet your need, is there something that we're missing? And that really helped I think understand for both sides. 

And I think one of the big elements for Ping, because they'd been doing this in some cases manually and in some cases solutions that they at home built, it took a little while to translate that into something new. And so I think the team was great in working through and being interactive and that was a big element of making the project successful is that going back and forth and, okay, here's what I meant when I needed this. Great, now we'll make it fit for the persona. We'll make it different for a buyer versus a whatever.

So Scott, what benefits did you derive from the initiative? What's different now?

Many. First and foremost, I'll go back to what we were trying to achieve in the first place, which was housing demand and supply planning together. It's not a demand forecast workbook over here that we then have to upload overnight that feeds a different workbook. It's truly all in the same place. It's different scenarios, of course, because it needs to be. They all behave differently. But we're able to develop the demand forecast, flow through an assembly plan, break it out into a component mix, feed a material plan.

And I touched on it in my presentation, our demand planners who sit in the marketing department actually sit in on our supply planners' buy meetings. To where the supply planner, once they've done all their level loading, laying out launch assembly numbers, things like that, and feeding in all the normal MRP variables, once they have a recommendation for what they think we should be buying on a month to month basis, the demand planners are a part of that and give their input. Even though you might think that their job is done, right? They've done their model forecasting, washed their hands of it, and moved on. Not the case for us. We like more circular input than that. So, they actually come back in at the end of the process and give the thumbs up or feedback, whatever it needs to be to our supply planners.

Well, I'm guessing that the job toward your supply chain improvement isn't done either. Looking forward, how might you think of yourself being able to continue to work with John Galt for future enhancements?

That's a big question for us, I think. So, we're on the tail end, while in the US, we're on post go live with a brand-new ERP. They've been a big help with getting us back up and running with that. We'll be subsequently onboarding our other global locations into that ERP. I mention that because following those go lives, I think it'll give us an opportunity once things have calmed down a bit to take on some of what Matt and others have talked about with more statistical modeling than we've done in the past scenarios, right? Scenario planning, certainly getting into some more AI related topics. And they already have the ability to do that and want us to undertake that with them, which hopefully soon we're able to start doing some of that.

Matt, from your perspective, what more can you bring to these folks in the future?

So I think Scott hit a few of them. With the ERP change, there was definitely some work that's been sitting on the wayside that everybody's excited to get to. And so a big element for us is AI and helping predict some of the launch models. So I always use the example, if you listen to Spotify or things of that nature, here's the different elements about a song that you like and here's how popular it may be. And so we've got the ability with AI to help with those different products.

And so today that's not done with as much data and as much veracity as it can be done. And that's an area where our AI can really go through and learn a lot and comb through years and years and years of different product launch and all the different features and be able to really improve that initial launch volume. I think another good one is going to be continuing to get smarter on how the team moves their capacity and make some of those decisions.