Sara Ittelson (00:00):

Hey everybody, I'm Sara Ittelson, a partner at Accel, and I'm here in the studio today with James Theuerkauf, the CEO and Co-founder of Syrup.

‍

James Theuerkauf (00:15):

Thanks for having me. Sara.

‍

Sara Ittelson (00:16):

Can you start off by sharing a little bit about yourself and kind of how that led to the founding of Syrup?

‍

James Theuerkauf (00:22):

Yeah, absolutely. I have a background in economics with a focus on econometric and statistics. He loved data and data inference models and what data can tell us about the world. I then went on to work at McKinsey with a consulting company where worked mostly in apparel, fashion, luxury, consumer retail, worked with close to a dozen different brands and retailers and really saw the pains and the complexities of inventory management firsthand and the pain that it produces for the people working as merchandise planners and the costs that it has on profitability as well as sustainability at the brands and retailers that we worked with. And that's where this idea came of saying, isn't there a world where we make use of the fact that we have access to more data today, orders of magnitudes of more data while living through this cambrian explosion in AI and ML to really help our customers make exponentially better inventory decisions. This is an age old problem that we're solving how to get the right product and put it in the right location at the right time. But with Covid and the supply chain crisis, we had an incredible variability in demand and in supply that created a lot of disruptions that showed retailers the huge advantage that comes with being agile and being able to respond rapidly to inventory. And that's been a big market pull for us, the desire to want to solve this problem in a different way.

‍

Sara Ittelson (01:44):

I show up at a store, I'm looking for an item in a particular color, a particular size, what ultimately leads to that, either being on the shelf or not when I want it to be there. And maybe how was it before and what is it like now for customers utilizing your technology?

‍

James Theuerkauf (02:00):

How does a red T-shirt in size M get into the New York Soho store of your favorite brand? It's a very difficult problem because you have these merchandisers and planners making decisions not just on that one skew, but on thousands or tens of thousands of products across a complex network that involve stores and wholesale partners and warehouses. And the way this is done usually is basically just looking at averages, averages in the averages of averages. Looking at historical sales, we sold two units over the last four weeks, so we expect to sell two units in the next four. It's a bit simplified, but that's really how the industry worked beforehand. What we are doing is we're using a lot more data, so we use the fact that what's the micro trend of that color in that location? What can we learn from search and social media? What's the weather pattern? What can we see about trends on our customer's website and how are people interacting with certain parts on our customer's website to make a much more granular assessment on what's the precise inventory need for that T-shirt in size, M in color red in that New York soho store? And we do that across the entire assortment of the entire network to get a lot better at figuring out how many products do you need and what parts do you not need in particular locations.

‍

Sara Ittelson (03:21):

And then I guess maybe speak to the complexity in this current era of there was a time when purely D2C was the in-vogue thing. I think now folks have generally moved to more of a belief of omnichannel being the winning strategy to be a long-term durable brand. How does that make this challenge more complex for a customer?

‍

James Theuerkauf (03:43):

Yeah, a hundred percent, exactly as you mentioned, Sarah, it used to be the case that, oh, we just have stores. And then that was one model and then it was, oh, we just have online or DTC, and that's another model that we thought of the customers being in different channels. That's changed. We all know this buy online pickup in store or buy online, it gets shipped from a store or I buy something online, I return it in a store. It means that demand is much, much more complex. A store in itself doesn't just sell to a retail customer, it also acts effectively as a micro fulfillment center. And that means that we have demand patterns, return patterns or negative demand from customers in a lot more interwoven ways. That means that you can't just look at the world anymore in a channel by channel way, but you really need to integrate demand and cross correlate all those different factors so that when you order something online or you go into a store and want something, that it's there, that it’s there at that moment. And AI is a huge, huge enabler to solve this really complex multi-tier stochastic optimization model in a solvable way.

‍

Sara Ittelson (04:46):

Maybe just on that point, you founded the company before AI was so prominent in the global discussion, and so what maybe gave you confidence that AI was going to be such a powerful tool to solve this problem? And then what has it been like building in this space during this sort of explosive time of evolution and iteration on the technology stack?

