In this episode
Hilary Smith, Chief Marketing Officer at Linnworks, argues that most ecommerce operations teams believe their processes are already automated, and that manual work quietly grows right alongside order volume. Speaking with Greg Kihlström, she traces where that hidden work lives — order routing, inventory alignment, carrier selection, listing management — and why the instinctive fix, hiring more people, is a margin squeeze rather than a solution. The conversation covers the shift of AI investment in retail from customer-facing chatbots and personalization into inventory, fulfillment and order management; the distinction between a rules engine that keeps sellers in control and AI that takes control away; what actually breaks first when peak-season volume runs 10X normal; and why a stockout is a brand moment that lands on a CMO who doesn’t own the warehouse.
Key takeaways
- The consequential AI work in retail is operational, not customer-facing. Smith says it is easy to point an AI tool at website copy or social posts, but what keeps sellers up at night is misaligned inventory, stocking out where they shouldn’t, and shipping the wrong product to the wrong customer.
- When operations break, the default fix is headcount — and that squeezes margin. She describes the most common thing she hears from small and mid-sized sellers as wanting to stay profitable while they grow, and hiring to paper over operational error works directly against that.
- Automation and autonomy are different things, and sellers want to keep control. The bulk of Linnworks’ automation lives in a rules engine — an if-this-then-that builder for cases like routing oversized packages to a different carrier, or applying different rules per marketplace — so operators aren’t pulling levers manually all day.
- Spotlight AI flags the work you haven’t automated yet, rather than automating it for you. Launched earlier in 2026, it scans continuously for actions a seller keeps repeating by hand and hands back a rule they can build and implement themselves.
- Order routing and inventory need consistency, not generative output. Smith’s point is that the same thing needs to go to the same place most of the time, which is what conventional rules-based automation is good at — and she notes how routinely “automation” and “AI” now get used interchangeably.
- Peak season is a volume-and-velocity problem, not a bigger-basket problem. With volumes at 10X the rest of the year — considerably higher for some sellers — an inventory error that happens once in a normal month happens ten times in peak.
- A stockout can cost a first-time buyer, not just a sale. Linnworks’ data shows a bad fulfillment experience can prevent a customer from becoming a repeat buyer or even a first-time buyer, which means losing the sale, the repeat sale, and the brand esteem already paid for.
- Sellers do their integration work in October, not November. Linnworks analyzed 207 million orders across 2025 and into 2026 and found roughly a 20% increase ahead of November in sellers integrating new channels and tools — repricing tools, shipping carriers, marketplace connections — so everything is in place before peak.
- Most transactions start on a marketplace, then finish where the buyer prefers. Shoppers go to marketplaces for reviews, shipping times and social proof, so consistent delivery promises across every channel is what earns the trust to bring them back.
- Scaling copy with AI moved her team further from the real story. Rewriting sales materials, the more they tried to scale with AI, the further it drifted from the human account — so they went back to listening to customer calls and interviewing sellers directly.
Chapters
- 0:00 — Intro
- 0:57 — Why manual work grows right alongside order volume
- 2:40 — Hilary Smith’s background and what Linnworks does
- 4:32 — Why AI’s retail investment moved to the back office
- 7:04 — Automation vs. autonomy: the rules engine and staying in control
- 8:07 — Spotlight AI: finding the work you haven’t automated yet
- 10:14 — Peak season is a volume-and-velocity problem
- 11:16 — Stockouts compound at 10X volume, and cost the repeat buyer
- 14:19 — Tech freezes and the 207-million-order peak report
- 16:18 — A stockout is a brand moment: what CMOs owe operations
- 19:11 — Why most transactions start on a marketplace
- 20:34 — AI slop, driverless cars, and what stays human
- 23:29 — One year out, and how she stays agile
Why the back office became the place worth betting on
For years, AI in retail meant the storefront: chatbots, recommendations, personalization. Smith’s read is that the money moved because the expensive failures were never on the front end. Inventory that says available when it isn’t, an order routed to the wrong place, a product shipped to the wrong customer — those are margin events and customer-experience events at the same time, and they scale with the business rather than shrinking as it matures.
