How can retailers make smarter pricing and markdown decisions during the most critical season of the year?
In this episode, John Gilbo, Sales Director for North America, Australia and New Zealand at Quicklizard, and Matt Schiffman, EVP Commercial AI at Quicklizard, explore how AI is changing holiday pricing, from managing inventory and markdowns to making faster, smarter decisions across thousands of products.
Transcript
John Gilbo
Welcome back to the Price Check podcast. You’ve got Matt and John today leading the way. Matt Schiffman, why don’t you do a quick intro on you and then I’ll jump in with a little bit about me if you don’t mind.
Matt Schiffman
Sure, happy to. Matt Schiffman. I’m the EVP of commercial AI here at Quicklizard Lizard. I spent about a decade at BCG X, the the part of BCG building AI-powered pricing tools, and then have moved over to Quicklizard, helping really expand the aperture of what we can do with AI and with our customers.
John Gilbo
Awesome.
Thanks, Matt. And and a little bio for me, a little background. My background has primarily been on the retail sides, pricing, merchandising, running pricing teams and strategy, leveraging a lot of a lot of data and a lot of tools to you know drive decision making. And I’m I’m you know I’ve been out of it a little bit, not too long. So I’m always always want know the current trends and where we’re headed. So I’m hoping to find out a little bit more on that today. So let’s get into it if it’s all right.
Matt Schiffman
Sounds good?
John Gilbo
looking at this time of year, I’m always I’m always thinking about markdowns and how that that plays out. They’re they’re interesting to me right now. What can you share heading into this this holiday season will be upon us quick? So you know what what comes to mind there for you?
Matt Schiffman
Yeah, I think that what I’m hearing from customers, what I’m seeing in the market
Is a lot of uncertainty both about you know the the future, where where are things going be going, especially when it comes to what’s the inventory position we have right now. a lot of that I I’m hearing is from tariffs of I didn’t know what tariffs were going to be. I stockpiled too much, I stockpile too little, I’m sitting on all this unsold inventory, and you know, my plan doesn’t make any sense. It’s a story we’ve heard before, you know, in 2020, 2021, COVID.
Years there was a similar like rush of interest and markdown uncertainty. You saw that during inflation. We’re seeing it again right now.
John Gilbo
Yeah.
Yeah. Yeah. Do you see this as markdown as being more than pricing when it comes to decision making? And if so, like where do you see that being impacted outside of just that price logic?
Matt Schiffman
Yeah, that’s a great question.
What I found is that markdown is kind of the the symptom that you see the most. There’s a lot of decisions that go into you know, how much stock do you have? Someone had to pick it’s going into your assortment, someone had to pick how much to buy, someone had to pick what to charge, when to promote it. And then the end of the year or at the end of the season, you’re kind of left holding the bag with what’s what’s left and what’s what are we going to do with it. So when it comes to markdown, like figuring out how to do markdown best is making a
bunch of decisions over a long period of time to to see where your inventory is at, where your merchandise is at. and it means that a lot of systems need to talk to each other to really get it right. it’s just such a cross-functional question that is also at the same time incredibly valuable, incredibly easy to see. You know, every dollar you’re slicing off that that that price tag is a dollar that’s not you know flowing back to you.
John Gilbo
Yeah, my experience of the even on the space side, you know, you’d have brick and mortar with with space challenges at the end of a season, new product coming in and and the the decision making around pricing in those stores would be very different than than our e commerce platform because e-commerce didn’t have the same space space challenges. Is that still a common concern with folks today or or is that has that been solved yet, this channel difference, if you will?
Matt Schiffman
no, a hundred percent. The e commerce, you talk about the space concerns. There’s also just you know, unless you’re using electric shelf labels, you’re someone is going and changing the ticket. It’s it’s not, you know, a click, it’s a it’s a real process. So it starts to get much more bespoke.
For how how is this store doing it? How is that store doing it? What’s in their inventory? What’s the warehouse that’s shipping to that store and in that inventory? You’re totally right that many of these questions we’ve been trying to solve for a decade. You’re starting to see at least better data there, but the the practical problems on the ground are exactly the same.
John Gilbo
Knowing that, those upstream effects, when you think about systems and and what good looks like to help understand that, what comes to mind do you, what are you seeing, what are you seeing best in class companies doing in that that when trying to solve that question, I guess.
Matt Schiffman
Yeah, I mean, I came through data science. I always like to talk about data first. that that’s the not sexy answer. If you think of like what’s possible with like Claude or Chat GPT, and you’re talking about a first use case in those spaces, I I’m I’m the one who says like, well, fix your data first. That’s that’s what it’s gonna come down to. Maybe you’re
You know, the buyers don’t tell you why they bought something and you don’t have an intent. You don’t know what season something is tagged to. you have a bunch of holes in your product hierarchy and seven different data sets from different that exist in different systems aren’t talking to each other. That’s the kind of thing that you know.
best in class is is taking that journey to try and really understand, you know, what is this item? Where does it live? How many do you have? How many have you sold? and then what are your competitors doing as well? That that that lens is
has has always been been a big deal. And more recently we’re seeing, you know, more and more people who just go on to like me personally, I was buying a robot vacuum. I’m like, what robot vacuum should I buy in Chat GPT? I’m using it to do the research. I think not necessarily making the purchasing decision there. But now all of a sudden I know I want to buy a a robot vacuum. I know what model I’m thinking about. I’m just asking it, hey, look at like 12 different stores, figure it out. And like maybe something is on Markdown right now that I wouldn’t
Of consider before that it gets surfaced. So that kind of competitive lens and that reactivity is key to really solving the problem these days. And what we’re seeing with the best retailers is that fast response time, is that dynamic consideration of all the all the many different factors.
