If you haven’t started forecasting peak season demand already, it’s high time to get started. Andrew Bagwell, the VP of Engineering at Whiplash, is joining us in this episode to tell us how 3PLs are getting their clients’ peak season ready.
Brian Weinstein: Welcome everybody to Sippin’ & Shippin’. I’m your host, Brian Weinstein. We’ll be kicking it here every other Thursday, quench your thirst for an insider’s take to enhance your customer’s experience. Grab your drink of choice, kick back. It’s Sippin’ & Shippin’ time. All right. Welcome everybody. To another episode of Sippin’ & Shippin’. I am your host, Brian Weinstein and keeping me company, and keeping me in line as she does every other week, and probably every day is Caitlin Postel.
Caitlin Postel: Hey Brian, how are you?
Brian Weinstein: What’s happening? I don’t know if I should really be relying on you to keep me in line.
Caitlin Postel: I know I was kind of thrown off by that, but okay. I guess today it’s my turn.
Brian Weinstein: Yeah. Exactly. Exactly. All right. And we have a special guest. Really the brain child, or the brain trust I should say of Whiplash, our VP of engineering, Andrew Bagwell, welcome.
Andrew Bagwell: Hey guys. Hey guys. How are you?
Caitlin Postel: Welcome. Welcome.
Brian Weinstein: Doing well. So Andrew runs our industrial engineering group, along with Sarah Drazetic, our chief engineer and process officer. And Andrew, why don’t you just give us a little bit of background on you and your roles and responsibility.
Andrew Bagwell: Sure. Thanks, Brian. Yeah. So I’ve been here with Whiplash for just over a decade now. Really living the true 3PL life. And so that’s supporting our operational team from a number of different perspectives. So layout, facility design, productivity, reporting, and all those types of items. And so from that perspective, we also take the lead in implementations and bringing in new customers, and making sure that we’ve got plans in place for them coming on board in their most important time of the year, which is peak.
Brian Weinstein: Yeah. Absolutely. Absolutely. And, and you are living and breathing in the weeds, so to speak, where Caitlin and I are sticking to the high level stuff. You get in the weeds under the hood, if you will, and dig in. So I have to say, and I’ve said this before, and I’m going to continue to repeat it, if you own your own facility, if you are with a 3PL, and you are not already engaged in conversations about what your forecasts look like for peak, you are already behind the eight ball. So here it is, we’re in August. You should be having those conversations if you have not already. And so Andrew, maybe you could talk a little bit about what your process is in engaging with the customers to understand where they are, and what they’re forecasting.
Andrew Bagwell: Yeah. Absolutely Brian, and you’re perfectly right there. This is the time and to make sure that we’re starting this process, and getting ahead of it as much as we can. And so the first and most important thing is what are our customers trying to do here? And so that’s two parts. One is what’s the forecast, and what the order magnitude is going to be. We know that things are going to change over the next couple months, but we need to get a sense so that we can start planning for what scale we need to get to during this very important time. And so that’s how many orders, how many units, and what the thought process is there.
Brian Weinstein: Yeah. And I would say it’s pretty safe to say that everybody recognizes that the only thing that we know for sure about a forecast four months out, is that it’s most likely wrong.
Andrew Bagwell: Exactly.
Caitlin Postel: Like our weather forecast. Just guess.
Brian Weinstein: That’s like a day out. So hopefully within a day, we’re pretty accurate, but I know it’s going to change, but at some point you have to at least put pencil to paper. To commit, not to commit, but to at least give directional guidance on what those forecasts are going to be, so we can get into the planning stages. Is that fair?
Andrew Bagwell: Yeah. Absolutely. And at this point things have changed. Peak used to just be, especially in the e-commerce side of the world. Black Friday, cyber Monday, but as we know now, strategies have changed. So some folks are still drilling in on those key days and trying to drive there, but some folks are spreading it out and looking to hit earlier in the month. So I’m thinking of it as a cyber seven, rather than a four day period. And if we can start to understand how those things are aligning from our customer strategy, and is this just a straight discount across the board? Is this a gift with purchase type scenario, or are there certain product types that are going to be trying to drive through? We can start to begin to formulate plans, not just from a pure number standpoint, but from a facility design and strategy standpoint.
Brian Weinstein: Yeah. And you know what I’m actually, that’s a little off topic, what I was thinking about, because it really hadn’t crossed my mind, but you do have to take into consideration. So we obviously, I know your team is focused on labor standards and units per hour, that you’re going to be able to pick and pack. If your idea is to add a gift with purchase, or something special into the equation, as we’re coming up towards peak, that’s something that the teams, both the client side, or that at least their design side and the operations team really have to collaborate on. Is that right?