Automation is not autonomy, and the difference is control
Linnworks keeps most of its automation in a rules engine: if this, then that, set by the operator. Exceptions for oversized packages routing to a different carrier, different rules for different marketplaces. Smith’s framing is that sellers should not have to pull levers manually all day, but they also should not hand over the levers. Where AI enters is above that layer — watching what is still being done by hand and proposing a rule — rather than replacing the rules themselves.
What Spotlight AI actually does
Smith describes Spotlight AI, launched earlier in 2026, as the opposite of an automation that takes over. It scans for the same manual action being performed repeatedly and surfaces it: you keep doing this by hand, here is the rule to stop doing it. The seller builds and implements the rule. That design is also why it survives a tech freeze — it won’t automate anything without being asked, so it doesn’t move a brand off a set path during peak.
What breaks first when peak volume hits
The first failure Smith names is misaligned inventory across channels, which produces stockouts, which produce a customer experience bad enough to cost future purchases. At 10X normal volume — higher for some sellers — a single instance becomes ten. Her operational recommendation is blunt: get the stack in line well before peak. Linnworks’ own data on 207 million orders shows sellers behave accordingly, with integrations of new channels and tools rising about 20% ahead of November.
Why a stockout lands on the CMO’s desk
CMOs own the brand promise but not the warehouse. Smith’s answer to that gap is alignment across the front and back of house: the operations team knows which deals are running, what is being promoted, which products are indexed, and which are at risk of selling out first. In a properly automated operation, a sales spike should not disturb the warehouse at all — order routing happens automatically and the team does digital picking and packing rather than working off paper or a spreadsheet.
Why buyers start on a marketplace and what that demands operationally
Being DTC-only, or single-channel, no longer matches how people shop. Smith says the majority of transactions start on a marketplace because that is where reviews, shipping times and visible social proof live — and then the purchase happens wherever the buyer prefers. What that demands operationally is consistency in shipping and delivery times across every channel, because the delivery experience is what determines whether the buyer comes back regardless of where they found the ad.
Where Smith draws the human line
Marketing, she argues, is the function most exposed to AI right now, and AI slop is the visible cost. Her team tried to scale a rewrite of internal sales materials with AI and found that the more they scaled, the further the output drifted from the real human story — so they reverted to customer calls and interviews with their own sellers. She compares operational caution to driverless cars taking years to reach the road, because physical goods carry physical risk. The other thing she doesn’t expect AI to own: the strategy of what you sell and who your customer actually is.
FAQ
Why is AI investment in retail shifting from customer-facing tools to operations? Because the expensive failures are operational. Hilary Smith points to misaligned inventory, stockouts, and mis-shipped orders as the problems that threaten profit margin and customer experience — and the default remedy, adding headcount, squeezes margin further.
What is the difference between automation and autonomy in ecommerce operations? Automation executes rules the operator sets — Linnworks’ rules engine is an if-this-then-that builder covering things like carrier selection by package size or per-marketplace exceptions. Autonomy removes the operator from the decision. Smith’s position is that sellers want the first without losing control to the second.
What does Linnworks’ Spotlight AI do? It scans for work a seller is still doing manually and repeatedly, flags it, and returns a rule the seller can build and implement themselves. It does not automate anything without being asked, which is why it can operate during a pre-peak tech freeze without disrupting a set path.
What breaks first during peak season? Misaligned inventory, leading to stockouts and a poor customer experience. Because peak volumes run around 10X the rest of the year, and higher for some sellers, an error that occurs once in a normal month compounds into many lost customers.
What did Linnworks find in its peak-season order analysis? Linnworks analyzed 207 million orders across 2025 and into 2026. One finding: roughly a 20% increase ahead of November in sellers integrating new channels and tools — repricing tools, shipping carriers, marketplace platforms — indicating the preparation work happens in October.
What should a CMO do about operational failures they don’t directly own? Smith recommends unifying front and back of house on what is being promoted, which products are being indexed, and which are at risk of selling out, so the operations team has awareness before volume arrives. In a well-automated operation, rising sales should not change what the warehouse has to do.
About Hilary Smith
Hilary Smith is Chief Marketing Officer at Linnworks. She brings 12+ years of experience driving growth through data-informed strategy across supply chain, ecommerce and FinTech. Hilary thrives at the intersection of brand, product and demand, translating complex customer and commercial data into campaigns that drive revenue, sharpen positioning and move the business forward. Her leadership style is grounded in empathy, curiosity and a deep understanding of how customers discover, convert and keep coming back.
