At the Top, Nothing is Beneath You | Cece Lee at Trendalytics

In the first part of my conversation with Cece Lee of Trendalytics, she talked about her efforts to use artificial intelligence to analyze trends in fashion.

In this second part of our conversation, Cece opens up about her experiences as a CEO of a startup in New York, and the humbling process of transitioning from an account manager to being the head of a company.

She offers some key perspectives on learning by doing and how being at the top can’t mean that you’re above anyone.

More info:

Cece Lee is the CEO of Trendalytics, a subscription-based, product intelligence company that empowers decision making with data-driven insights. Over the past two years, Cece has overseen the doubling of business and personnel within the company while building repeatable, scalable processes to quadruple the efficiency of each member of her team.

Prior to making the jump to the fashion tech world, she worked at a variety of different companies, from Target to Jimmy Choo to Michael Kors, across buying and planning functions. Cece holds a B.A. in economics and government from Cornell University.

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Cece Lee 0:00
Retailers I think are famous for essentially having product too early or too late. Because their designers are like, Oh, we definitely think people are interested in this thing because they’re seeing it on the streets in New York, they’re seeing it in Europe. But whether or not your actual consumer is ready for it is an entirely different thing.

Alexander Ferguson 0:23
One of the biggest goals of artificial intelligence is to process human communication. But communication can change depending on the context, the way you talk to your family may be different than the way you talk at work. And if you work in pharmaceuticals, you’ll talk a lot differently than if you work in fashion. And it’s the fashion industry that Cece Lee at Trendalytics is working to crack. In this episode of UpTech Report, Cece dives into her efforts to predict fashion trends so that manufacturers know what to make, and retailers know what to stock. Cece, I’m excited to chat with you today and learn a bit more about trend analytics and the journey that that has been on in you have been on as a leader. I’m gonna ask you to start with describing your company in five seconds. What would you say?

Cece Lee 1:08
Trendanalytics is an AI and ML based tool that essentially helps merchandisers in fashion and beauty space, pick better product,

Alexander Ferguson 1:18
Pick better products, using an awesome technology just pick better products. So the market segment is is fashion. Right? Is it helping many retailers? And manufacturers? What’s the main area that you then serve? Yeah,

Cece Lee 1:35
so the interesting thing about the way that our company works is that there’s such a broad swath of information that’s out there essentially to be mined, that we can help anybody along the chain. So we help manufacturers, we help retailers, we help off prices. We help basically anybody who’s tangential to the fashion industry, who needs to know what people are interested in.

Alexander Ferguson 1:56
And with this, this problem that you saw, give me a good use case of one of your clients like this is where our technology really shone and shined.

Cece Lee 2:06
Yeah, so there’s a variety of different roles within those types of companies that can benefit from what we do. But I think, you know, on a micro level, it can help. So here’s a silly kind of example. We had a retailer, essentially, who had a product and it was jeans, right? So they had jeans on their site. And they were the remember the like holey jeans, the ones of all the rips, and all of that kind of stuff. So essentially, they had it on their site. And the fashion way to talk about that particular thing is by calling it distressed denim. But literally zero consumers are searching for the term distressed denim, because to every other person gets ripped jeans. And so they were able to essentially just make a quick copy, edit and increase their click through rate by 20%. Because even if you have the right product at the right time in the right place, if your consumers can’t find it, it doesn’t matter.

Alexander Ferguson 2:55
Gotcha. So it’s understanding those naming structures, the trends that you can I make those modifications and see that insight.

Cece Lee 3:03
Yeah, so it’s kind of a combination, like we use, obviously, like natural language processing. But we also think about just what’s the colloquial like the fashion person way to talk about something versus the way that consumers are actually searching for it. And another part of that also is understanding when you should have it. So retailers, I think are famous for essentially having product too early or too late. Because their designers are like, Oh, we definitely think people are interested in this thing. Because they’re seeing it on the streets in New York, they’re seeing it in Europe, but whether or not your actual consumer is ready for it is an entirely different thing.

Alexander Ferguson 3:36
So you’re really just trying to help them connect what’s actually out there, and they just are not seeing for whatever reason.

Cece Lee 3:43
Yeah, you know what every retail company kind of says the like marrying the art with the science. But that’s literally what we’re doing. As well, as you know, in a nutshell, it’s more hits less misses.

Alexander Ferguson 3:53
And your business model, subscription base, is it? Is it heavy on the SAT software’s or service side? I mean, yeah, it’s

Cece Lee 4:01
heavier on the software side. And I think that’s changed over time. But at the end of the day, as a company, a software company, especially you can’t really scale if you’re so reliant on services, because obviously that scale says a one for one. But with a software that people can kind of ala carte grab as you go customize it, however they want to, you know, the possibilities are a lot more limitless. I guess the way that I want to put it.

