Supply chain optimization: When to automate and when to wait | Jonathan Tam, COO of FIGS

Watch on YouTubeListen on SpotifyListen on Apple Podcast

Featured Speakers

Ben Eachus

Flowspace Co-Founder & CEO

Jonathan Tam

FIGS Chief Operating Officer

About this episode

When does warehouse automation actually make sense for a growing brand? When does the right partner matter? In this episode of Well Delivered, Flowspace CEO Ben Eachus sits down with Jonathan Tam, COO of FIGS, to discuss supply chain optimization and the decisions that help brands scale efficiently.

Drawing on his experience scaling Revolve through its IPO and now leading operations at FIGS, Jonathan shares how smarter fulfillment processes can drive meaningful improvements, why strategic supply chain partnerships matter, and how AI is helping teams uncover new opportunities to improve performance.

Key takeaways

  • Automate with intention: Invest in warehouse automation where it delivers lasting value, and stay flexible as your business evolves.
  • Optimize fulfillment processes: Smarter order processing can improve efficiency without adding more equipment.
  • Build strategic partnerships: Strong supply chain partners help brands expand capabilities without proportionally increasing headcount.
  • Put AI to work: Use supply chain technology to uncover shipping inefficiencies and make better operational decisions.
  • Prioritize relationships: Long-term partnerships and shared goals are essential to successful supply chain optimization.

About the guest

Jonathan Tam is the COO of FIGS, a leading healthcare apparel brand. Before joining FIGS, he spent seven and a half years scaling fulfillment and operations at Revolve Clothing through its IPO.

With a background in engineering and ecommerce operations at Forever 21, Jonathan brings a data-driven approach to supply chain optimization, fulfillment technology, and operational growth.

Narrator (00:00) - Welcome to Well Delivered by Flowspace, the podcast unpacking real insight for building and scaling modern supply chains. Today, we're joined by the COO of FIGS, who spent seven and a half years scaling operations at Revolve, to talk about when automation actually pays off, the hidden math behind great fulfillment, and why the right partners matter more than people think.

Ben Eachus (00:26) - Hi, my name is Ben Eachus and I'm a co-founder and CEO of Flowspace, a fulfillment company that helps brands scale their operations. Joining us on the Well Delivered podcast is John Tam. John spent seven and a half years scaling operations at Revolve Clothing up until their IPO, and is now the COO of FIGS, the leading clothing company in the medical space. So John, thank you so much for joining the call.

Jonathan Tam (00:51) - Great to be here.

Ben Eachus (00:52) - Awesome. Well, John, I would love to just know a little about how you got into operations. I know, I think you were an engineer by trade, and what drew you to the operational world?

Jonathan Tam (01:04) - Great question. So yes, technical engineer by trade, did a mechanical undergrad and then a master's in engineering management science. And it was really a friend of mine that said, "Hey, we have a lot of problems at this company called Forever 21 in its heyday, scaling like crazy. We have this number one performing store called eStore. Nobody understands it. You have an engineering background. Come here and figure it out." And honestly, I love those types of problems, I think, looking back. Once I got into e-commerce, I found that a lot of the engineering principles really connected, but it also took my appetite — I didn't love the technical side of engineering. I like the problem-solving, but being in front of the computer doing CAD or design all the time wasn't what I wanted to do. So this was like a perfect blend where I could use the expertise and then apply it in more of a business mindset, and I'm still in it.

Ben Eachus (02:04) - That's awesome. Now, talk to me a little about when you joined Revolve. What did their kind of fulfillment footprint look like, their operations team? You were an early employee there, and then my next — I'm going to ask you what it looked like when you ended. But what was the start of it?