Hilary Smith on LinkedIn: https://www.linkedin.com/in/hilarypaigesmith/
———- Resources ———-
Linnworks: https://www.linworks.com
The Agile Brand podcast is brought to you by TEKsystems. Learn more here: https://aglbrnd.co/r/2868abd8085a9703
The hardest part of martech right now is telling which tools are actually earning their keep. That’s the question running through the MarTech Summit Boston agenda on October 27. We’re attending MarTech Summit Boston on October 27 as a media partner. If you want to join, the code AGILE15 saves you 15%. https://aglbrnd.co/r/c70c250b8d52dff3
We’re proud to be a media partner for #MAICON26 – Oct. 13-15! Learn how AI can power your marketing and business and help you grow smarter. Use code AGILE150 to save! https://aglbrnd.co/r/7fe458ced0f04658
Need to evaluate martech, AI, or CX platforms and don’t know where to start? The Agile Brand Guide List is now in beta. We’re building the most useful, free-to-use, and up-to-date way to assess what truly matters when evaluating platforms. See more here: https://aglbrnd.co/r/f76be3a33cc491bb
Reach your customers with Reddit. Spend $500 in ad spend, get $500 back in ad credit! Learn more: https://aglbrnd.co/r/491818c79fb1873f
Chaser is the only Slack-native project management platform that helps teams turn messages into tracked tasks, automate follow-ups, and maintain team-wide visibility, without adopting another tool. Now integrated with Claude and other GenAI tools. Learn more at trychaser.com and use code AGILEBRAND for a 3-month free trial (normal trial is 14 days).
Start building your own apps with Replit and get $20 off. Learn more: https://aglbrnd.co/r/93531742a7625a20
Enjoyed the show? Tell us more and give us a rating so others can find the show: https://aglbrnd.co/r/faaed112fc9887f3
Connect with Greg on LinkedIn: https://www.linkedin.com/in/gregkihlstrom
Don’t miss a thing: get the latest episodes, sign up for our newsletter, and check out Martechipedia, the wiki for marketing technology: https://aglbrnd.co/r/35ded3ccfb6716ba
The Agile Brand is produced by Missing Link—a Latina-owned strategy-driven, creatively fueled production co-op. From ideation to creation, they craft human connections through intelligent, engaging and informative content: https://aglbrnd.co/r/f1782ab12211f765
Transcript
Linnworks CMO Hilary Smith on the manual work hiding inside “automated” operations
==================================================================================
[00:00:57] Greg Kihlström: Hi, I’m Greg Kihlström, host of the Agile Brand, and here’s a question for you. Most retail operations teams believe their processes are already automated, so why does the manual work seem to grow right alongside the order volume? The brands that handle scale well aren’t just faster. They can see where the work is actually happening and act on it before the customer feels it. Today we’re going to talk about why AI’s most consequential move in retail right now isn’t customer facing. It’s happening inside inventory, fulfillment, and order management. What peak season really tests, not bigger baskets, but whether your operating model holds together under volume and velocity, and the difference between automating a task and building a system that spots the problem before anyone reports it. To help me discuss this topic, I’d like to welcome Hilary Smith, CMO at Linnworks.
[00:02:29] Greg Kihlström: Hilary, welcome to the show.
[00:02:31] Hilary Smith: Hi Greg. Thanks for having me.
[00:02:32] Greg Kihlström: Yeah, looking forward to talking about this with you. Before we dive in though, why don’t you give a little background on yourself and your role at Linnworks?
[00:02:40] Hilary Smith: Yeah, absolutely. so I’m very fresh, a very fresh face to Linnworks. I just joined the team in March of this year. but I joined because I have a very deep e-commerce background. So prior to this I spent six years at Jungle Scout, really diving into, e-commerce sales data, specifically Amazon sales data, which I love. I’m a bit of a data nerd. and I’ve been working with small to mid-size businesses for years. I also worked in, fintech in finance automation. So small business, e-commerce, and automation are really the three buzzwords in my background. [chuckles]
[00:03:18] Greg Kihlström: Yeah. Yeah. Well, and, and to give a little context here, for those that are l- are less familiar with Linnworks, can you give a little background? You know, what’s the company’s mission and who’s the, who are the primary customers that you’re serving?