Alexander Ferguson 4:24
What size of in retailers? Are they you working with? A winner like what’s the cutoff on both sides

Cece Lee 4:31
are? Yeah, it really there’s no real minimum threshold? I think it just has to be the the real way that we think about our customers from a like an ideal client profile perspective, is you have to be interested in data. You have to understand directional data and not just you’re like, Oh, what did I sell last year? What did I sell through what does it look like? But there’s no real bottom or top threshold. It does tend to be better I’d say with your largest retailer dealers that have a focus and what consumers are interested in. But at the end of the day, if you care about data and you care about, you know, what people are interested in, anyone can kind of find a home with us.

Alexander Ferguson 5:12
Obviously, it has changed over the years what what the product is everything evolves. So tell me a bit more about the technology over the past several years of how it’s kind of morphed and changed into where it is today.

Cece Lee 5:24
Yeah, so I think at the beginning, really the, the place that we went was a little bit broad. So we have three different data sources. So we use search social, as well as market. And the funny thing is for our kind of target audience, every single company in our space, no matter how close to it, or how far uses the exact same language. But for us, you know, the real differentiator has been using all three of those sources in equal measure. So we have competitors in the space that focus very heavily with my dog on only the market segment, or on only this social media segment. But there are no other people in the space that are using all three of these things together to really create a predictive algorithm for based on the, you know, the leading indicators, where does that say that your trend is going to go in the next three months? Six months in a year?

Alexander Ferguson 6:14
You mentioned those three things, can you expand on on a bit more than of how those come to play and then visually show up for the for your clients when they’re interacting and seeing this?

Cece Lee 6:25
Yeah. So you know, when we really think about it, so search, essentially, we’re pulling from Google, right? Because, you know, what is a better indication of intent to buy then Google query. We’re also scraping social media sites, mostly from influencers. So it’s not from every single person, not from me and you, unless you’re an influencer? I certainly am not. But it’s, you know, who are the people who are leading the conversation forward, that people are looking to, to then get their indicators of what they want to be wearing, or buying and seasons to come? And then EECOM sites, so we’re also crawling a variety of different ecommerce sites, to then give you pricing information to understand. Okay, so similar types of products? What’s their average retail? What’s their markdown rate? Is it new to markdown? Is it highly discounted? Is it full price for a long time? Is your market saturated? So you can get an understanding of what’s going on in your competitors, proverbial yards?

Alexander Ferguson 7:21
Gotcha. The then in use case on someone’s logging in and utilizing this they putting in their own fashion content and, and then being able to see, is this good? Is this bad? How does that work, then?

Cece Lee 7:34
It’s more of a discovery engine, really, because we want people to be able to come in and either validate things that they’ve kind of seed on the street and say, Is this something that is data data supported? And should I bring it to my divisional? Or what should I do with this, but then also to be able to just say, you know, what surprised me show me something that I don’t know, in tops, because my job is 85,000 different things. And this is just one of them. As opposed to for us, like our sole focus is to okay, this is also a very silly turn, but surprise and delight our users. So when they log in, they can say, You know what, I knew that puff sleeves were a trend, but I actually didn’t realize that. It was also just specifically organza fabric puffed sleeves. So it’s it’s a very, it can be very specific or very general. But the idea is to help retailers push the ideas forward. Even if it’s a totally unsexy concept to understand also when it’s right for their consumers, because, you know, if you’re too early or too late, people aren’t going to buy it from you anyway.

Alexander Ferguson 8:37
Hmm. Does do you get predictions on based off of those three different inputs that are getting that there is a timeline of when this will no longer be in fashion anymore?

Cece Lee 8:48
Yeah. So right now, the way that our prediction engine works is we do three months, six months in a year out, but based on where you as a retailer, or you as a brand, or you as a manufacturer, sit on the adoption curve that can be as far as two or three years out. So for example, if for example, your customers adopt or late adopters, something that we’re showing as an emerging trend that will grow for another year might actually have two years of life for you because it’s an emerging trend that the influencers are wearing. However, your you know, a soccer mom in Dubuque, Iowa might not be ready for it yet.

Alexander Ferguson 9:22
Your clients then do they customize their profiles, so they only see that data or do they, when they’re going through it, that they they make the choices saying, okay, for I know my audiences, so I know this is good. Do you help them with that?