‍

James Theuerkauf (05:11):

Yeah, a hundred percent. So when we founded this company, my background is in econometric statistical modeling, and over the course of the time we're at Harvard, at the time we were founded the company, we got exposed to the amazing advances that were happening at the time in machine and machine learning modeling in particular. And that was this big eyeopening moment of, wow, there's something really big happening here and something that's really changing the rules of the game. The statistical modeling and forecasting that I'd learned was being way outpaced by these new approaches to doing it. And that was the moment to say it's worth doubling down here. We definitely didn't predict the massive boom in AI that is got the trend, I think, we got the shape of it. And then it's been an amazing time to be building in this space, particularly I think from an explainability and customer excitement point of view, it's a bit overwhelming for people sometimes, but this is overall sense of wanting to try out something new because people just see the massive impact that AI can have with large language models. And if you think you can tell a chatbot to tell me a joke, if you are George Clooney sitting on a beach in Spain, what can AI do with your data for your business?

‍

Sara Ittelson (06:33):

What are they using as sort of the core metrics where they're evaluating success and using your technology?

‍

James Theuerkauf (06:40):

If you can drive your in stock rate up, so your availability up with your weeks on hand or your access down. So we allow our customers to sell more product while holding less product, and that leads at the end to a sell through a profitability metric that we can measure.

‍

Sara Ittelson (06:57):

And I mean it's so powerful both for the bottom line for your customers, but also I think from a society perspective, I think people have gained additional awareness about the waste in this category. And so maybe do you want to just spend a minute on what happens when they create too much inventory? What is that sort of downstream liquidation cycle that you're helping these brands avoid? 

‍

James Theuerkauf (07:22):

Yeah, a hundred percent. I mean, this is an industry that works on a model today, which is we produce 10 items to sell three or four at full price, and then maybe we discount three others and then the three at the end either go to a liquidation channel or go to a landfill or get that. And that's a huge environmental cost. Apparel and fashion in particular, it's the most globalized industry. It's one of the world's most complex industries. It's one of the oldest industries, but it's also one of the ones that operates on the least agile demand chains or supply chains. And the big impact there is saying, how can we help reduce some of that excess that I don't need to produce 10 units to just sell four at full price, but how can I produce eight? How can I produce seven even or how can I produce 10 but 10? The right 10, right? The ones that people actually want.

‍

Sara Ittelson (08:20):

Get the 10 to the right place. 

‍

James Theuerkauf (08:22):

Exactly. Exactly. I mean at a societal level, it's 20,000 football stadiums that we could fill with all the garments that go to waste every year. So any change that we can drive there has a big impact on our customer's bottom line, but also on the planet.

‍

Sara Ittelson (08:38):

Yeah. We've talked a lot about how you've been applying AI, the current history of the company, but what has you most excited about what AI can do in the coming years, the next 24 months? How are you expecting to harness it and what are you most excited about?

James Theuerkauf (08:57):

A hundred percent. There's a lot that we're really excited about because there's a lot of product to be built, and we're really fortunate to work with an amazing group of customers who have been phenomenal design partners for us and to keep pushing us to help them with more modules and more use cases. And we started with a set of use cases around optimizing the inventory that you already own today. It's a really nice wedge. It's really easy for our customers to implement and to see the benefits quickly, but there's a lot more that we are looking to build right around how can we help our customers get the right product in the first place? How can we help them with their upstream suppliers and their upstream partners to produce the right inventory to hold the right raw materials? And the more you look at it, it's really this multi-stage problem that's really complex. The demand that someone shows you on a website or in a store, how does that get fed into our yarn manufacturers and figuring out how much stock to hold there. It's a beautiful problem for AI to solve and we could not do it without. Totally, and that's what gets us so very excited about this problem.

‍

Sara Ittelson (10:05):

It's awesome. Any concerns or what are the challenges about building in such an emergent space?

‍

James Theuerkauf (10:11):

There's definitely problems. I think one is people being overwhelmed by AI, right? It's on all the headlines, but then if you are a merchandise planner or an IT professional at a brand, what does that mean for you? And then there's so much out there that how can we really bring this down to a level where it's implementable, where it's usable short term and immediately where it's not just, oh, the art of the possible, what could be in the future, but how can we help your business today and how can we help drive impact for your business today? I'll say that's one of the two big ones. The other, there's a lot of noise in the industry. Whenever there's hype, everyone starts saying, we're AI powered. We use AI. We're an AI company. And it's hard for both for customers as well as for employees and people in the overall space to understand what's real and what's just noise. And that's a big one for us is how do we cut through that noise and show that we're not a legacy software, some AI sprinkled on top, but it's an AI company from the ground up.