[00:03:30] Hilary Smith: Yeah, absolutely. So Linnworks, we’ve been around for quite a while. So we actually were founded in 2005. Our founder, was actually selling, I think it was used DVDs or CDs on eBay, like a very 2005 story and [chuckles] and found it was like frustrating to wait for labels to print. And so he coded up this like lightweight tool to like automate the process of printing shipping labels. And from there, the tool really just grew and expanded. He actually put the tool up for sale in like the eBay software section, which I think they had at the time. And then it grew and was informed by, sellers just like him. We’re now, a multi-channel e-commerce operations platform, so we help with the inventory and order management specifically, as well as listing management, shipping carrier management, and the overall process of multi-channel selling for small to mid-sized businesses. we have headquarters in the UK as well as the United States.
[00:04:32] Greg Kihlström: Great. Great. So yeah, let’s dive in here. And want to start with, one of the things I teed up in the intro and, talking about operations and, and AI. And, you know, for years AI and retail mostly meant customer facing things, you know, chatbots, recommendations, personalization. Now it seems like at least a, a significant part of the investment is moving into more operational things, inventory, fulfillment, order management. What has changed? You know, there’s certainly plenty of talk about the customer facing stuff still, but you know, what changed to make the back end the place that’s worth betting on?
[00:05:14] Hilary Smith: Yeah, that’s a great question. I feel like it’s pretty easy to look at an AI tool and be like, “Oh great, this can write my website copy,” or, “This can help me,” you know, make it a little easier to post things on social media, right? All of those things that you do like around your business to promote it. But the really hard stuff and the stuff that keeps people up at night over their profit margins or over bad customer experiences is unfortunately when inventory is misaligned, you are stocking out where you shouldn’t, things get shipped to the wrong place, you mis-ship the wrong product to the, to the, you know, wrong customer. And you start to look at all of your operations and think that the only way to solve it is to hire more people. And that is a huge margin squeeze and that is so important. That is like the everyday most important thing that I hear small to mid-size businesses talk about is they want to make sure that they can be profitable while they grow.
[00:06:11] Greg Kihlström: Yeah, yeah. Well, no, and I think it’s a, you know, the, I think the more glamorous thing is the fancy customer-facing stuff, but to your point, it’s the business needs, needs to run, needs to be more efficient and, and keep up and, and everything. And so, you know, some of the, some of the more substantive in- investments are in that, that operational, those operational aspects. And, you know, I think there’s, there’s an important distinction that, that you’ve made between automation and autonomy as well. And so maybe if you could talk a little bit about, you know, what, what exactly that means, and, and where does operational AI cross the l- that line into doing something a human couldn’t even do at all and, and where is it, more augmentation?
[00:07:04] Hilary Smith: Yeah, absolutely. We like to make sure that the businesses that we work with are still in control of what gets automated. So the bulk of the automation inside of our platform lives in something called the rules engine, which is really, like, an if this, then that rule builder that automates every rule that you set. a good example is, like, exception for packages of certain sizes, they’ll ship to, through another carrier versus another one just based on, like, the size of that shipment, right? Or maybe different rules set up by, by different marketplaces that you’re selling on. All of this is to make sure that you still have a degree of control, but you don’t have to, like, manually be pulling the levers all day every day, because that’s the thing that is the most challenging and eats away at that, like, productivity that really levels up to your profitability, frankly. and then on the AI side, where we’ve brought AI into the business in a way that [laughs] yes, the operational side is, like, somewhat unsexy, but we, we like to think it, it has a fairly sexy output-
[00:08:07] Greg Kihlström: Right, right
[00:08:07] Hilary Smith: … in what it ultimately does is instead of, just, like, taking the control away from you, we have a tool that launched earlier this year called Spotlight AI that actually highlights the things that you haven’t automated yet. So it will, like, scan for those all of the time and then say, “Hey,” like, “you keep doing the same thing manually over and over again. How about you stop doing that?” And then give you the rule that you can build and implement yourself. So that it’s doing the work for you. It’s doing the automation. It’s flagging where you’re not being productive, but it’s not taking the control away from you that you may still want or need.
