Pantry AI

Jordan Buckner: [00:00:00] Building a CPG brand is a bit like playing a game of which came first, the chicken or the egg. You need inventory to feel confident making sales, but you also need sales to have the money to pay for the inventory. This is especially true with products that have a short shelf life, like a lot of food and beverage products listening in.

So what's the answer to utilizing proper forecasting tools and how much product you should really make? So I've invited on Ian Leaman, who's the co founder and CEO of Pantry to discuss how to improve your forecasting. Ian, welcome. 

Ian Leaman: Thanks for having me, Jordan. 

Jordan Buckner: So, tell me a little bit about the problems that founders are facing when it comes to accurately forecasting their demand.

Ian Leaman: Yeah, that's a really good question. I think, especially in this environment, like, where VC money is a little tight, and like, companies are trying to grow profitably rather than trying to grow at all costs where you you could increase your sales by spending a million dollars on Instagram ads.

Yeah. Forecasting [00:01:00] demand has become more important as people want to basically increase the margins on all of their business on all of their production. It's really interesting. I think ultimately like sales and operations planning is like the root of how we've seen a lot of brands trying to grow quickly.

So, you could get like a million dollar deal with Whole Foods, but like that could actually end up bankrupting a small CPG brand. If you're not like planning your inventory correctly. So forecasting super hard and super important. Ultimately, as it like relates to cash flow and, just. Making sure that like, you're not left flat footed.

So it's been really interesting to like dive into the space. 

Jordan Buckner: Yeah. You know, I've been just looking at my own experiences running my brand Teasquares and like you mentioned, like when you actually get a new customer, it's really challenging because sometimes, you know, how much they'll place for an initial order, but then the reorders after that are completely variable.

And even like, if you're in say Whole Foods and you're launching in Target, right? Like their reorder rates. And are going to be completely different. And so it leads to a lot of just confusion in terms of [00:02:00] like, how much you can rely on historical data and how you can use that to forecast the future. And I've seen just a lot of kind of put your finger in the air and see where the wind is blowing kind of mentality of kind of making up a as you go.

Ian Leaman: Totally. I've seen so many times, like you get a really big initial order and then like maybe even a really big second order and then all of a sudden they drop off for six months and that can like cause panic because if you're running production on the size of those first orders. You've just sunk all of your working capital into inventory that could expire.

So I've seen this like across like every different category as we've kind of dug into this problem. 

Jordan Buckner: Yeah. And as you mentioned, right, like the ability to just have like a large lump sum of cash is even harder to come by these days with investment dollars. Drying up and so you don't want to be sitting on a ton of inventory, but you also don't want to have no inventory because of opportunities come up and it's like, oh, well, we'll have to wait 3 months to produce.

Right? Because let's be real. A lot of the manufacturing cycles are maybe every 3 months, 6 months, sometimes once a year for certain products. And so you really have to be kind [00:03:00] of, you know, have a really strong strategy for your forecasting, your planning. 

Ian Leaman: Totally, totally. I find even just like bringing in some like basic data into the conversation can be helpful.

It's like we're working with 12 brands right now and like beta and even like just wrangling data from either like QuickBooks since 7 or like NetSuite if you're using it is Sometimes people like aren't crunching some basic numbers like distributor velocity that I think could be quite valuable for making some of the decisions that you're talking through.

Jordan Buckner: Yeah. So let's talk through that a little bit. Like what data , are you finding is like really helpful to have? And then how are you thinking about analyzing that data to help founders make better decisions? 

Ian Leaman: Yeah if you think about, like, what a brand particularly in retail is trying to do, like, you ultimately want, like, in store velocity.

So there's a lot of ways to get to that number. But I find , the best companies we're seeing basically have, like, a weekly SNOP meeting where they're looking at their in store velocity by their major accounts, as well as just like their sales targets. And then you can kind of work backwards into like the inventory planning.

