[00:00:00] Jay Topper: The world of commerce is undergoing a revolution. Today's consumer expects a buying experience that is nothing short of perfection. Your company's digital IQ has quickly become a new standard that drives growth and loyalty. Welcome to Chiefly Digital: the digital leader's guide to modern commerce.
Welcome to Chiefly Digital. Today I have the pleasure of meeting with Prashant Agrawal, the CEO of Impact Analytics, and I'll tell you a little, two little stories. Number one is Impact Analytics is a company that focuses on the allocation, the planning, the buying, and the pricing of inventory, which I think is a holy grail, uh, within retail.
Prashant and Impact Analytics I've known for several years, and they are the real deal. I've come to both appreciate the company. And appreciate the person. There's an old Colin Powell, uh, leadership principle, don't be buffaloed by elitists, where sometimes people can use big words and, and put shiny objects in front of retailers to get them drawn to a certain technology, especially in the AI, ML, uh, arena.
And, uh, Impact Analytics is, is real. We've done deep dives on this technology, and I'm so excited and so pleased to have you with me, Prashant. So, first of all, just thank you for joining.
[00:01:30] Prashant Agrawal: Jay, thank you. I actually just want to stop right here. That was the best introduction I've ever had. Thank you. That was very kind of you.
[00:01:38] Jay Topper: No, you're welcome. And I love your story. So you have an interesting background. You've started up a few companies. You've been back and forth across the world, and you've worked for some big companies. Maybe just, uh, a two or three minute summary of how we got here sitting here today, uh, your life, your life story would be great for the audience to hear.
[00:01:59] Prashant Agrawal: So, uh, I'm Indian American, right? Uh, grew up in Baltimore. Ravens, Orioles fan. Both are gonna make the playoffs. I actually started my career at McKinsey in 1999. I did a JD MBA at Columbia. I started at McKinsey and When I started, I basically had traveled the world for a year and I said to them, look, I'll do anything.
Um, but I want to be in New York City. I just traveled the world. I don't want to do retail. Not interesting to me. And I don't want to do supply chain. And they said, so great, we can give you something. And look, it's a company in New York City. There's insurance, there's banking, there's lots of stuff. The next day they sent me on a plane to Land's End to do supply chain in retail, obviously in Wisconsin.
And I was like, I need to work on my communication skills because something. Here, it's not working for me. But it did start off a 25 year love affair with supply chain, retail, and everything that's in there. You look at it, retail is an art and science, and science sometimes gets neglected. The art is beautiful, right?
The art is beautiful. I moved with McKinsey to India in 2002, and it's part of our story. Most people moved from India to the U. S. I went the other way. I was there for 12 years. I was at McKinsey, then BCG. I actually left McKinsey to work at a hedge fund, start my own hedge fund. I actually ended up joining another one, and it was good.
Uh, it was with the Finance Ministry's son, McKinsey partner. I said, what can go wrong? Hedge funds is great. What can go wrong is a 2008 financial crisis, and that blew up. I actually did a startup in India that was what I'll call quote unquote moonshot, where we were trying to beat Facebook and Twitter for India.
And that's in 2008, and while today it's very obvious that Facebook and Twitter of India are Facebook and Twitter, it wasn't so obvious then. Anyway, there was lessons learned from every Startup failure that I had. I went back to consulting with BCG, led the private equity group for them, but the entrepreneurial bug stuck.
And I said, I want to do something that's a little bit more tangible, something that's closer to what I know and what I've done. And what I did is actually came back to the US, worked at Staples, um, for a year, um, with the chief transformation officer there. And we did a bunch of stuff, uh, over there. I was with another consulting firm as I was doing this.
And . What I saw is that whatever we were doing in 2014 was what we were doing in 1999. The same Excel, the same decision making product. And I said, there's a better way. And in 2015 we started, we do have an onshore offshore model. We have about 60 people here in the us. An incredible amount of X retailers is, as you've seen, Jay, and then an incredible development team in India.
And that's what's led, right? We're now 700 people. We're working with retailers literally across the world we want in New Zealand now. We've won in Italy, we're working in the Middle East, we're working in Asia, and then obviously North America is where we started and have most of our clients.
[00:04:54] Jay Topper: That's fantastic, and I'll summarize as a career retailer myself what I think Impact Analytics does, and then, but then maybe you can highlight what problems you're solving for retailers.
And in my view, in the interactions we've had, which have been numerous with you and your great team, is I've always been very fascinated with merchandising. It's all about the product. And in your space, my understanding is that you focus on, number one, buy the right amount of product at the SKU level.