[00:08:41] Greg Kihlström: Yeah, yeah. Well, and I think the, y- the important thing to highlight also about the rules engine is, there are many things that generative AI is, is great at doing and everything, but there are a lot of things that, I think to, to what you’re saying, you don’t, you don’t need rules and, and processes and things invented by an LLM [laughs] sometimes too, right? You know, so it’s like, it’s like the right use of… You know, ’cause a rule, if this, then that is technically AI. It’s not the AI we talk about a lot, but it’s the right tool for the job because you need things to be precise and, and repeatable. And then where gen AI is useful and, and, you know, can augment, it’s utilized. And, you know, that, to me that i- that goes, I mean, certainly there’s talk about tokens and, and stuff too, but it’s also just about dependability, right?
[00:09:32] Hilary Smith: Yeah, exactly. When it comes to order routing and inventory, it’s often, like, the same thing needs to go to the same place the majority-
[00:09:41] Greg Kihlström: Right
[00:09:41] Hilary Smith: … of the time, right? You just need that consistency. That’s where, like, more conventional automation, which [laughs] it, it’s funny to see, like, as someone who considers themselves quite AI native at this point, how much automation as a, as a thing and AI as a thing are getting just, like, intermingled together all the time. But it’s that, like, level of consistency, repeatability, and control that despite the fact that we would love to probably have AI run things for us all the time, you still need that certain lens and consistency that is set by you.
[00:10:14] Greg Kihlström: Yeah. Yeah. it’s, I mean, it’s the right, right tool for the job, right? So it’s, you know… Yeah, definitely. and so, you know, we’re coming up on, well, soon enough we’ll be coming up on a peak, shopping season. I guess, I mean, it’s probably already started, hasn’t it? [laughs] but, uh-
[00:10:32] Hilary Smith: Well, I know. I keep waiting. I’m like, I’m bracing myself for the Black Friday email-
[00:10:37] Greg Kihlström: [laughs]
[00:10:37] Hilary Smith: … and I’m, like, placing my bets. I’m like, I think I’ll get my first one on maybe October 10th is, like, where I’m gonna place my bet. I don’t know if you wanna place a bet too, but [laughs]
[00:10:44] Greg Kihlström: Right. Right. I know. but it, either way, yeah, soon enough it’ll, it’ll be here and, and, you know, the, the pressure from that, like, sheer volume and, and velocity of orders, you know, there, there’s a lot of pressure there, not, not even necessarily larger individual purchases. So, know, what does that change about how a brand should prepare for it, and, which parts of traditional peak break first under some of these kinds of volume and, you know, volume and velocity loads?
[00:11:16] Hilary Smith: The biggest pitfalls we see is misaligned inventory, and that causes stock-outs, which causes a poor customer experience. Like, if you are trying to sell something on your own website as well as another channel, and that buyer goes to buy it from your website, you say that you have it available, but when push comes to shove, you know, ship goes to, to f- to, to order or order goes to ship, and it’s not available, that creates a really bad customer experience. And we actually find in the data that, like, it can prevent that customer from becoming a repeat buyer or even a first-time buyer, right? That bad customer experience sets up everything, and you could potentially lose that customer forever. So if you look at peak season, where, like, volumes are 10X what they are for the rest of the year, and for some, for some of our sellers it’s much higher than 10X. Like, it’s really, really high. that kind of error, it just compounds, right? If you think about it happening once, all of a sudden it could be happening 10 times, and then you’ve got 10 people that aren’t coming back, right? And that’s not just losing the sale, but it’s losing the repeat sale, and it’s losing the esteem that your brand has already spent a lot of time gathering.
[00:14:19] Greg Kihlström: One way of trying to solve for just keeping things consistent and, and, you know, not making last-minute changes, a lot of retailers implement a tech freeze well before Black Friday, just trying to keep a handle on things. How does a tool like, Spotlight AI function within that constraint? You know, does it provide intelligence on the existing stack? Does it require changes when teams are most risk-averse? You know, how, how does it work, and, and still keep things optimized?