So like when I say SNOP meeting, [00:04:00] like the head of sales and the head of ops should probably and maybe even the head of finance should be kind of like, looking at velocities. And trends like, basically every week to adjust as the plan. Changes in addition to, like, the trends, like, what you did last year can be very helpful as just like a benchmark to know, like, if you're like, what you did last year, combined with your seasonality combined with.

Any upcoming events that you might be doing is like a pretty good way to finger in the wind, understand, like, the direction that your business is going and accurately produce the right amount of inventory. 

Jordan Buckner: Yeah, I think you kind of hit on something that's really underlying that data, as you mentioned, are the variables that go into planning.

Right. So maybe there's a sense of a baseline velocity that you'll learn once you're in the market for each retailer and it can be different. There's probably like regionality because every market can buy products a little differently. There's seasonality, depending on the weather. If you're a frozen ice cream, you'll sell better in the summer or the winter, depending on which part of the country that you're in.

And then what other kind of factors are you looking at? Like are those all of them? Are there any more? 

Ian Leaman: Oh, no, so it's actually really interesting. I think , there's [00:05:00] all the factors you can look at. And then there's like, how do you get to like, a base level of accuracy? It's like, when we go out, a lot of people are asking if we use like point of sale data from stores, like , the sell through rates and if you have like 2 3 years historical data, you can get pretty decent accuracy by just looking at, like, the P. O. S. because you can start to estimate the P. O. S. into distributors.

You can start to estimate a lot of those other factors. And then you kind of nailed it. It's like seasonality which region you're in. Conceptually, like, if a brand wants a ton of work, they could look at things like weather patterns and like census data to understand, like, is this, because like you could be selling into like a distributor or something.

And, but there's like saturation points. It's like, if you're already in all the whole foods in a specific region, then like, there's not much more to go in that area. So understanding like your saturation is maybe like the biggest variable. That you didn't mention. Otherwise it's basic seasonality, history, growth rates, upcoming events.

Jordan Buckner: So, You know, maybe it was pranking some of the secret sauce of pantry, but how do you think about like combining all those things to kind of forecast , the sales? Are you looking at just like a weighted average across the different [00:06:00] variables? Kind of , what are you testing out in terms of like finding.

Accurate models , of looking at , the demand. 

Ian Leaman: Yeah, , that's a really good point. I mean, for us, we're constantly like tweaking it. So, diving under the hood a little bit. We basically run a bunch of different algorithms. And , if you have a lot of experience with supply chain, you probably have learned similar things, but like, basically we run a bunch of different algorithms and then choose the best performing one on a SKU by SKU basis.

But we're constantly looking at like new data points. As well to, like, kind of weave into the algorithm to try and understand, , like, maybe it's history. Maybe seasonality is important. Like, maybe saturation. Maybe it literally is census data. So, like, trying to understand what can get us more accuracy in a forecast is constantly working on it.

Jordan Buckner: Yeah, I think that's really key, right? Because, like, I can tell you, most founders listen to this. They're like teams of 1 to. Maybe 10 people. And so like , they don't have someone who's like constantly looking at like, what's the most accurate way of predicting. It's like you put together an Excel spreadsheet once, and that's what you're using for years.

Right. You don't like go back and really edit that just because there's not enough time or understanding of it. And so I think it's really cool that [00:07:00] your team is like, this is the work that you're doing. And so it's a sense of like, constant improvement where other people and other founders, like, just aren't don't have the bandwidth to focus on that.

Amongst , the companies that you've been working with on your beta. What have been the biggest learnings that either you've had or they've had in terms of thinking about demand plan, like , any kind of features that they've wanted or anything that you've learned from that? 

Ian Leaman: Yeah we've learned a lot, but one of the really interesting learnings for me was just understanding that like every department and everyone who like, neat touches the forecast is probably speaking in a different language. So, like, sales might be speaking in like, revenue and or like, like, sellable case units, whereas ops is probably thinking of each is or bottles or something like that. So, having a way for 1 of the things we learned is, like, even though we're providing 1 forecast, having a way to quickly switch between a bunch of different views.