Number two, Place it in the right location so you have the right amount of stock at the right place for the point of intent. And number three, price it right so you can optimize your margin.
[00:05:41] Prashant Agrawal: It's an interesting thing, right? I'll start off with a very simple thing. I have got the best t shirt ever. It's a great t shirt.
White t shirt, you've got a black one on. We all love t shirts, and I've got a new one, right? And you've got a couple of, a bunch of these that have come out. In the last decade, you start off with a t shirt and then you say, you know what, I need, I've got a medium t shirt, I need large, I need small, XL, extra small, so on and so forth.
Then you're like, I just don't want white, I want black, blue, yellow. And then you say, okay, you know what, I want V neck, crew, I want all these other things. And I'll give you a real life example, which is Bombas. Bombas is a sock company. That is how they started as a company. And yet today, if you look at it, they're selling everything.
And what happens then is this becomes not just an art, it becomes a science. How deep, how broad do I want to go against any of these things? So how much do I want to buy? Now in retail, we have the challenge that what we're buying is actually six to nine months out. So what do I want to buy six to nine months out?
And if you talk to many merchants, and this is how they should be, this jacket is the best jacket ever. I want to get 2 percent more here. And everything is a little bit more than what we have today. But mathematically and statistically, that's generally not true. So if you can help on the buy guide and give a better first draft as to the depth and breadth that you want across size categories, across colors, across any of these things, that's great, right?
Now you've got a better buy. Part of that is also how we're going to sort to each store. How much are we going to put in Baltimore? How much are we going to put in Boston? How much are we going to put in Baton Rouge? A little bit different, different places, different Uh, demographics, I'll give you a cute story from Joanne, which is, as we're trying to build out our assortment and our buy, they do Christmas ornaments, right?
And they do beach Christmas ornaments. Now, they used to think that word beach Christmas ornaments sell the most of. Maybe California, maybe Florida, just, that's where the beach is. It's actually the Midwest, because people are remembering where they were, and they want that as a part of their Christmas tree, like, oh, we went to Florida, or we went to California.
If I live in Florida and California, I don't need the beach, right? That's there. So the data can actually tell you things and inform these decisions as to what the assortment, and then where do I need to allocate. And the last piece is pricing. Pricing in retail has always been. Like this. It's a little bit, uh, you know, up in the air.
Like, is it 10? Is it 11? Is it 12? Pricing is still an art and science, but there's a lot of science involved. Everything around clearance is a science. It's a straight line. Tell me how many weeks I want something to live. There's a markdown associated, a clearance rate associated with that. Base pricing.
How much do I start with? We work with furniture companies and you buy a sofa, But then do you want leather? Do you want this texture? I want the USB, right? You can now put USBs in this. I want it to recline. How much do I charge for each of these upgrades? There's just so much around this that's science that retailers kind of, because it's very car y, just kind of do up in the air and continue what we call same as last year.
Sally, why are we doing it? Same as last year, right? That's what we did last year. And what we're trying to do across the board is that. The other one, and I'll just, is automate. There's so much Excel paralysis that we do in retail, where we're doing this stuff in Excel, and we just don't need to be doing that anymore.
There should be so much more automation, and allowing our people to think. Allow them to think, do the things that they really want to do, that they enjoy, but they're stuck sitting there, typing away, doing this. You just don't need to do that anymore.
[00:09:32] Jay Topper: So, one question I want to ask, and to go one level deeper, but we don't have to go super deep, but when you think of buying product, allocating it, pricing it, how does AI and machine learning come into play for your platform?
And we talked last time about how over time that gets smarter and smarter, so just give a little bit of a peek underneath the hood. on one or more of those capabilities that you provide retailers?
[00:10:01] Prashant Agrawal: No, so it's a great question. A couple of quick things on this. One, we are always forecasting, right? As people, we're always trying to see what's going to happen tomorrow.
And we use yesterday, right? That is the basis, generally, of what you do and what's the data out there. Now, in retail, before COVID, we were basically just using some kind of time series. Last year was, X, and we're going to go up a little or down a little. Now, the challenge with that is that there is so many SKUs, so many stores, potentially, you kind of just peanut butter spread it, right?
It's 2 percent across all shirts, or it's 2 percent across all this or that. With machine learning, what you're actually doing is figuring out what a store SKU or a SKU has been doing over the last year or two years. And understanding what that velocity has been over the last 12 months, last 8 months, last 2 weeks, and then getting the best fit for the next 2 weeks, 4 weeks, or, uh, more.