[00:14:52] Hilary Smith: Yeah, for sure. It’s a good question. It’s not going to, it’s not gonna automate something without you asking it to, so it’s not going to, take you off of a set path when you’re in, when you’re in peak season. So it’s not gonna disrupt anything. Our recommendation would be, make sure that you have your tech stack in line and ready to go well before peak. We’ve actually just released… We’re calling it, like, our, like, peak season data report, for lack of a better word. My, my, my team might kill me ’cause I can’t remember the exact title of our, of our-
[00:15:25] Greg Kihlström: We’ll, we’ll put a link in the show notes. How about that?
[00:15:27] Hilary Smith: Yes, that would be amazing. It’s, it’s got some really great data. So we analyzed 207 million orders across all of 2025, as well as some data into 2026 to just look at order volume and trends over the course of the year to see how and when things are compounding across not just, like, orders, GMV basket size by marketplace, but also, like, what technology are people deploying ahead of peak season to be ready. And one of the things we see is that there’s actually, like, a 20% increase ahead of November for people who are integrating new channels, integrating new tools. Like, picture all the, the tools that a, a, a platform like Linnworks would integrate with, like your repricing tools, different, like, shipping carriers, like various different types of platforms, and people are doing that in October well ahead of peak so that they have it ready to go by November.
[00:16:18] Greg Kihlström: Yeah, yeah. So I wanna get back to something you were talking about earlier and just that, that impact that, you know, certainly good, good customer experiences have, have a p- have a, an impact. But, you know, some of those negative moments, you know, stockout, fulfillment error, and, you know, you were talking about how this can prevent even a first-time customer, let alone a, a repeat customer from, from happening. A lot of times, you know, there’s a lot of CMOs listening to this show. There’s a lot of CMOs out there that are building brands, and, and they’re tasked with that, that brand promise part of, of the equation, but they don’t own the warehouse. You know, how do you… Y- you know, what, what should CMOs do to, to kinda connect the dots there?
[00:17:05] Hilary Smith: Yeah, that’s a great question. Often what happens at the front of the house, like your marketing, your storefronts, your listing, all of that can directly impact the back of the house, right? That’s your operations. That’s your, you know, your warehouse team, everything. So of course your team needs to be unified across the board on what deals you’re running, what you’re promoting, which products that you’re indexing on, which products are maybe at risk of selling out first just so that there’s some awareness there. Ideally, you’re gonna have enough healthy stock ordered well ahead of time to be able to make it all the way through peak, but, you know, viral moments catch us all, right? [laughs]
[00:17:45] Greg Kihlström: Right, right. [laughs]
[00:17:45] Hilary Smith: So to make sure that you are well aware and prepared for that season and that there’s consistency in your messaging and awareness across your entire team. But in an ideal world where things are automated, your warehouse should not be impacted by an increase in sales. There should already be the tooling in place that is gonna help with order routing automatically, and then they’re just gonna have to go and do the, you know, the, the digital picking and packing versus trying to run everything off of paper or running everything off of a spreadsheet. There is gonna be consistency in what people are buying, what you have available in your warehouse, and then which carriers and how it goes out and when it goes out and to whom it goes out to.
[00:18:29] Greg Kihlström: Yeah, yeah. Well, and I think ano- another key part on that that you, you kinda touched on there is just- You know, consumers are jumping around from channel to channel. They’re, they’re– they wanna access things exactly when and where they, and on what device, and so on and so forth that, that they want to at the, at the moment. So not only do we have rising order volume, but we also have potential inconsistency where, you know, you, you buy something through social and it feels almost like a different company than if you go to the website. What is– what has to be true operationally for a brand to really feel like one company at, at that scale, you know, from a multi-channel perspective as well?
[00:19:11] Hilary Smith: Yeah. That’s a great point. I think what is especially important is being all of the places that you think your buyer will be. So being, one channel or being DTC just isn’t enough anymore for the way people shop and the way people buy. We know, with our own data as well as data that we see from other providers, is that the majority of transactions actually start on a marketplace because people wanna go and see the reviews. They wanna, they wanna see what the shipping times were. They wanna see what the quality of the product looks like. They wanna be able to see the, you know, the real social proof of, of someone who’s got the lived experience of buying that product. But then they wanna go and buy it where they are and how they prefer to shop. So one of the things that they’re gonna be looking for is shipping delivery, shipping and delivery times. So if you’re able to make that consistent across all of your channels and you’re able to make sure that, hey, if I’m selling through this channel, it’s gonna have a very similar delivery time and delivery experience, and have that consistency, that is, is key for shoppers because it, it helps increase that trust, right? It’s gonna mean that no matter what, they’re gonna wanna go back to wherever they find your ad. Maybe they’re getting an Amazon ad or maybe they’re discovering you in a, you know, a sidebar, on their TikTok platform, right? They wanna be able to, trust the experience that they’re gonna get everywhere that you sell.