So that people can kind of like speak in the language that they are used to has been really important and also can just help like that cross team collaboration a little 

Jordan Buckner: bit. Yeah, I think that's interesting, right? Because as the decision maker moves from the founder to actually multiple [00:08:00] people can touch into different variables.

That language that people uses is really important and having a streamlined communication. Right? And you know, a long time listeners of the podcast, we did an episode earlier this year about a company that was selling coffee into KeHe and KeHe, the rep told them of certain volume and they use the word units, but to them, units make cases versus to the company units meant bags of coffee.

And so they said like, Oh, we need something like. You know, 10, 000 units. They're like, cool, 10, 000 bags of coffee, but their case size was like six. And so then their sales were like, Oh no, we actually make cases. So it literally went up six times and it was really hard for them to process that. And they're working like crazy.

And there's a whole nother side of that story. So if you want to listen to that podcast, go back and find it. But, you know, from personal experience as well, running Teasquares, I ran the sales and co founder ran. And it was interesting because even as we were thinking about growth, right? Like with my sales hat, I was like, in finance, I knew that we needed to [00:09:00] see exponential growth in order to get the business to where we wanted it to go, but it was very aspirational and then on the upside, he was looking at like, okay, how do I optimize for what we currently have? And so I remember this number, but he's like, yeah, we're seeing a 11 percent month over month growth rate. I'm like, yeah, but we need to get that to like, 20 or 30%. And. He's like, well, I don't want to make forecasts based on that amount.

If we don't have the sales, we're like, but the sales are coming. And so it was a constant kind of battle between those like sales and operations side. So I'm curious to learn, like, what have you seen in terms of that? Like the sales and operations planning , from those clients as well. 

Ian Leaman: Yeah, it's been something, it's really interesting.

Like accuracy versus aspiration is like, I think a constant struggle at all. Like basically like every brand it's like, you may want to double, but like. Whereas that understanding where that growth comes from, like , if you basically produce to that aspiration, like, you can put yourself out of business is like, I'm sure you saw it Teasquares.

I think the way we've started to figure it out is basically, if you understand, like, almost the [00:10:00] revenue segments from, like, where your growth is coming from, I think it makes it a lot easier to have that conversation between, like, head of sales, head of ops. So if you understand, like, the way we think about it, at least for pantry is like, here's your baseline business.

It's , like you don't touch anything like this is how much you're gonna sell through all the, like, all the distribution you already have, all the retailers you already have, but obviously you're putting in a ton of work to like improve that and get new distribution points, run promotions, like run advertisements.

So if you can kind of stack. Your existing business on top of like all the new business that you want to do and we've run into like some of the more like some of the bigger like 100, 000, 000 dollar more sophisticated brands kind of do it this way as well, where they'll having that segmentation is really valuable because then you can be like we want to double but actually like we can't because like our growth rates only 20%.

So, you know, , we're gonna have to run a ton of sales activities on top of that if we want to double and like, is that possible? I think it's probably a more productive way to have that conversation. 

Jordan Buckner: Yeah, I think it's cool that you had like the multiple views within there as well. So you can like see the data from different perspectives because it allows people to have that common [00:11:00] point of reference and like source of truth in terms of like, okay, this is how much we're all agreeing on that.

We're producing. I think that conversation around like actual numbers is probably just as valuable as the accuracy itself. 

Ian Leaman: It totally, it totally is. And changing because things change. It's like no one can predict the future a hundred percent. It's just like how quickly can you iterate and adjust as you get new information.

Jordan Buckner: As you're thinking about kind of building out , the models, are you also looking in terms of like running different scenarios or are you seeing clients kind of saying like, Oh, well, if we launch in this account, Then our inventory will needs will go up by this amount. kind of looking into that?