Now, when you're doing the buy, you're guessing 9 months into the, to the wind, right? That is always a difficult decision no matter what you're doing. What AI ML forecasting can do is give you a better first draft. So instead of actually having to manipulate this in Excel in any way, You can actually get a good first draft that is at a skew level and then hone your intuition, your experience around that.
Now, what else does it do? There's a lot of data out there. There's a lot of data sets out there. Now, we know Christmas moves. Christmas moves. I mean, well, it's the 25th, but Thanksgiving and how many days you have after Black Friday, that moves. Whether it's three weeks or four weeks has a huge impact. On sales, you can actually use past data to understand what that impact is.
We work with one retailer where, okay, Halloween, Halloween is always October 31st. But if it's Saturday versus Tuesday, the number of women's costumes that are sold is different. It's always more on the Saturday than a Tuesday. So you can use data and understand what that delta will be so that you can order better next year and actually get the product to the right place.
Well, that's good. But there's more. There's school openings and closings, which differ across the country. They open earlier in the West, they open later in the East. And that has your children's clothing store, you know that with school openings, you're going to sell more. So if you have that opening and closing, you can do that.
And you can get signals from different areas to put in to the later areas. State events, state events, move things. Um, spring training is big in Florida where you are, but it's the biggest event in Arizona. Most tourists, you're going to get more goods. Mardi Gras, number one event in New Orleans. Actually, it's the number one event in Alabama.
So if you actually have stores in Alabama, you have a surge of tourists. I never knew that, right? I never knew that Mardi Gras was in Alabama, but you can keep adding these data sets and at this point you just get more and more accuracy around it. So that helps you all the way from the buy to the allocation and understanding where these are at.
And then pricing, right? What's happening? And, you know, we're working with one retailer who has the same price. Now this is a little bit on the QSR side, but the same price in California as you do. And I'll go to Alabama. And you know the cost is the difference, right? You can't have the same price. You can get better localized pricing around this stuff.
And it's just impossible. It's just impossible to do in Excel. It's impossible to do with current systems.
[00:13:33] Jay Topper: Yeah, I've seen and worked with some of those Excel sheets before, and they're massive. And usually only one person really fully understands them as well. And so, you also have a little bit of a chain of custody challenge when you're trying to make changes and recommendations.
One thing I've, uh, in today's world where, where, you know, speedy, uh, ROI and capital is super protected, and, and the management teams and the boards are looking for ROI back from their investment. When you engage with a retailer, uh, on pricing or allocation or buying, what can a retailer usually expect?
With regards to getting a return on the investment they make with you.
[00:14:12] Prashant Agrawal: The payback period is very quick. I was at Boston Consulting Group. I co led the PE and corporate finance practice and private equity really likes pricing because you can change pricing the next day with pricing. The payback is three months, right?
You're paying it back. It's actually money into the coffers for a long time because you're adding science there. Allocation is also relatively quick until the next four to six months, because you're getting the right product, the right size to the right place. Buying is a little bit more long term and it has the best impact.
But it's, again, 9 to 12 months away. So, the impact is larger, but so is the payback period because you actually have to take some time to get it right. The thing is, and Gartner has said that the number one driver of profitability for retailers in the next five years is AIM. It is going to be that, and everybody's going to need it.
And to your point, even with a recession looming over us, a soft landing, whatever it is, we know that 2025, 2026 are going to be slightly tougher. The thing that retailers can do to help cushion that blow is use systems either like ours or, well, nobody else, but, or something else out there on AI and ML, uh, to help.
[00:15:27] Jay Topper: Yeah, and I can tell you, I looked at, uh, plenty of these in, in my time having the, the fortune of, of overseeing technology and the supply chain in, in the last several companies I worked for and, It just became so clear that even if you can make this 50 percent better than what it was, than what it is today, and then the next year, you know, then you can increment your way to perfection, but even the big swaths, that is where the ROI, where I see it clearly coming into play.
Not only being able to sell more, uh, having the right product at the right time, but then on the gross margin level, not having to sell it off or ship it to another store or back to the DC. Which every single touch of that inventory impacts gross margin. So it's a, it's, it's just really a great gross margin and revenue story.
And when you go into a client and we, even when, uh, the companies I've worked for, when we looked at your company, is, are people more focused on the upside in the revenue or are they zoned in on the gross margin?
[00:16:28] Prashant Agrawal: I think, you know, it's interesting, right? We in retail had such a good 22 and 23 to a good part.
Because of the COVID bump, right? The government threw money at us. We couldn't spend. So, and revenue really grew and then the bottom fell out, right? I think people are hopeful for revenue, but anything you can do to protect margin, increase margin, people are very excited about right now. That's just the macro part.