Andrew Bagwell: Exactly. And that’s somewhere where we can look to make it a true win-win situation. And so if it’s a single same gift with every purchase, that’s something we can be placing above our pack stations. And making sure that we’ve got the appropriate lead time to be able to make sure we’ve got whatever physical thing we need to keep that product right there, or kind of carve out an appropriate area nearby. So that it’s a nice, easy pick. And so we can go ahead and fulfill those, without putting tremendous stress on the operation.
Brian Weinstein: Right. And so when you’re looking at that, okay. And let’s just say for arguments sake, that a customer does not have anything special, they’re going to add. What does your team do with those forecasts to understand volumes, what it’s going to take to get that down in a period of time, and work through that backlog?
Andrew Bagwell: Yeah. So I mean, it does end up being a fairly straightforward math problem. Which is if we understand the forecasts that we have, what the plan is, and we know what our current capabilities are, it’s fairly straightforward and a rigid process that we like to use, to really just put that into a model. And so we’re looking at daily activity through the whatever peak period that we’re talking about, where here are the orders that we’re expecting to drop. And then based on our current set of capacity, which is driven by our number of pack stations, our number of pick carts, our number of robots in the facility, what are we going to be able to accomplish?
Whether that’s on an eight hour shift, a 10 hour shift, two shifts and through the weekends. So that we can lay out very plainly, if nothing changed at all, this is how we would perform through peak. This one is when we would be caught up shipping same day orders, and just look at that. And whether that’s something that’s an acceptable situation, or more likely, because it’s peak, we’re going to have to make some modifications here. We’re going to have to tweak some of those variables to provide the experience that our customer’s customer is ultimately looking for here.
Caitlin Postel: Yeah. It’s interesting. Because, so to go in the other direction of just when I think of peak, I just think of massive amounts of volume. But you brought up a good point about maybe strung away from the norm. So maybe there is a gift of purchase. Maybe there is something that you’re not… They’re going to introduce. The client’s going to introduce, that’s not part of the norm, part of that historical data that you can dive into and make your assessment from there. What do you do in that case to be able to properly plan, and how much notice do you need around that?
Andrew Bagwell: Well, as much as we possibly can. And it depends on the scale. But what we do is try to incorporate that into the model. And so a classic example might be that somebody’s got a traditional order profile of three units per order. And so our customers often talk in terms of orders because that’s sale. And so 1000 orders might be 3000 pieces that we need to go pick. Well, if it’s gift with purchase and it’s something that’s not just thrown into the poly mailer. It’s not something that’s thrown into the box, but something that has to go be picked, all of a sudden, those 1000 goes from 3000 units to 4,000 units, which is a pretty appreciable difference.
But if we’re able to get ahead of that, so one, we plan with that, but if we know that’s something we can set up nicely, we might be able to increase our productivity rates associated with picking there. And so that’s something that we like to just talk through. Make sure we understand, okay, is this 50 different skews that are being picked from. So normally we’re picking X units per hour. Well, maybe that’s a 10% productivity increase based on that, or some other percentage so that we can build that into the model and understand exactly how these things are being impacted.
Caitlin Postel: So regardless of scope, just that communication factor?
Andrew Postel: That’s beyond the key. And so what we like to do is just lay this in front of everybody. This is a very collaborative discussion. We’re not hiding these models, they’re designed to be how we communicate. And once we actually get into the peak period, it’s something we measure against. And so we’re looking at daily performance, because we know those numbers are going to come in slightly different. Well, did we get what we said we were going to do done, which we should be, because that’s how we plan to it. And then if forecasts are coming in over what was anticipated, what does that look like now? And what other levers can we pull, and what other variables can we tweak to increased capacity.
Brian Weinstein: And my favorite part of what you said, is it’s all simple math. It’s not emotional math, it’s not guesswork math. It’s very simple, you model it out, and from that you know, and you have that ability to work through with the customer and say, this is what it looks like. If you get 10,000 orders on black Friday, and here’s what we have. We have X amount of pack stations. We’re running an eight hour shift. We can add a shift, we can do all these things. And what does that look like when you get through Saturday, Sunday, and then get another 15,000 orders on cyber Monday, and what does that look like for a backlog. So when do those types of conversations, when are you really getting in and putting those out in front of the customer for them to see. What time of the year is that typically?
Andrew Bagwell: So hopefully at this point we’ve received at least the initial forecast from our customers, and then over the next few weeks. So through the month of August is really when those conversations are happening and being solidified. Because that’s the sweet spot for timing, because some of these things, if we do want to pull the trigger and make some adjustments, some of these things do have lead times. Whether that’s hardware, which is increasingly a tricky topic in the year 2021, or racking, which is another tricky topic in 2021, or space. Those are things that-
Brian Weinstein: Also another tricky topic in 2021.