Ian Leaman: Yeah, totally. Scenario planning, I think is a pretty requested ask. We basically look at it in terms of like, what's the worst case scenario, the likely case scenario, and then the best case scenario, it's like the confidence interval on the forecasts, but I thought it was really interesting. I like we've crashed a couple of our customers, like weekly planning meetings.

And. The question I heard the most from ops is like, what actually happens if everything goes right? So it's like, not necessarily what happens if things are like kind of smooth sailing. It's like, yeah, what if [00:12:00] we actually land all these accounts that we're talking to? So I think that's where scenario planning becomes really important is like figuring out what you do if you actually get all the business you have in the pipeline.

Jordan Buckner: Yeah, I think what's interesting is as well that I found is just like that. Lag time between when you might need to deliver order versus when you need to start production of it. Because like we talked about earlier, it could be months and so like, have you seen clients talking through those issues in terms of like, okay, how, like when do I even need to produce this stuff if I need it by this date?

Ian Leaman: Oh yeah. I mean , we have customers who have like, I've seen, like, all the way up to, like, a year lead times necessary, which is crazy because it's, like inventory you need to purchase of, like, raw materials that are just gonna be sitting in a warehouse and the amount of cash you need to, like, lay up up front for that is kind of staggering.

I find that they Tend to over order because I think the cost of stocking out is just like so high. But if you can reduce that order by even a little bit, it can be quite valuable to a brand. It's like, that's 100, 000 of working capital that you now have to spend on growing your business [00:13:00] rather than.

Sitting in a warehouse somewhere. 

Jordan Buckner: Yeah. I mean, it's interesting, right? Because you say that like a year lead time to some of our listeners, they might be thinking like, wow, like that's a really long time. Like, why does it take that much? And sometimes it might be even just be like one ingredient. And I was talking to a founder and they were like, yeah, we had to place our orders for 2024 almonds, right?

And it's like, we need to get the contracts. We can get in the price. Like, sure. You can find other stock on like from a wholesaler distributor after that. But they were like, in order to get our best price, we need to like place our orders now for the whole year. And that might limit the rest of , their product that they're ordering as well.

Ian Leaman: Yeah, yeah, . Exactly. And even like more common is like one to three monthly times if you're like manufacturing overseas or something like that. And even that's like nerve wracking. It's like, if you don't have enough, on the flip side, if you don't have enough product and then you need to wait a month for like the container to get here from overseas, like that can be pretty.

Stressful as well, and you're missing revenue the whole time, 

Jordan Buckner: you know, are you finding that listening to these conversations that the founders you're talking to, at least are like, used to using data and decision making, or is that [00:14:00] process and almost like muscle around it still a little new .

Ian Leaman: I'd say like 75 percent new, 25 percent for the more sophisticated, like they, they're doing it, but even then it's like a lot of their time is going into like wrangling it and not as much as going into like using the data in those decisions. Like, almost everyone we talked to is like emailing spreadsheets back and forth to like, who , then like you email the spreadsheet and then you like spent like a whole day crunching all the numbers and like building out all the views you'd want.

And a lot less of the time is going into like, alright, here's the velocities, here's our targets, like, here are all the different, like, slices we need. So I haven't really seen anyone actually totally using the data in the way at least I'm used to from like the SAS world, but there are varying degrees of it for sure.

Jordan Buckner: Yeah, you know, like, I was just thinking back because if I'm being honest, I was probably searching for data that supported my viewpoint and ignoring the data points that did it. And so I think I watched it. I know that happens a lot within companies, right? Because they're like, oh, I have this idea. And right.

[00:15:00] The thing about days, you can always find data in a certain view or a certain time frame that supports , your case. But I think what's really interesting in that the. Businesses going more and more is having kind of a common set , of data truths that you're all kind of basing decisions around.