[00:16:58] Jay Topper: Gross margin is king for sure. And then, uh, interesting before we get to a couple of closing questions, you have a tie with the, uh, Columbia business school. So, tell us a little bit about that. So, I,
[00:17:11] Prashant Agrawal: I went to Columbia Business School. I did a JD MBA from Columbia. And last year, I actually taught, uh, AI and advanced analytics and retail to the second year MBAs.
And it was fantastic. It was fantastic. And hopefully we can get you up there, uh, this year, Jay. We had, uh, guest speakers. We had the CEO of Saks. We had the CTO of Ralph Lauren. But it was a three hour course. It was. Uh, for nine weeks or so, and it was oversubscribed, and I think the important thing here is not because it was oversubscribed because of me, I wish it was, they didn't know who I was, but the amount of interest that is there for AI, Gen AI among them.
What, J, no offense, the younger generation, much younger than us, is tremendous. They know this is coming. And instead of like potentially saying, Oh, I'll, I'll do it. They're embracing it. They're going full for it. And because they have to, they know that this is where the workforce is headed. They know where this is, where the world is headed.
It was eyeopening to me. Like. Our company is there, but people have to stay ahead of it because the best talent in the world is going to want to use these systems. They're going to embrace companies that use these systems.
[00:18:22] Jay Topper: Now, I know, and I have four kids of varying ages, and even the one that just recently graduated from Indiana University, you know, from a practical standpoint, day to day, knows as much about AI as, A lot of the people that I meet just in the general workforce because they use it, they embrace it, even if they're not technical, it's just what, what, what can I harness to make my job easier and to make what I do easier and faster?
So there's, there's, they're fearless. And I think there is some fear inside of companies, which leads me to my next question, When you go into a company and, and retail companies can, can cover the gamut of being, you know, modern, innovative, technologically savvy to super conservative in some cases, and not always in a bad way.
But you do get a range of possibilities there. What are the blockers that you usually find when you go into a company when you're touting, you know, the machines and the computers to take some level of responsibility away from the associates? What type of blockers do you usually run into? What type of scrutiny do you get?
So, one, I think that The
[00:19:33] Prashant Agrawal: top, and I mean the CXO level, like you have to have that buy in from the top. It's not going to be bottoms up driven. It's going to be top down. The good news is the rest will buy in if the top drives, drives it. I think that's one. I think the other thing is There is some healthy skepticism out there.
There have been solutions that were there in the noughts, and I'll call it the teens, where people were burnt. People were burnt. They tried it. There was a lot of promise. And I think, you know, I think a lot of companies before did try it, and they were doing their best. This is just the right time, right place.
Ten years earlier, we would have been too early. Ten years later, we would have been too late. This is the time to do it. The technology is there. The people are there. And the processes to enable it are there. So, I think that if you can get over the fear factor and you have buy in from the top, these two things will really help speed it.
That said, I think the third quick thing is, it's not a smooth journey. There are ups and downs. It's, it's for anyone, whether you're embracing this in our merchandising supply chain or marketing or wherever it is. You've got to stick with it, understand there are going to be downs and ups and downs, and just keep moving with pace.
[00:20:52] Jay Topper: Yeah, I find too that, uh, that it's not just the technology that changes with platforms such as yours, but a lot of the business processes change, a lot of the roles and responsibilities in an organization change. It actually gives opportunity for people, uh, to, to have, you know, to use advanced merchandising skills and let some of the, the rote learning and the rote recommendations come from your platform.
So even though there are some people that have fear, in the long run, what I've seen is that it breeds opportunity. Well said.
[00:21:25] Prashant Agrawal: You have people that just all of a sudden free up time in their day to do strategy, to think, do things they wanted to do.
[00:21:32] Jay Topper: Now, this, uh, next to last question is. If you're going in to show your platform, uh, to a company, there's, there's two constituents that are interested in that and probably more.
You have one, you have the users that, that look at the interface and, you know, like to see how it works, how it looks, how it feels. And then you have the technicians, the data scientists, I would even throw data in there, the engineers, they want to make sure that it's real technology and not just something, a great UX that was put on top of something that's maybe not real.
Where do you find those decisions being made inside of companies right now? Is it equally balanced? Is it starting to shift one way or another? Where do you see that? Where do you see the rub with that?
[00:22:20] Prashant Agrawal: So that's a great question. I think that to a good degree where we see it evenly balanced is probably where you see the most success.
If you see people too focused on the data technology and the data science, then they're not focused on the user experience, which is huge for adoption. If they're only focused on the user adoption and user experience, and they don't focus on the science and the technology to make sure it fits into their framework, that the data and the science is actually helping them move forward, then it also doesn't move because then people are also questioning, oh, is this helping or not?