John Gilbo
Yeah, and it’s it’s in my past I was I was very lucky to to be exposed to some rules based solutions back in as early as like five two thousand five two thousand six started to dabble with what we had for like basic algorithms and and you know some machine learning calculations if you will. how’s that changed? And and what are you seeing today? What are you seeing, you know, that’s new in in the forward view?
Matt Schiffman
Yeah, I think that many of the core algorithms haven’t changed too much. It’s just how they’re applied and how the humans are are involved. So when I say that, like you’re seeing more and more like, you know, we work for Quick Lizard, we have a pricing platform, there’s a lot of decision making that goes, but you’re not gonna be able to look at every single item at every single store and figure out every single markdown.
And before it was really like painting with a broad brush. You were trying to say, hey, this sweater, I want to discount 30% versus 40% for like all my sweaters right now, because of how they’re doing. And you would be running some algorithms suggesting that. But now more and more is really possible to get to that more discrete level that even though the the business context can be similar, you’re still sitting on this pile of inventory.
We’re starting to be able to really operationalize that as here’s the here’s what I need to look at as a merchant. Here are the most important decisions that I need to make. And here’s where where the AI can can offload that. And now the difference is, and here’s why I can trust it to tell me what I need to be worried about and not not focus on on all the long tail, focus on the the decisions that matter.
John Gilbo
Has data gotten better? Has the challenges of all the inputs gotten better? Has has that been solved for? And and you know, in my time it was always messy, it was frustrating, and and I thought I would have solved it by then, but I I haven’t, you know, purely solved it yet. So has that gotten easier or better in any way?
Matt Schiffman
I’m laughing because yes and no. what what I see all the time is there’s more data. There is far more data than there used to be. There are more systems, there’s more information out there. any any given system, you know, there there are these tools I was talking about earlier, easier to to deal with, but that means the picture is just more complicated now. That like
when you’re looking at how often your competitors are pricing 10 times a day from from like four different companies that are pulling in the that data. When you’re looking at you know three different inventory systems and the vendor portal that you’re working with, you’re the the data is exploding. it’s it’s out there, but it it’s definitely that that’s still very much a an issue that’s you know top of mind with retailers these days.
John Gilbo
And as as the AI capabilities help solve for that in any way? Is it making it easy or is there?
Matt Schiffman
Yeah, I think that understanding the your data, visualizing your data, talking to your data is becoming more and more possible, especially for non-technical folks. Like we have an MCP. That sounds like a very foreign thing, but what it really means for our platform is being able to
you know, talk to your data in in human language and ask it a question of like, you know, we’re talking about markdown. It’s like, well, what are the markdown decisions that I need to review? I just heard from the store that, you know, there was a viral viral TikTok about this one product. What should I do about it? and the the AIs are letting you see you know,
construct these more customized things, look at these more these things without requiring like a full BI suite data analytics team to really answer those questions.
John Gilbo
You know, thinking about what we do in in our customers, you know, being the the the the merchants, the corporate pricing teams, those that are leveraging our tools, our solutions. for me, I always think about the end output, right? Going to that customer and going to you know, the cons you, the consumer, buying your your vacuum robot, which I envision is like this big riding motorcycle type thing you’re on, but I know it’s not that exciting.
How has AI changed changed that process for the customer side? That the, you know, you think about the pricing and the markdowns for the very end consumer, about to buy the products from our clients. can you share a little bit there?
Matt Schiffman
Yeah, I think that the
w the searching for products has become much easier. that researching and looking across different vendors makes that kind of competitive lens so much more important that if you know sure you had Google shopping for a long time but
Now you can compare across products in a way that you couldn’t before. It’s not just like skew by skew. what what does that mean for retailers? So if you’re having, if you have your own brand of a product, someone’s looking for the name brand product, now they might surface your your your white label version right next to the the own brand without them having to you know type in the very specific keyword. we’re we’re seeing a lot more of this kind of explosion of
more strategic comparison shopping.
John Gilbo
Love it, yeah.
Matt Schiffman
and yeah, and also just kind of exposing random parts of your long tail that weren’t weren’t relevant before that all of a sudden, you know, your 700th best seller is not if if it’s the cheapest in the market, you might see a big bump in volume right there. Not someone’s not looking at it week to week because it’s you know a tiny product. All of a sudden it triples in sales and you need to think, okay, what do I do right now?
John Gilbo
Yeah, thinking about that that impact back to the retailer, or are there other elements that the consumer side of the AI changes? What does that mean for them, for for our customers, the retailers?