‍

Sara Ittelson (11:23):

We talked about the impact you have for customers. Maybe you can just make it concrete within a particular customer story. What did those core metrics look like for them before Syrup and then after Syrup?

‍

James Theuerkauf (11:36):

Yeah, a hundred percent. So what we often do with our customers when we start working with them is we take jeans, a portion of their business that could be a portion of their assortment. So say we do all the genes and we let our customers do what they're doing today with all the rest, or we take a portion of a regional, like a country or the northeast of the United States, for example, and then we compare and contrast and with one of our customers, to give you an example, it's a fast growth American omnichannel brand coming up to close to a hundred stores. We look at the in stock rates before pre post and compared treatment and control group, and we saw an increase of 8% of inventory availability while at the same time looking at on hand level. So how much did they have and reducing that by 25%.

‍

(12:25): And the combination of the two was really powerful because we allowed them to sell more while holding less inventory. And if you look at it from a profitability standpoint, it was an increase in net margin by over 4% for our customers. Coupled with a workflow component, which is looking at what's the amount of time that our end users spend on working on manual tasks, we reduced that by close to 80%. So it's a lot of time they were able to give back to them to not do the manual boring number crunching, but to actually be able to think more analytically and more strategically about their business.

‍

Sara Ittelson (13:04): And in the prior model where I was in Excel saying that we should do four mediums in this store, what were the inputs I was using to make those decisions as sort of an individual planner pre syrup?

‍

James Theuerkauf (13:16):

Yeah, it's really, really hard to do this manually. And what you resort to as a planner is often just transaction history. And so what we are doing is we are taking that painful manual number crunching and not just looking at transaction history. So it's looking at not just how many units of this red T-shirt in size M and Soho sold over the last two weeks, but it's saying, oh, black Friday's coming up in a few weeks time. It's a pretty warm November for New York more than usually there's a trend in the color red because of a certain influencer story or a certain movie or whatever it is to get a lot more precise.

‍

Sara Ittelson (13:55):

Are there particular product lines or categories that are especially hard to crack? You know, women's blazers? Are there categories that are harder than others? I'm curious.

‍

James Theuerkauf (14:09):

Yeah, the more standard the product, the easier it's to forecast, right? A standard white T-shirt, there is actually a variability, but there's much less a variability than say a red woman's sweat blazer the world has never seen before because it's a fashion item. So forecast accuracy on the white T is definitely going to be higher than on the red blazer by the end of the day. What really matters for us is the relative improvement that we can drive. So on the white T-shirt, you are better at it, but we can make it even better on the red blazer. We might be worse at forecasting it than the white, but relative to other models or relative to what you're doing today, we can still be significantly better. So for us, in fact, the more complexity, the better, the more sizes you have, the more colors you have, the more newness in the assortment. All of those are really, really good inputs to power this complex AI engine and to drive real business value.

‍

Sara Ittelson (15:16):

And this is a place where actually consumer expectations is such a driver for your product where the norms have evolved such that people are expecting quicker turns of fashion, but doing so in a way where you can deliver those quicker turns while reducing the waste and not having all that inventory glut. That feels like the magic equation. That's been hard for retailers, but you guys are enabling.

‍

James Theuerkauf (15:42):

A hundred percent. Exactly. And if you look at customer loyalty, for example, stockouts are really costly and we all know this, right? You walk into your favorite store, you find the perfect jeans or the perfect T-shirt, but the size that you want isn't there. Or you're on a website and you're scrolling through all of those products and then finally you find the one you want. And of course your size isn't there, and that's a really frustrating consumer experience. And if you look at loyalty of consumers to brands with stockouts, they reduce significantly. So there's not just a real monetary cost for a brand of not being in stock, but there's also a lifetime value cost for not being in stock. So getting that right is really, really powerful. And you're exactly right on the excess as well. The rules of the game have changed. We can no longer live in a world where it's okay to just overproduce massively and then burn everything that we didn't sell, but there is a consumer expectation that we work in a society responsible way and environmentally responsible way and enabling the two of them. Exactly. That's the magic.

‍

Sara Ittelson (16:53):

What does this mean for the bargain hunters out there? What's the implication for them?