And sometimes those would be like a positive indicator. Sometimes there'll be negative indicators, but it's very. Reminiscing like, this is what's happening with our business and we shouldn't shy away from it. We should address it head on. 

Ian Leaman: Yeah, exactly. You can only ignore the truth for so long. Oh, you have too much inventory sitting in a warehouse somewhere that's going bad.

Jordan Buckner: I love that. I wanted to take a sec too and just learn a little bit more like why, what led you to start Pantry and to help you kind of figure out the, the issues that founders are dealing with on the demand and forecasting 

Ian Leaman: side? Yeah, it's a really good question. I think , it's been quite a long journey.

I actually did a startup before pantry. We were doing like predictive lead scoring, although we started with, as a startups are constant pivoting, we started as like a database of researcher profiles [00:16:00] that like we sold to biotech companies to figure out like who was using their stuff. And we kind of transitioned through the years into like predictive lead scoring.

And then we ended up getting acquired by HubSpot. And I always knew I wanted to do another startup. But, like, you kind of have to have, like, some unique insight into the world to, like, do it. So, for the my time at HubSpot, I was basically just, like, keeping an open mind. It was like, kind of just like looking all over the place and I've always loved food.

Like I used to sell cookies on the playground. Like I love to cook and honestly, like I like people who like food and it's a huge industry. So I was interested in that like sector. 

Jordan Buckner: Yeah, I just stopped you right there real quick. Sorry to interrupt. But like, I love a good founder story where they used to sell like food or candy at school.

That was the same way. Like I was like, my mom was a baker. And so she would like make cookies and I would sell them at school too. That's so cool. No, nevermind. Like having them in my backpack. So anyway, I just love that. 

Ian Leaman: Yeah. That's so funny. My parents literally thought I was selling drugs. Cause I would come home with like wads of like 1 bills and stuff like that.

Jordan Buckner: So anyway, you were like thinking, looking at problems in the food space. 

Ian Leaman: Yeah. And honestly, it's crazy. As I dug [00:17:00] into it, it was just like a lot of the software in this industry is like out of the nineties. Like, I think people honestly, like running a CPG brand is so freaking hard. Like you get hit with like a million leaky buckets and like, there's not really good solutions out there in my opinion, which is kind of like what led me to this.

Like people just have to throw people at the problem, which is very expensive, like. Automation isn't kind of present in the same way that I've seen in other industries and like the SAS world, which I'm used to from HubSpot and my startup before that. So, like we saw a lot of problems and we basically just interviewed hundreds of people and that kind of like, we've iterated a bunch on the idea, but, like, it seems that getting control, like running a data driven sales and ops process is Far and away, like the best way to like unlock cash to like grow your business faster and like made data driven decisions.

So that's where we landed, but I mean, , we've explored like every idea under the sun, like we were going to do inventory management for a second. We're going to do, like more of the shipping logistics stuff. So. But I think we're just trying to keep an open mind for , where are the problems that people are really struggling with?

Jordan Buckner: Well, I'm a big supporter of people and companies who are solving problems [00:18:00] for emerging CPG brands, because it's definitely a challenge. And every business is challenging, but in particular in this space, a lot of the managers are like. Having inventory, having expiring inventory and relatively low margins, right?

Like it's really tough, so you have to be able to get it right. So I think that that's awesome. When is Pantry playing to launch and how do people sign up for the waitlist if they're interested in learning more? 

Ian Leaman: Totally. You can go to pantry. ai. You can sign up for the waitlist. If you use QuickBooks or SYN7 slash gear, we're starting to let people off now.

We're, we basically just want to like, make sure we can be helpful. So , we're kind of restricting signups to people who like the existing tool can work for them, but we should be letting more people off the waitlist in the next three, four months. But the waitlist is the place to go. 

Jordan Buckner: That's awesome.

And we'll make sure to drop that in the show notes as well. Ian, thanks so much for joining me today. 

Ian Leaman: Thank you so much.