How Taylor Swift Helped Optimize Inventory with Chantal Schweizer of Pivotree

Key Takeaways:

(02:12) Taxonomy helps structure product data for better organization and searchability.
(05:20) Spreadsheets remain the default tool for managing business and product data.
(10:29) Taxonomy creates order from chaos, whether in nature or commerce.
(12:35) Spreadsheets can’t handle large-scale operations.
(16:35) Understanding SKU count and categories reveals the scale of data challenges.
(19:16) Discovery projects identify data issues before implementing long-term solutions.
(21:58) Automation creates efficiency but often sparks fears about job security.
(25:10) Buyers need fast, intuitive search tools to find products easily.
(29:59) An ontology connects multiple taxonomies to create structured relationships.
(34:55) A retailer adjusted inventory based on Taylor Swift’s concert schedule.

Chantal: One of our clients set up their stocking mechanisms to follow Taylor Swift’s concerts. Wherever her concerts were, they would focus on the store in that area and increase their stock. It worked.

Aaron: Welcome back to the B2B Uncut podcast, sponsored by OroCommerce. I am your host, Aaron Sheehan, and with me today is Chantal—am I saying that right? Correct me if I’m wrong.

Chantal: Perfect. Chantal Schweizer.

Aaron: Okay, great. Chantal has an awesome resume and background in the B2B space, especially on the data side. She has quite a pedigree.

I think you might be the sixth person this year—or in the last 12 months—with a Grainger connection. That name has come up a lot. We’ve had a lot of McMaster-Carr folks, and now we’re getting more Grainger.

Another famous name. Always nice to mix it up. So, Chantal, I’d love to give you a chance to introduce yourself. Tell us—what do you do?

Chantal: Sounds good! I’m Chantal Schweizer, the Practice Director at Pivotree for Strategic Data Services. What it really comes down to is, I have a team of taxonomists, and we help people organize their data—especially product data.

We build taxonomies to structure attributes that describe products, create web taxonomies to make things more findable, and everything in between. We fill in the data, normalize it, make it all pretty, tie it up with a bow, so people can have a great customer experience on their websites.

That’s what I do now. Before that, as you mentioned, Grainger—I got my start there, and that’s really where I dove into product data. I have an art degree, and I didn’t really know what to do with it.

I found out you can’t make a lot of money with an art degree.

Aaron: So I’ve heard.

Chantal: Yeah, right? Oh man, that was a tough one. Art, I still love you. But there is an art to data—I see it come into play all the time.

I got my start at Grainger, fell in love with organizing the chaos that is product data—herding all the cats, getting data from different manufacturers into our system. I did that for about eight years.

Then I moved over to Schneider Electric, where I was working for a manufacturer, getting data ready for distributors. So I got to see both sides of the fence.

Eventually, I moved into consulting—doing PIM implementations, mostly taxonomy work, and then managing taxonomy teams. Now, I help companies get their product data looking pristine and lovely.

Aaron: That is pretty awesome. I’m curious—you said team of taxonomists, which has a nice alliteration to it. It made me wonder, what’s the plural?

You have a murder of crows, a herd of something… what’s the plural for taxonomists? We need to invent something.

Chantal: We can have fun with that. I don’t know—I’m going to have to think about it and come up with something.

Aaron: If we do, I promise, listeners, we will put it in the show notes. And if not, follow us both on LinkedIn—we’ll post it there.

You mentioned Schneider Electric—actually, we had Steven from Schneider on about a year and a half ago. Big name in the industry. In fact, I think he might be coming back on soon, which would be awesome.

It’s interesting, your background. If people have been listening to this podcast—by the way, please subscribe, mash the like and subscribe button, leave us a nice review—you’ll notice a pattern.

Last year, we had a run of guests from the supply chain world. This year, the theme seems to be product data. We’ve had a lot of PIM, product data folks.

I haven’t had anyone on who actually identifies as a taxonomist, though, so that’s pretty awesome as a job title.

You know Jason Hein, of course. There’s a lot of overlap between you guys—I see you following each other and engaging on LinkedIn all the time.

The problem this raises for me is, I’ve now asked all the obvious questions about product data, PIMs, and taxonomy. So instead of asking you the usual, “What’s a category?” type questions, I’m going to have to pivot.

So, I was looking at your LinkedIn, and I noticed you’ve spent about 20 years in B2B—which means 20 years fixing other people’s spreadsheets.