Jonathan Tam (02:21) - It was awesome. It was about 25,000 square feet and 20 people picking off of paper. That was like a lot of the companies at that time, because Revolve was an earlier e-com company — they were started, I think, 2007, 2008 — and whatever worked, worked. Labor back then and real estate back then was much, much lower than it is today, so there wasn't as much of a drive toward all this automation and technology. I helped them evolve, stay scrappy, and continue to grow. As volume grew, we were always very methodical about technology we would adopt versus technology we would wait on. That patience was really important, because there was a wave of e-com companies growing, so it was very easy to jump on and say, "Everyone's doing this, I want to do this." There was a lot of waiting and seeing what ended up maturing, what didn't mature or didn't get enough backing. Now I feel like it's a much clearer space in terms of which technologies work well, depending on different profiles.

Ben Eachus (03:39) - That's super interesting. I think one of the mistakes I made when I was at The Honest Company is we tried to automate certain processes way too early in the life cycle, because our channel mix was still very dynamic, and even the product mix was really dynamic. So we ended up with automation that fit the business a year ago, and we didn't get the full benefit of it. So I'm curious — when you're reviewing these automation decisions, I think it's very glamorous to say "everything's going to be automated," but there's a stage in a company's life cycle where it doesn't actually make that much sense to do it yet — you need to see what you have first. Tell me more about how you evaluate those decisions.

Jonathan Tam (04:25) - Yeah, valid points all around. I think the main key is you have to understand what elements of the business are likely to remain for the long term — the profile of the business. For example, I'll use FIGS: we are always going to sell sets of scrubs. It's very, very common — you look at the data, somebody wants a top and a bottom, matching colors, matching styles. So any automation I can implement that continues to optimize that aspect of the business, I can lean into, because I know it has longevity.

The other side is things like accessories — we just recently launched a charms business, our footwear business is growing, a lot of the non-scrubwear. I really don't know how that's going to evolve as part of the business, so it's very dangerous to overinvest in things you don't know. And then the mix in the middle is a dynamic profile — how many units per order, what type of automation, is it going on conveyor, can it convey everything. That's the middle space where it makes sense to invest some, but not too much — keep your investments flexible and adaptable, so as the business changes — now we have a bunch of stores, a huge B2B business, international volume growing faster than domestic — those things can be accounted for early on.

Ben Eachus (05:55) - That's such a good point. One of the questions I had for you — you described how Revolve was when you started, but over the seven years you were there, it grew pretty tremendously. What did it look like when you ended? The level of automation, square footage, how many facilities — what did that look like?

Jonathan Tam (06:17) - Yeah, so I opened a facility when I was at Revolve almost every year — 10 out of 12 years we opened a facility. We shut down facilities too, so we ended with about four. In each of them, what you'd see is a fair amount of automation equipment, but not a ton — not overly populated with automation equipment, but a ton of intelligent processing. That's where we spent our time really optimizing.

For example, at Revolve, single-item orders and two-item orders made up 80-plus percent of the units and 90-plus percent of the orders — because we had free exchanges, free returns, and fast shipping, so it made no sense to add items or combine outfits. We developed a lot of processes that I think are bespoke, even to today, for two-item orders, because not a lot of businesses look at that. They look at single-item orders, because dynamically they're very different from multi-item orders, but then everyone else just lumps the rest into "multis." So you have singles and multis. When we looked at it at Revolve, we realized: the data isn't split into singles and multis — it's singles, two-item orders, and then multis. Two-item orders, when grouped with multis, are actually the worst multi, because you're not getting a lot of benefit — you're holding one order just for one other order. But you can optimize it really well if you have like-sized products, because you're really only holding one item. Once we leaned into that, we kept digging deeper and found tons of opportunity.