So I think where it's equal, And they're balanced, right? Ying and Yang, uh, in some ways. That's where it works best.
[00:23:02] Jay Topper: Yeah, I find that, uh, if a company over indexes on the business making a decision, you really have to push technology into the process. And if a company overly pushes on Tech leading the decision.
You have to really lean into the business side. The biggest mistake I've made, and I'm not going to name the company here, but when somebody says, Hey, what's the biggest mistake you ever made? I chose a platform at a company where I didn't do enough technical due diligence on the architecture and, and actually how it was built.
And so once I picked it, all the front end stuff was great, but boy, it just never quite got off the ground because of the back end complexities. And so. It was just a real lesson learned to me that that balanced approach you're talking about is, is super, super critical. And to be honest, there's a lot of companies out there that tout AI that don't really have it.
And so you have to be on the watch for that. As a CIO, You have to have a little bit of natural skepticism, uh, and, and, and, and, and get into that. The parts that you do fear the most and are most ignorant about to bring that comfort. So I'm a merchant, or I'm a product person leading a process inside of a company.
So last question for you, and I'm getting ready to explore, open up this whole can of of allocation, buying, and pricing. And I either am using a whole bunch of Excel sheets, you know, multi tabbed, huge pivot table messes, or I've got a big old legacy technology inside that just isn't advancing. So I'm going to start this process.
If you were to give a recommendation to the people that are actually involved in this process, What counsel would you give them to, to embark on a journey of, of solving some of those problems that, that you, that your company does in fact solve?
[00:24:53] Prashant Agrawal: I, I think that one, perfection is the enemy of success. It is not going to be perfect no matter what you do.
And I think over and over, and you were a good example of this. Go with the data you have. You go to war with the army, you have Donald Rumsfeld. I mean That's what he said and it's true, right? I mean, I wish you could pick the time for some of this stuff, but you go to war with what you have. Go to war with the data you have.
Release these things out in the open. Let them be used. Let people be like, Oh my god, this isn't working, that isn't working. Then you'll fix it. Every time you think, oh, I'm going to fix everything and then release it. That's just, it's not how it's done. And I think you're a great example of this, right?
Let's try it. Let's test it. Let's get it going and keep improving it. We put this system in, you know, into Tapestry's credit and the CIO and the team there. In seven months, across the world, two brands, three channels, um, and three continents. And they said, look, let's just go. And at the end, was it perfect at the end?
No. But was it great? Much better, right? They put us in the quarterly report saying, hey, look, this is materially impacting financial performance in a good way. But that's what you gotta do,
[00:26:06] Jay Topper: just go out
[00:26:07] Prashant Agrawal: and do it.
[00:26:08] Jay Topper: Yeah, in today's world, you gotta ship it. Get it out there, beat on it, and then make it better.
And increment yourself to greatness. And in this area, I still maintain that there is so much opportunity that you can get. Partway there and still see material, material positive impact on your, on your gross margin in a product like this. So anyway, well, I loved having you on. I'm sorry it took two days to go across, but it's always good to talk to you.
I'm always going to be an advocate of your company. I bring it up at least once a week and it was a real joy having you on and thanks, thanks for joining.
[00:26:42] Prashant Agrawal: Thank you, Jay. Thank you so much.
[00:26:47] Jay Topper: Well, that wraps up another episode of Chiefly Digital. And I really enjoyed talking to Prashant. Uh, impact analytics. Uh, you heard it here first. If you haven't heard it before, they specialize in the buying, allocating, and pricing of inventory. And I think in my 27 years as a see something in retail, I'm not sure I've seen an area as ripe for rapid ROI as I have, uh, the merchandising tools of the type that impact analytic sales.
A couple tips and tricks that came out from this that I, that I also embrace and that Prashant brought up is, one is, When you're getting ready to make a selection of a big platform like that, you need to involve your merchants and your technology team. You don't want to over index too much on one or the other, because sometimes the usage of it is stronger than the technology, and sometimes the technology can be weaker, but the UI looks really fantastic.
So you really have to go in with a balanced approach. And, like most software platforms this day, it's a progress versus perfection. If you can get 50 percent of your ROI in three months, it's better than waiting two years to do some big bang rollout. And then the third point is, you need some top down buy in to this, because AI can be a little bit scary for some.
For most people that embrace this early, in fact, probably all people that embrace it early, there is opportunity. So anyway, thank you for joining Chiefly Digital. You can find us on YouTube, Spotify, and Apple. See you next episode.
Jay Topper 
31M 38S