There’s a saying, all business software competes with Excel. No matter what we’re building, Excel is always the real competition.

That is definitely true for us on the eCommerce platform side. It’s absolutely true on the PIM side.

So, why is the spreadsheet still the default medium for schlepping data?

Chantal: I think it’s just what we’re used to. We get so good at the different functions—concatenating things, using VLOOKUPs, pivot tables—we can be wizards in there, work magic.

Then, when you switch to a new system, it’s like you have to relearn everything all over again. Even something as simple as going from Excel to Google Sheets—there’s a huge learning curve.

Going from one system to another? Forget it. So I think we’re all just so comfortable in Excel that we don’t want to step away from it.

It’s kind of like—I grew up driving a stick shift, and then one day I had an Uber pick me up, and I couldn’t even figure out how to open the door.

Aaron: That is a great metaphor.

I wonder if AI is going to become the automatic transmission in this scenario—where you no longer need to worry about VLOOKUPs and pivot tables because AI will just do it.

Instead of writing macros, you’ll just ask for the data—“Hey, show me this,”—and it’ll just figure it out.

No more messing around with formulas. What a weird mix of terrifying and amazing.

You spend years mastering something super niche—like, forbidden knowledge of how to manipulate spreadsheets—and then suddenly it’s irrelevant.

And I don’t know how that makes me feel. It’s like growing up driving a stick—which I did—and now I haven’t seen a new manual transmission car for sale in… what, 20 years?

Used, maybe. But new? Can you even buy one?

Chantal: Probably, but you’d have to really hunt for it.

Aaron: Listeners—have you bought a manual transmission car in the last 15 years? You may be eligible for a class-action settlement.

Chantal: [laughs]

Aaron: You mentioned art earlier, and when we were talking before recording, you have a lot of hobbies. That really stood out to me because—let’s be honest—a lot of people in B2B are kind of… boring.

So your list of hobbies was actually pretty interesting.

You mentioned art, watercolors, nature, archery, hiking, camping—I know I’m missing something. So here’s the obvious podcast question: how many spreadsheets do you maintain to manage all these hobbies? And what’s the connection between the hobbies and the day job?

Chantal: Oh, there are definitely a lot of spreadsheets! We have ones for managing travel, investments, household tasks—all that stuff.

For the nature side, I do a lot of conservation volunteering. One of the projects I worked on for a couple of years was focused on phenology, which is the study of seasonal changes in plants and animals.

I was in charge of tracking wildflowers at a specific nature preserve. Every week, I’d hike through the site and record what was blooming, how many blooms there were, and how long they lasted. I also logged temperature, rainfall, and other conditions to track long-term patterns.

And, of course, it all went into a spreadsheet.

Aaron: That is kind of amazing. And terrifying.

Had no one thought of using SQL for this data?

Wait—are you telling me that all of this data was just sitting in an Excel spreadsheet?

Chantal: Yep! Some of the older records were handwritten, but eventually, they were digitized. And as far as I know, they’re still being maintained in spreadsheets today.

Aaron: Oh, that’s awesome. I actually have a similar story—this is totally irrelevant to the topic, but it’s my first chance to bring it up.

I studied geosciences in college, and one of our projects was digitizing cemetery data.

For decades, local historians had been writing down where all the grave plots were—in some cases, in the middle of nowhere in a field. They compiled everything into these massive paper books.

We had to digitize that, go out with GPS devices, and turn it all into a database. The county wanted it online so residents could look up records.

Chantal: Oh, that’s fascinating!

Aaron: Yeah, it was a mix of fieldwork and data entry. The funny thing is, unlike your wildflowers, the cemetery data didn’t really change.

No seasonal updates needed, thankfully.

Chantal: [laughs] Yeah, that’s definitely a key difference. But both projects had the same fundamental challenge—categorizing and preserving important information for long-term use.

Aaron: So, switching gears—I was looking at your LinkedIn, because I troll everybody’s profile before we record.

I like to get a sense of the person, or at least the version of themselves they put out into the world. Those aren’t always the same thing, I’ve found.

I think in your case, they probably are.

But one thing that stood out to me—you mention digital data transformation a lot. It’s in your job descriptions, your bio, it comes up over and over again.

I’ve seen digital transformation, I’ve seen data transformation, but digital data transformation—that’s a new one for me.

You said it so many times in your bio, so I assume you meant it on purpose.

What is that? Are you trying to get people to stop using spreadsheets, or what’s the deal?