So what you'd see on the floor is, "there are still a lot of people, not a lot of robots or equipment doing things people could be doing" — but when you look into the processes, you'd realize this process is extremely optimized for this business. That's why, the last three or four years I was there, I got to a point where I felt good and ready to leave and explore a new opportunity, because we'd gotten the business to a point where it could support its growth for a short-to-medium amount of time — but the next step would've been, once all these leases run out—

Ben Eachus (08:58) - That's awesome. I think you raised such an interesting point — a lot of the automation people talk about in the supply chain isn't necessarily visible when you walk onto a facility floor. A lot of it is figuring out which work should go together, and that's happening in software before you even give it to someone to work on. I think that's what's interesting about our space — it's a lot of math, and there's a physical element, but there's a lot of work you can do upstream to optimize not just shipping costs, but which items go together, whether the same person is working on a set of orders. All of that's invisible — if you walked into a facility, you'd wonder where the automation is, because it's hidden. I think that's exciting.

So you moved to FIGS, and you're running operations at a public company. One of the things you mentioned in an interview was the importance of picking really good partners, especially in the supply chain. What did you mean by that? Why is that so important?

Jonathan Tam (10:05) - I'd say in today's world, the success of any company, if you really dissect it, is tied to a lot of the strategic partnerships with all the pieces that work on that company. There's very little that FIGS does completely independent of other partners — we're literally a hub working with all these partners. The number I use and share with my team: for every one employee I have on my team or at the company, there are 10 to 20 people working on FIGS indirectly through our partners. That leverage ratio means how you interact and work with your partners is extremely critical.

On top of that, from a development standpoint, the things I need to build to support our business are often the same things our partners want and need to build to support other brands. Finding strategic alignment so they're building it for us instead of us building it ourselves — they're testing it, we're giving them feedback — that relationship is extremely valuable for both sides. FIGS gets the requirements and functionality we need from our partners, and in return, our partners get a roadmap: if you do this for us, you know every other customer coming to you will need it too, and you're getting ahead of it.

The last piece is that changing partners is extremely time-consuming and expensive. So whenever a partner isn't working out, the first thing I think about is: have we done enough to really understand how to make this partnership successful? Sometimes a partner is just going in a different direction than you need, and that happens — identifying that and finding a good fit is critical too. But if all your partners are constantly not performing to your expectations, that's often a sign you're not really investing in them — you're asking things of them without giving anything back, and you're setting yourself up for failure. That's what I really meant by that.

Ben Eachus (12:17) - I love that. The leverage point is so interesting, because you and I both started in e-commerce a long time ago — at The Honest Company, there was at one point 100 engineers. If you looked at a direct-to-consumer brand today, that just wouldn't be possible, because Shopify — not exactly a partner, but a vendor — is doing all that front-end development for you. A lot of software tools have grown up and evolved so you don't necessarily need to build your own order management system or warehouse management system, and 3PLs have evolved too. So the companies we're supporting now have revenue in the hundreds of millions of dollars, but headcounts that would've been unimaginable 10 or 15 years ago — 40 or 50 people scaling a business like this — because you're leveraging all these outsourced vendors, or partners, as you call them. Such a good point.

John, one of the questions I had for you — you oversee not just fulfillment, but a broader purview at the company. What are some of the technologies and developments over the last couple of years that you're excited about?

Jonathan Tam (13:44) - Well, it'd be hard to do any podcast without mentioning AI. We're exploring it across a lot of different aspects of operations and the supply chain. From a shipping standpoint, just extracting and aggregating data — we're shipping hundreds of thousands of packages, and we're able to find nuances like "this is slower, this is faster, this is cheaper, this is more expensive" — that would've been much harder to do manually without AI.

I also oversee trade, which is really interesting right now, specifically in the US, but worldwide trade is interesting too — also something AI can play a part in. We're publicly traded, so there's a lot of inventory control and SOX compliance that AI is helpful with. And a lot of inbound shipping and containerization optimization — I think AI just helps. I wouldn't say in operations it's really replacing anything, but it's helping you understand where you should be looking. The tools are highlighting pain points or insights you didn't even know were a huge opportunity. We're using AI on exchange recommendations, for example — that's a big one, we've seen a major impact improving our exchanges. Customers are telling you these things, but you're not able to visualize or aggregate the data sufficiently without AI's support, and then that helps you realize the impact and explore it further.