Matt Schiffman
Yeah, I think that reactivity becomes much more important. it’s important to be able to think strategically, act tactically and not get stuck in the operational side. That when you’re, you know, if you have an Excel workflow and it takes a week and three meetings to to change something, you’re you’re not gonna be able to understand, you know, when you see that triples at sales and the long tail for a viral product example, that you
what needs to happen for for a retailer become such a a faster question. But when you when you have these faster questions, when you’re trying to react to the market, when you’re trying to make these decisions, you need to have good guidance already baked in that that you need to be able to have your system think strategically in ways that weren’t really possible before, weren’t weren’t weren’t as relevant before when you know you’re just walking into a store or when you’re just
And you know, you have your your captive customer.
John Gilbo
I love it. And and kind of did dovetail onto that. you know, I think about AI and and you know, a lot of a lot of AI is being thrown out there these days. And I and I think of it as more around problems what problem am I trying to solve? What what decision am I am I trying to make? I don’t I don’t throw AI tools and and other things at it. I try to f think about the problem first and then the and then the
What what do I need to to solve that? And as I think about LLMs, I think about agents, MCP servers you talked about, where where do they fit in? You know, if if some of the stuff’s not new, how do these new enhancements, if you will, how do they help?
Matt Schiffman
Yeah, I mean I could talk for hours about different possibilities and how to get agents involved in the decision-making process. But ultimately, where they’re helping the most, I’m seeing at retailers, is when you have a process where people are making decisions, but they can’t make decisions on everything. So, you know, let let’s let’s take the holiday season. You you have
thousands of items, you know, you’re you’re sitting on thousands of items of stock, you’re trying to figure out what to do about it. and historically what would have happened is you looked at your top fifty products and you thought a lot about them and then you left the the long tail alone. the
the future state has to be is becoming more and more agentic, meaning that long tail is now being considered with with the same amount of scrutiny, with with the right guardrails to make decisions at scale where before it was just, you know, you had a person, now it’s a human in the loop agentic system. the
John Gilbo
Love it. Yeah.
That long tail too is where you can you can reap some decent benefit. And we know we know margins are thin in retail and and that long tail can definitely help out in that side of it. y you know, thinking about that as you think about
some of the systems that we talked about a little bit earlier, asked about you know what systems are, you know, besides pricing and things of that nature. It’s not really an if they’re gonna talk to each other, but when, when they do. what’s that do for for you know the companies we work with? What what that what change will that bring?
Matt Schiffman
Yeah, it’s
Every piece in the chain is another source of value. So when you’re making a better buying decision, you have to carry less inventory. that that’s saving you working capital. you’re marking things down less because you had the right amount to begin with. you’re promoting to really drive traffic and drive people into stores, not just you know, because you’ve trained your customers to only buy on promo.
And then each of those s decisions touch each other. You need to have enough stock to have a successful promo, but you don’t want too much, then you have to clear it all out at the end of the season. So that means you need to figure out what you want to buy. But when you’re choosing what to buy, you have to figure out can I have a balanced assortment? Am I are am I filling up the shelves correctly? so each of these are individual decisions. They’re often handled by separate teams. And that shared context, that shared
language need is really the key to having a a fluent merchant organization and really making sure that everything that your system your your systems think you’re you’re solving the full product lifecycle problem you’re not just solving a single a single step in it at a time
John Gilbo
Matt,
I’m assuming you’re not AI. I’m assuming you’re real. So so how do we keep humans plugged in and and where should they stay plugged into this process?
Matt Schiffman
Yeah, I think that when it comes to AI, the the obvious answer is that humans are humans need to make the high impact decisions, they need to make the strategic decisions, they need to make the long-term decisions. AI is great in the details. AI is great,
at the for that long tail when you give it guidance. You really want a human thinking, here’s what I want to do. This is the category that I’m trying to be competitive in versus this competitor. High high level guardrails and high level objectives. But then the question becomes, you know, you’re you know what you want to do with your KVIs. You know roughly, hey, I wanna I’m I’m competing against Amazon and I want to be about the same price for my long tail, like on average, but I don’t know, two percent here, three percent there.
Let’s let the AI decide. That’s really what it comes down to is humans need to think about the strategy, humans need to think about the systems and the broader context. And then the AI starts to handle more and more of the details.
John Gilbo
Awesome. Last question for you. What’s your what’s your go to karaoke song? Wh what where what do
Matt Schiffman
Oof.
John Gilbo
you hit with?
Matt Schiffman
I am a horrendous singer. I pick something loud and angry like a green day, so at least I can pretend that I can just shout into
John Gilbo
There you go. I love it.
Matt Schiffman
the microphone. What about you, John?
John Gilbo
I love it. maybe, maybe do Don’t Stop Believing by Journey. It’s kind of a classic
and and can get everybody going. So well thank you, Matt. appreciate the insights. This was good for me to learn a little bit more about the agentic side, some of the some of the trends and the tools. So really appreciate the updates and and you know look forward to working with you more on on some of these customer problems we have.
Matt Schiffman
Yeah, thanks so much, John. It was a great conversation.