Cece Lee 9:33
Yeah. So it can actually be both. So the way that the platform is built is endlessly customizable. And so you can track only the categories that you’re interested in. And you can save your filter. So every time you log in, you see the same thing. We actually do like to direct our users to kind of what do I want to say, get out of their own heads, right? Because you can actually self select yourself into something that’s so specific, that you miss all of the context that’s around it. And so So the the best way to use it really is to kind of come in and say, Let’s see everything. But then I want to only see my own juniors boys graphic tees, but it’s just so specific that you have to sometimes, you know, because the trade off might be, you know, from pants to dresses. But if you’re only ever looking at pants, you’re gonna miss the fact that dresses as a category is growing. How long has the company been around now, gentlemen? Yeah, so the company’s been around since 2013. I’d say in our latest iteration, though, really? The last four, four ish years?

Alexander Ferguson 10:33
Four years. Okay, gotcha. And the number of clients as I’m sure I’m sure change over the years, how many active people are being able to utilize your platform now?

Cece Lee 10:43
Yeah, so it’s basically because we’re b2b, it really varies on how they disseminate the information. So in terms of eyeballs on our data, it’s in the 1000s. But if you look at the users, you know, they might be taking the information, packaging it and forwarding it out. So a lot of companies have actually found success and having like a group of power users that then disseminate the information to everybody.

Alexander Ferguson 11:08
Because even you still get the data someone has to make figure, okay, this is what we need to do take action, not everyone loves data.

Cece Lee 11:14
Yes, exactly. And so that’s why we target very specific roles within those companies as well, because some people like pretty pictures, some people don’t care about the pictures at all. And so I think, you know, the kind of intermediary is the people who influence the products that actually go into the stores and stores can be calm, it can be actual physical stores, it can be wherever. And you’re right, some people don’t like data, and that’s been a learning for us over the years is who to target and who not to target. But I won’t discount people who you might not think would love data, because some of our most active cohorts of users are the trend offices that a variety of the largest retailers, because they’re also interested in what’s new, and next. And so I think they’ve actually a lot of them have been, like ducks to water. So I think don’t put anyone in a box of you like data, you don’t like data, either.

Alexander Ferguson 12:03
Gotcha. Looking forward, then where do you see the company in the near term and long term? So the next year or so? And then long term? Like five or 10? years?

Cece Lee 12:14
That’s a really good question, actually. So I think near term, it’s, you know, we have such an amazing opportunity at this point, with everything happening with COVID. And with all the stores being closed, I think, because you can’t just rely on your current selling data anymore, because you don’t have any current selling data. But companies are still going to have to buy their assortments for next year, they’re still going to have to buy their assortments for fall on holiday. And so the question is, in the absence of data, what do you use directionally, to fill that hole. And I think that’s actually where we come into play. And it’s, it’s been interesting, because in the past, it’s been mostly the most data savvy companies that have really signed up with us, because they can not only understand their own data, but also understand the importance of directional data outside of their four walls as well. And so I think more and more people are gonna have to get right, get right with the man for lack of a better term, and really understand how that influences what they’re picking, because especially, you know, they’re not going to be able to rely on the old ways of doing things. Because in a lot of ways, the old way of doing businesses is dead, right. And you have to figure out how to innovate or die right is essentially the the mantra and retail. And so I think that for the next year, it will be how do we strike while the iron is hot, but also just on a human element help people when they’re being tasked to do more with less. And that’s also that’s kind of been the the stage of retail for the past five to 10 years really is going through a period of contraction, where the age of malls was huge and fat for everybody. And now you have to cut down processes, you have to be a little bit more nimble, you have to behave a little bit more like a startup. And so I think it will be helping people get through the most challenging time of you know, their their whole careers, right. We’ve had people who’ve been working for 3035 years say this is the most disruption that they’ve ever seen, and how do you navigate through that? And then I think longer term, obviously, there is so much rich data that we have, it’s really going to be supercharging, our data scientists to then say, what does it all mean? And how can we use it in a smarter way? Because from my perspective, we’ve really only scratched the surface on a lot of it.

Alexander Ferguson 14:26
Where can people go to learn more? And what’s a good first step for them to take?

Cece Lee 14:33
That should be such a softball question, but I’m like, well, that’s a great question. You can really go to you can go to our website. So it’s a trend to Linux, t r e n d al YT But you can also reach out to us at Hello at trend to very important not com co but yeah, you can really reach out you can follow us on Instagram. You can see us on Twitter, there’s a variety of different ways to get in touch with us but the easiest way is to just shoot us an email and say I’m interested because we’re always happy To chat through people’s businesses also get an understanding of what they’re looking for and even just to do a demo,

Alexander Ferguson 15:07
be sure to check out part two of my conversation with Cece, in which she offers some important insights including the necessity of understanding the human perspective. Why no meeting is a bad meeting, and how the rampant adoption of online working and schooling will change business.



YouTube | LinkedIn | Twitter| Podcast

You Are Not Your Company | Susan Stevens at Wespire

Predicting Preppy: AI in Fashion | Cece Lee at Trendalytics