‍

James Theuerkauf (16:56):

Yeah, I mean there’s this worry of how our discounts going to go away, discounts are not going to go away. And fans, in fact, discount products because they want to discount products, the problem is when they're forced to discount products because no one wants the items, and that means that we live in this weird world where there's 70, 80% discounts because it was overproduced stuff that no one wanted. If we get better at determining what do people want, brands can pass on those savings to consumers. So discounts are they're not going to go away, but it's the unwanted discounts that really, really hurt brands, the wanted discounts, they're fantastic, and hopefully we can enable more of those and so brands can reach the consumers that they want to reach and for consumers that they get the bargains that they want to. 

‍

Sara Ittelson (17:51):

Totally, but reapplying it as a growth lever as opposed to sort of a loss reduction, more reactive as opposed to proactive lever.

‍

James Theuerkauf (18:01):

Exactly. Exactly.

‍

Sara Ittelson (18:02):

Now, well you shared in your background you personally are quite global, the company is as well. So maybe you could talk about the team and building a globally distributed team and what that's been like. Your customer base is simultaneously probably not unrelated. So yeah, probably useful for some of our listeners to hear about how you've managed that.

‍

James Theuerkauf (18:27):

Yeah, a hundred percent. So as you mentioned, I'm half British, half German. I grew up in Spain. We founded this company in the US so this has been a global endeavor from the get go. We founded the company in Boston. Our first customers were in Switzerland and in Germany, and so we've been global from the beginning. We also found this company during Covid when actually there wasn't an ability to co-locate. So our default was we'll hire people wherever, and we've carried that with us up until today. The team today is spread between Hawaii and Cape Town in that time zone, in that time zone span, it's worked really, really well for us, particularly it's worked well to be able to find the exact right talent that we needed at the moment that we needed it to not to be bound by the radius of commuting into Manhattan, for example.

‍

(19:21): But we're looking for a data scientist with a particular skill set in demand forecasting, and we find him in the north of England, that's who we can hire. Or we're looking for a specialist software engineer and turns out she's based in southern Portugal, we can hire her not bound by those constraints. Really, I think the one thing that is harder by being a remote and globally distributed team is building trust. Just seeing people on those little tiles in video conferencing tiles doesn't build trust, particularly when you come from these different cultural backgrounds. And so what we do is we bring the entire company together once a quarter in person to work on that team building and that culture building and to build trust between people, which has worked phenomenally well. They're always really, really, really fun, Syrup summits, and you can just see when people then go back to their home towns and home countries, it's a new level of energy and excitement and carrying that through as the company grows is something that we really, really want to do.

‍

Sara Ittelson (20:29):

Yeah, it's a great case study in sort of maximizing the benefits that you can get from being distributed while being really intentional about sort of guarding against some of the failure cases. And I had the pleasure of joining the Syrup summit for a bit most recently, and it was really powerful to see that in action and see the trust building that you did amongst the team and just a really incredible special group that you've been building. So kudos to you all for being so thoughtful about it. Maybe we can go back in time a bit. I remember leaving our first call and actually incredibly excited, and one of the things that I felt so excited about was you finished and you said, I usually don't let anyone talk to Furie. I usually try and guard his time, but that at the end of our call you said, I want to connect you up with Ferdy. And so it was like this real, I don't know, sort of permission granted to spend more time with the team, get to know you guys better, and a thank you. I appreciated the trust. But maybe we can talk about a little bit about our journey from first call to now joining your board. What's your recollection of that first call?

‍

James Theuerkauf (21:51):

Yeah, it was a great call, I think particularly Sara, because you really understood our problem space and that made a real difference for us. There's a lot of very, very smart venture investors out there. There's not that many that truly understand the pains that our retail customers face and the opportunities. And I think in that first call, you made it very, very clear that you did and that you had some really cool ideas of what we could do. And that was me saying, wow, yeah, let's show you the product, right? Let's see what you actually think and how you can help us. And that was our first call. I think from that moment on until we actually signed our series A, it was a six to eight month journey. We met in Las Vegas at a trade show Shoptalk, one of the very large conferences, and in the middle of all of that craziness, I thought it was really cool that you were at that conference as well, looking at what does the market look like and really building your own understanding of why is Syrup maybe something different to what the rest of players are doing, and what's the real differentiation.

‍

(23:03): I think having gotten to know you and Accel over that period of time, you helping us with product ideas, introducing us to people, that was really, really powerful and means that the moment that we said, yeah, come join our board and come lead our series A, it felt like we knew each other already, that we'd actually had ample opportunities to test the waters of what does it feel like working together. 