Chantal: That’s pretty much what it is.

It’s about getting people out of their homegrown legacy systems—whether that’s spreadsheets, old databases, or, in some cases, even paper catalogs—and moving them into a mature system that lets them actually manage their data in a structured way.

Something that makes it easier to govern, enrich, and distribute data to all their channels.

Aaron: Makes sense. Okay, I’m going to play devil’s advocate here.

I’ve had this conversation many times with people who run long-standing distribution businesses.

They’ll say, “What’s wrong with my homegrown system? It’s worked for 40 years. It worked for my dad. It worked for my granddad. We run this business on little green notebooks. Why should I spend money on something new?”

How do you answer that?

Chantal: We hear that all the time. The biggest thing is scalability.

If you’re happy managing a few thousand products, sure, you can stick with spreadsheets. But if you want to scale up—if you’re talking hundreds of thousands or millions of SKUs—you’re going to hit a wall.

Spreadsheets aren’t designed for that kind of volume. They’re not easy to govern over time, and they don’t give you the reporting, workflow, or automation capabilities you need to manage complex data efficiently.

Aaron: And there is a row limit in Excel. I just want to point that out.

Chantal: Exactly!

Aaron: Do you know what the Excel row limit is?

Chantal: I think it’s around a million?

Aaron: It’s a little over a million. It’s a very specific number, based on how memory is allocated, but yeah—it’s a million and change.

So yeah, you can’t exactly scale to millions of products using Excel.

Chantal: You could try with multiple spreadsheets, but then you have another problem—you’re managing multiple sources of truth, and they’re bound to get out of sync.

Aaron: Oh, wait, I’ve got it!

What if you had multiple spreadsheets with a spreadsheet that tracks the multiple spreadsheets?

Chantal: [laughs] You joke, but I’ve seen it happen!

Aaron: See? Who needs a PIM when you can just manage your PIM inside a spreadsheet?

Aaron: Okay, shifting to another topic that’s near and dear to me—food.

I was scrolling LinkedIn and saw your post a couple of weeks ago. You posted a picture from your annual pantry reorganization.

First off, I love that.

Second, I promise this is going somewhere.

What’s the first step of a proper pantry reorganization?

Chantal: The first thing—and I actually had a before picture, but I shamed myself out of posting it—is clearing off the floor.

Quick context: I have four boys, ages 11 to 14. They are evil little pantry goblins.

They open bags of cereal in there, and inevitably, the cereal falls out and gets crushed. They leave half-eaten sleeves of Ritz crackers lying around. They put things back wherever is most convenient for them.

So, zero data governance happening in my pantry.

Aaron: Shocking.

Chantal: Right? It’s total chaos.

So first, I take everything off the floor, put it on a table, sweep it all up, and get the space clean. Then, I empty one shelf at a time, categorize everything, and start putting things back where they belong.

For example, I’ll designate one shelf for pasta, pasta sauce, and olive oil. I go through the whole pantry, collect those items, and put them on that shelf.

Then I move on—next shelf is for tortillas, refried beans, salsa, and so on.

I keep going systematically until everything is organized, and it all looks perfect… for about three days. Then the pantry goblins strike again.

Aaron: That makes sense.

And I’ll just say—two nights ago, I did take a half-empty sleeve of Ritz crackers and put it somewhere it shouldn’t have been.

So maybe I am a pantry goblin.

But I’m fascinated by how similar this sounds to what you do professionally.

You’re essentially going through the same process—taking inventory of what’s there, categorizing it, and structuring it so it’s easier to navigate.

So when you walk into a new client engagement, what are the first questions you ask to figure out how much of a mess you’re dealing with?

And how much does it resemble your pantry reorg?

Chantal: Oh, it’s very similar!

I usually start by asking a few key questions:

• How many categories do you have?
• How many SKUs do you have?
• What system are you using to manage your product data?
• What governance processes do you have in place today?

That gives me a rough sense of size and complexity. Then I ask about pain points—what’s keeping them up at night when it comes to product data?

If they’re using spreadsheets or an outdated system, I know they probably have duplicate data points scattered across multiple files. That tells me they need a centralized repository.

If they’re a distributor struggling with long vendor onboarding times—getting inconsistent product data from suppliers in different formats—they probably need a vendor portal to help streamline that.

And sometimes, companies go through all the effort to clean up their data, but they don’t put governance in place.

So three years later, they come back, and it’s a mess again.