Andrew Bagwell: Yeah. I know near and dear to your heart. This is the time to be doing it. September, October, we’re getting extremely close and the entire industry is pinching. And so if we can be getting at least close to where we need to, and understand we need to add eight pack stations. And maybe that gets cut down to six or increases to 10, a little later on. We’re at least able to get the gets started, get those layouts being tweaked and set up appropriately, so we can get all parts moving as they need to.
Brian Weinstein: Right. Right. And that also, to your point, I mean, right now in 2021, IT hardware equipment, and material handling equipment is a little bit trickier to get your hands on. But there are other approaches. So you could theoretically take a single shift five day a week operation, and add a split shift or a second shift to offset that?
Andrew Bagwell: Yeah. Absolutely. And those are usually some of the first levers that we’re looking to pull in this. Is how do we increase capacity without going into a capital expenditure? Which doesn’t, is the last resort through this process. And so, yeah, we’re obviously looking at other shifts, extended shifts and those types of things to be able to increase capacity as much as possible.
Brian Weinstein: Right. Right. And so I know from a commercial perspective, what Caitlin and I find is really been a real help for us as an organization, because we hear it from the customers, is helping them set expectation for their customers, for their clients. The ones that are buying this stuff and waiting for it. They’ve just made a purchase for holidays and they’re expecting it, and allowing them to set expectations.
I know we had a couple of situations with our brands last year where their projections, which were already quite sizeable for, let’s say a black Friday came in at 50 to 100% above their projections. So walk through the process of the adjustments that your team needs to make, and work on with the operator. So we can get out, what needs to happen with the customer so they can communicate those SLA changes to their end consumer?
Andrew Bagwell: Yeah. So I think the first thing is, as we’ve talked even further upstream too, which is to do a bit of sensitivity analysis on some of these models. Which is to say, here’s what we forecasted and built this for, what happens if it’s 20% higher, what happens if it’s 50%, or 100%, or 150, which can get a pretty messy situation, one that everyone likes, and then doesn’t like immediately, but something that we can work through. So then as soon as we start seeing that first they come through and it be significantly higher than the rest. At that point, we revisit our model and say, what, what other levers can we pull.
And so are we already on two shifts? Are we already running as longer shift as we can, and those types of things to make sure that we’re hitting as much of the available capacity as possible. But then we’re able to then look around. We have the benefit of being in large facilities ourselves that have other parts of capacity. And so do we have someone else who maybe isn’t quite hitting the numbers that they thought they were going to be. What other resources can we leverage, are there more mobile type pack stations that can be set up, can we be packing off of iPads to available printers? And things like that we have in the wings available to juice up capacity where we need to across our network.
Caitlin Postel: Is automation ever brought on as a seasonal solution?
Andrew Bagwell: So automation’s a very interesting topic when it comes to peak, because it can help. And so we do for instance, have operations where we use robotics. And that’s one of the reasons we went with that type of solution is because we feel like we can scale it, and grow it towards peak. And so I think those in the industry have looked at, and seen traditionally as some of the classic pieces of automation, your tilt tray sorters, your unit sorters, carton sorters as great for the everyday operations, but can pose up challenge during peak, where you are truly running into capacity limitations there. And so that’s why, especially in this high growth eCommerce world, we’ve tried to look at automated solutions that are able to scale. And so going from X robots to X plus 30 robots during that timeframe to be able to increase that capacity, and then be able to send them on back home after that peak period’s over and eliminate that expense.
Caitlin Postel: Yeah. That makes sense. Whereas with the labor, temp labor, maybe thank you for the season and on your way you could scale up, scale back with the robots. Yeah. That makes a lot of sense. Yeah.
Brian Weinstein: Yeah. But that’s also an interesting point too Andrew, I think with the certain types of automation, like you mentioned, a tilt tray sorter has limitations and is not going to be able to scale, whereas other pieces of automation can, or maybe you can add, I mean, tilt tray is a huge investment in and of itself. But if you go a different direction where you can add multiple induction points into different types of operations, or different types of automation, I mean, that probably makes more sense otherwise you’re going to stunt your capacity.
Andrew Bagwell: Sure. I’m absolutely not saying that you ignore unit sortation just for the sake of peak, because there’s all sorts of different models and opportunities for being able to set that up, and then set up some sort of supplemental process to be able to get through peak. If you think that automation’s going to help for 10 months out of the year to such a level, and be able to support that business, then a strategy should be built in on how do we supplement that for this peak period. But as we look at some of these guys that are truly growing at really high rates, the design criteria, what year of throughput capacity are you shooting for can get a little bit tricky, because in a high growth world, you always need to be able to move to higher levels of throughput without impacting the current operation.