[00:20:34] Greg Kihlström: Yeah. Yeah. And you know, it’s certainly more, more and more of this, can be automated, will be automated and, you know, so, so systems are gonna continually handle, more of this. But, you know, from your standpoint, what should stay something that humans are, are driving and, you know, certainly there, there’s human in the loop, but there’s also humans still managing [chuckles] something. So, you know, where, where’s the line that you wouldn’t want an operations or even a marketing team to hand over even once technology can technically handle it?
[00:21:12] Hilary Smith: Yeah. I mean, as someone in marketing, I, I don’t think we hear about any industry that should be informed by AI as much as marketing right now.
[00:21:20] Greg Kihlström: Right. Right.
[00:21:21] Hilary Smith: So much of all of the copy that you read, you’re hearing the term AI slop everywhere. and that’s something that really can only be avoided if there is a human touch and a human lens on it. AI is not necessarily at the stage right now where it can compete with real human thought and, like, the spirit behind it. We just went through, like, a pretty incredible exercise where we were, like, rewriting, some of our sales materials internally, and we were like, “Oh,” like, “let’s see how much of this we can scale with AI.” And the more we tried to scale more of it with AI, the further it got from, like, the real human story. And so we ended up having to do it the old-fashioned way and saying, “We’re gonna sit down and we’re gonna listen to customer calls, and we are gonna interview every seller on our team and find out really what customers are facing every day, so that when we go into those conversations, when we are creating content, we are getting the real lived human experience, not just the AI layer that’s been applied over that real human experience, so that we can get at the heart and, like, the spirit of that.” So that’s, like, a bit of a soapbox moment for me but, [chuckles] but to get back to your, to get back to your original question, like, the operational stuff is great to, to automate with AI. You still wanna have a certain degree of control there. it’s just like, you know, driverless cars took a long time to get on the road, right? Because they deal with the safety of physical things and people, right? We also feel the same about physical products. We want the small business owners that we work with, some of whom are still risk-averse, in some regards, right? We need to make sure that they feel comfortable with what we’ve, we’ve offered them from an automation and AI standpoint, and that they still have that degree of control. Another piece that I don’t think will ever be truly AI-owned is, like, the strategy in what you sell and who your customer really is. Ultimately, you know as a person what is gonna resonate with your buyer. and there is just so much, like, fun and spirit that goes into adding new products to your catalog.
[00:22:54] Greg Kihlström: ove it. Well, Hilary, thanks so much for joining today. I got a couple last questions for you as we wrap up.
[00:23:29] Greg Kihlström: the first one: if we were having this interview one year from today, what is one thing that we would definitely be talking about?
[00:23:36] Hilary Smith: If we were having this one year from today, I really hope we will be talking about how we’ve layered in even more exciting AI functionality into Linworks, as well as how we’re layering in some, like, more broad data across our, across the brands, that we see and that we work with. So, I hope our product team’s not listening. We can just cover their ears. But I would love to see an even better layer of, like, benchmarking data and analysis in the product.
[00:24:06] Greg Kihlström: Yeah. Love it. Love it. And last question: what do you do to stay agile in your role, and how do you find a way to do it consistently?
[00:24:14] Hilary Smith: Yeah. Well, I do rely on, my best friend Claude [chuckles] fair amount to stay agile in my role. I will often test a lot of my assumptions against, like, two different projects that I’ve built in Claude that analyze all of our, like, customer content, as well as one that analyzes, like, a lot of our marketing data. So it’s helped me be really agile in my role because I don’t spend a ton of time, like, manually doing reports anymore. I can get answer-based, answer-based answers, I guess is probably the way to put it from my own data. That’s what I like to do and how it helps me, like, stay ahead. I would say another way to stay agile, uh- I rely on my team for a lot. They’re really incredible. So, I’m lucky to have, like, really great people behind me that I work with that I feel really confident every day that we can just go in and, like, get stuff done.