On the technology side more broadly, I'd say shipping in general — it's a really big line item, and there are services now doing more orchestration, auditing, even negotiations. They've turned that into less of an art and more of a science, where you have enough information to just say "I need this, this is better, this is better" — it becomes more formulaic.

Ben Eachus (16:13) - Totally — I agree with your point about it highlighting areas you might not have focused on. We've used it to analyze our Slack channels with a lot of our customers — why are they having to reach out to us, and what could we do to proactively flag issues before they have to reach out. From our vantage point, a merchant should never have to reach out about a delayed order — we should proactively tell them so they can message their own consumer. That would've been really difficult to do in the past, but now there are tools available to quickly analyze it. Obviously all our engineers are using it, everyone at the company is using it, but from a product roadmap standpoint, it's been super helpful to flag areas of improvement. Certainly an exciting time.

So John, as we wrap up — because I know you're very busy — you've had a tremendous amount of success in the operational world scaling these businesses. For another leader at a fast-growing organization, whether in apparel or other products, what's the one thing you'd tell them to focus on early as they're scaling? I know that's a hard question.

Jonathan Tam (17:38) - I do believe it's relationships. I mentioned strategic partners — you and I met a long time ago, and I've always made time to focus on the human relationship element of operations. When you do that, you realize there are fewer people and fewer players in this space than you think. When you get started, you think there are all these people, all these titles, all these companies — then you start to develop a network, talk to different partners, and realize that for everything, there are really only five or six companies that do it well, and of those, only two or three are probably a good fit for what you're doing. If you already know them, it makes things ten times easier.

In the future, when AI is doing a lot of the computation, analysis, and data work, the one thing it'll never do is replace human interaction and relationships. I'd attribute a lot of my success to prioritizing that — understanding that everybody comes from a different background and has different goals. My role, whenever I meet anybody, is to see if I can be helpful to that person and build the relationship from there; if they're helpful to me too, great, but that's never the intention. I haven't seen that as much in up-and-coming, younger people — the hunger is more toward getting a result or an outcome, but the relationship really gets rooted in the connection. I'd encourage younger people to evaluate that, because I think it's extremely helpful now and will be paramount in the future.

Ben Eachus (19:27) - I totally agree. The number of people I've crossed paths with years ago, and then suddenly see on a Zoom call now, is crazy — peers I knew from different companies, doing the same job at the same time. One of our first customers is actually a company I interviewed with and didn't get the job, but I kept in touch with that CEO. It really is a very small industry. And of course we're all tasked with hitting metrics and driving down costs — everyone knows that — but the personal element does make a difference. I think the challenge, especially for businesses like ours, is to blend technology and performance while still making sure customers know they're not just a number, that their business matters. Really well said.

John, I appreciate you taking the time to speak with us here — you're one of the smartest people in the industry, and I appreciate you taking twenty minutes to impart some of your wisdom.

Jonathan Tam (20:40) - You're welcome. Happy to do it.

Narrator (20:42) - Thanks for listening to Well Delivered: How Operators Keep Fulfillment Moving. Follow for more episodes like this, and we'll be back next time.

Other Resources

Lessons on marketplace growth from Fartech | Melissa Marks, Co-Founder of Elfin
Podcast

Lessons on marketplace growth from Fartech | Melissa Marks, Co-Founder of Elfin

Operational Excellence: How To Scale a Business From an Expert | Trevor Weed, President of Albatross
Podcast

Operational Excellence: How To Scale a Business From an Expert | Trevor Weed, President of Albatross

How Ecommerce Brands Build an Omnichannel Strategy | Alexander Lans, Amazon Marketplace Consultant
Podcast

How Ecommerce Brands Build an Omnichannel Strategy | Alexander Lans, Amazon Marketplace Consultant

Start simplifying fulfillment with Flowspace today

Discover how Flowspace can simplify fulfillment and help your brand scale faster, with no long-term contracts or hidden commitments.

Talk to an Expert