Brian Weinstein: Right. And so in terms of just the overall collaboration, so when you’re building these models, and you’re making those decisions, are we adding, is there capital investment in additional pack stations, or other types of automation? Are you going to split shifts, two shifts? Are you going to go, I think at one point last year we were running 2 12s. So we were basically, we were a 24 hour operation for a period of, I don’t know, 10 days.
Andrew Bagwell: Yeah. Switches, it’s just of course the very last lever that we like to pull. We end up with, true diminishing returns in all senses to the word when we do that. But so that is the plus 50% type scenario that you were referring to earlier.
Brian Weinstein: Right. Well, and assuming for a second that there wasn’t any issues with, labor, which we know there is, but even if we’re able to staff at 100% on 2 12, it’s obviously not ideal to your point diminishing returns. You’ve got fatigue that sets in, you’ve got, I’m sure by the time you get to the eighth, ninth, then 12th hours, your UPH starts to get skewed pretty drastically.
Andrew Bagwell: Yeah. Absolutely. It’s not something that we try to plan for it all. And if we were to try to plan for requiring 24 hours of capacity, that is a time that you’d like to look at a third shift to be able to do that. It’s not something, if it’s not something you’re running traditionally, it can be a little trickier to set up and then take down post peak. But that would probably be the more reasonable solution rather than pushing everyone 12s.
Brian Weinstein: Right. Are you seeing, and I’m not sure how much you’ve gotten in terms of the forecast yet. Are you seeing a pull forward that’s heavier than it was traditionally? I mean, is that volume in those sales? Have you seen on those weekly forecasts that is starting earlier and earlier November?
Andrew Bagwell: Yeah. We have certainly seen starting probably last year, and then continuing on this year, definitely a pull forward. I think a large number of folks are referring it to it as the cyber seven now, but then looking to also hit dates earlier in the month in November to grab more eyes. And so with that, and an understanding of those types of schedules, it allows us to look within our full facility, and look at our labor, and look at our resources in a slightly different standpoint. That being said, those four days are around black Friday through cyber Monday, even though things are being pulled forward are increasing year over year in a significant capacity.
Brian Weinstein: Right. Yeah. And not going away.
Andrew Bagwell: Yes.
Brian Weinstein: And not going away. It’s become, it’s become the holiday tradition now.
Andrew Bagwell: Indeed.
Brian Weinstein: Yes. Exactly. Which is interesting, because you don’t really have to wait till the last minute anymore. I mean, you can order stuff now, and have it for the holidays, but we don’t actually do that.
Andrew Bagwell: No. Not so much.
Brian Weinstein: So just, as we leave everybody, what would be your top focus points that you would stress to people that are listening to really focus on as they’re going through this forecasting process?
Andrew Bagwell: So I think the most important thing is to start early, and to start with a broader strategy discussion. It’s great to get the numbers, and the expectation as quickly as possible. But it’s almost as important, not quite, but almost to understand what the approach is, and what the thought process is behind there. Because although the hard numbers that drive pack station numbers, pick carts, robot counts, and things like that do need to be resolved, are extremely important.
Knowing what the strategy is with products, and what’s going to be on sale is equally as important. Because understanding how it’s going to impact our process, do we need to have a different type of active pick location for this time period, to make sure that when we do replenishment from reserve to active, we don’t run into a significant bottleneck. Do we need to be setting up some additional locations closer to those pack station? Those types of things are probably equally as important in setting up the facility for success, and making sure that kind of peak season relationship continues to be a success.
Brian Weinstein: Awesome. Andrew Bagwell, rockstar, VP of engineering for Whiplash. Appreciate you coming on.
Andrew Bagwell: Great. Thanks a lot for having me. Always exciting to talk about peak.
Brian Weinstein: Yeah. Peak has become the thing we talk about year round, even though it’s only four days long.
Andrew Bagwell: That’s right. Not anymore. Seven days or longer.
Brian Weinstein: Seven days or longer. Exactly. All right. Appreciate it. Andrew Bagwell. Thank you very much for coming on. Caitlin as always will walk us out of here. Appreciate everybody tuning in, and we will speak to you again in two weeks.
Caitlin Postel: Thank you Andrew, and thank you everyone for tuning in. Check us out at sippinandshippin.com or on your favorite podcast platform. We’ll see you every other Thursday here at sippinandshippin.com. Thank you everybody.
Brian Weinstein: Thank you. Enjoy. Take care.