‍

Sara Ittelson (23:26):

Is there anything that you'd want to share with our listeners that has you most excited about what's happening for syrup specifically within retail, broadly within AI, maybe open floor for founders are uniquely passionate and observers of interesting trends and dynamics, so anything else you'd want to share that you're excited about?

‍

James Theuerkauf (23:52):

A lot of things that I'm really excited about. I think retail in particular, it's such an interesting moment in an industry that's worked on an operating model for a really long time operating model of very long lead times being very unresponsive, not very agile, being slow to adapt, really that's really changing really rapidly and where the fact that you can become a data company is really distinguishing some from the rest, right? Amazon is famously effectively a data company. For all of its pitfalls is also a data company with a very, very strong foundation. And that means that we can actually drive real change in retail and that this historically fairly slow industry is really, really modernizing. And that's an incredibly exciting time I think for the brands and retailers that are in the space today for new entrants and also for people coming to join working in the industry, the old ways of these really clunky legacy systems and spreadsheets that's changing and we see it live with some of our customers who are at the forefront there.

‍

(25:02):

And it's always difficult to be a pioneer, but it's so exciting, and that's true obviously for the inventory related topics, but it's also true across advertising and e-commerce experience in store experiences. I think on the broader AI wave, this is just the moment in time and it's so, so exciting to be in the middle of that. It's been such a crazy rapid evolution of models changing and new updates coming. GPT3 and three and a half and four, and there's so much more to go. I'm just fascinated by the amount of data that we can get and that we can process in all different application use cases. For us, it's how can we help our customers understand demand better and then shape supply. But the same is true in a whole bunch of areas and the impact that it can have on our working lives of spending less time on the mundane tasks, but actually being able to elevate ourselves to do much more creative strategic thinking, not doing the repeatable, leaving the repeatable to the machine and us being able to focus on the non repeatable, the creative. That sounds like a very exciting proposition.

‍

Sara Ittelson (26:17):

We have a lot of founders who listen to this podcast. And so I'm curious what advice you would have to other early stage founders folks building now and maybe actually on this point around sort of cutting through the noise on AI and what's real and what isn't, maybe to founders or to people considering joining early stage companies, how do they discern who's really packing the punch that they are suggesting they can?

‍

James Theuerkauf (26:47):

Yeah, I think the big one is customer impact. What value can we drive for our customers? It doesn't matter if we have the best AI models out there or the most sophisticated engine if it doesn't actually lead to a measurable meaningful impact for our customers. And that impact could be as in our case, on the revenue, on the cost side, it could also be on the workflow side and for our customers or on user engagement side or on the waste side. I think there's multiple metrics that one could look at, but that's really what it comes down to. Our customers don't buy syrup because we have the fanciest AI. They buy us because we can meaningfully impact the metrics that they really care about. And that will be my advice to founders out there as well. Fall in love with a problem, solve the problem, and then figure out what's the right tech to get there as opposed to here's this fancy tech.

‍

Sara Ittelson (27:44):

We just went through a fundraising process together. What was it like fundraising as an AI company in the most recent market?

‍

James Theuerkauf (27:53):

So we started this company in 2020. We did our first round in 2021 when it was the mega hype, the super buzz years. We had a really hard time fundraising and we took us a really long time to get money. We got no traction with just rejection after rejection after rejection until we finally found out our first investors who end up giving us a million dollars overall to get this companie up. But it was a really difficult process and we did a lot of things wrong. And then we did our second round in early 2022, which is when the market started to come crashing down and that we ran a full process there. The typical, we spoke to 60 different VCs, ended up thankfully getting a handful of offers and went with the one that we liked the best, but it was a very long, very involved process.

‍

(28:44): And now for the series A, the way that we chose to do it was to meet a few people over the course of a year, people that really respected that we thought could lead this round like yourself. And we built a relationship with these people over a period of six to 12 months, which I think allowed us to get to know you, to get to know us. So then the moment that we said, I think now we're ready to go fundraising, it was a very natural conversation and a much shorter one as well than we've previously done.

‍

Sara Ittelson (29:14):

Yeah. Well, we're thrilled to have you in the Accel family. I personally have just loved working with you, thrilled to be joining the board and joining you on this journey. So thank you so much for spending this time in the studio with us today, and we're excited to keep our listeners posted on Syrup's continued successes.

‍

James Theuerkauf (29:33):

That's awesome. Thank you so much, Sara. Thanks.

‍