Aaron: So basically, somebody put the Ritz crackers in the wrong place.

Chantal: Exactly!

Aaron: The parallels here are eerily close.

So whether it’s product data or a pantry, it sounds like the process is the same—you start by defining the desired structure.

How many jars of olive oil do I have? How many different kinds of pasta sauce? How many bananas?

Then, you take stock of what you actually have and build a roadmap to get from chaos to order.

Chantal: Yep, exactly. Current state, future state, and here’s the data path to get there.

Aaron: What’s the biggest SKU count you’ve ever had to clean up?

Chantal: 7.5 million. That was a beast.

Aaron: Oof. We actually have a customer with 7.5 million SKUs. Probably not the same one, but maybe we should talk offline and compare notes.

What industry?

Chantal: Electrical products.

Aaron: Okay, totally different company, then. Our guy is in forklift parts.

Chantal: Gotcha.

Aaron: Have you ever had to walk away from a prospect or an engagement because they weren’t ready to do the work?

Because this is a lot of work.

Chantal: It is. It’s a ton of work.

A lot of times, before diving in, we’ll suggest starting with a discovery phase—a short project, maybe a month—where we analyze their data, do stakeholder interviews, and figure out exactly where the issues are.

These interviews often turn into data therapy sessions where people vent about everything driving them crazy.

We help them map out what’s wrong, what needs to be fixed, and how to build a business case to secure funding. Because half the battle is getting leadership to actually approve the budget for a full data transformation.

And yeah, there have been times where we get through discovery, present the findings, and realize… they’re just not ready.

Aaron: Like, digitally not mature enough?

Chantal: Exactly. They don’t have the processes in place, or they don’t really want change.

I’ve had cases where we start building an attribution taxonomy, and they push back—“Well, we’ve always done it this way.”

It’s like the spreadsheet conversation all over again.

Most of the time, we can get them on board. But every once in a while, you run into someone who just does not want to move forward, no matter how much you try to show them the benefits.

Aaron: Yeah, this goes back to that earlier conversation.

You’re asking people to switch from a stick to an automatic. Or from Excel to a proper system.

And there’s always a tax to adopting a new way of working. People get less efficient at first because they’re learning.

It’s the same as hiring—when you add people to a team, you don’t instantly become more productive. There’s a dip before you get better.

I see this in tech and operations all the time.

Is it the same for product data? Like, is there a point where a company is so busy onboarding the new system that it feels like they’re sliding backward?

Chantal: Oh, absolutely.

And on top of that, there’s always a little fear—like, “If this gets more efficient, does that mean I lose my job?”

If you have a whole team whose job is gathering and cleaning supplier data, and suddenly we automate half of that process, what happens to those roles?

Aaron: Right.

Chantal: We try to reinforce that the goal isn’t to take jobs away—it’s to make your job more efficient.

So you can focus on more important things—like auditing data, improving processes, or making your user experience more successful.

Aaron: Okay, let’s pause there because you just said something really important—user experience.

I got into a little debate online the other day about this. I think customer experience drives everything.

Retention, revenue, satisfaction—people want to work with companies that make things easy.

And in eCommerce, good product data helps you sell stuff. That’s true whether it’s a paper catalog or an online store.

But here’s the problem: supplier data is often… well, there’s a technical term for it—crap.

Chantal: [laughs] Yes.

Aaron: I’ve talked to so many distributors over the years who get hundreds of CSV files from suppliers, all in different formats.

Some have missing fields, some concatenate values, some get truncated mid-sentence.

So, if I’m a buyer shopping on a distributor’s site, what makes for a good user experience in that situation?

Chantal: A few things.

First, findability. Buyers need to be able to get to the right product quickly, whether they’re navigating through a taxonomy or using search.

Second, good filters. If I’m looking for chapstick, and I want vanilla flavor, I need a flavor filter.

If the data isn’t structured properly, that filter won’t exist—and I won’t be able to find the product I actually want.

Third, completeness. The product page needs to answer all the buyer’s questions.

If three different suppliers sell chapstick, and one provides complete data while the others don’t, guess which one shows up when buyers filter for vanilla?

The one with all the data. The other two are invisible.

Aaron: So, the company with better product data wins.

Chantal: Every time.

Aaron: That’s something I wish more manufacturers and suppliers understood.

If you’re a supplier listening to this—please provide complete data. And if you’re a distributor getting bad data, invest the time to clean it up.

Because another problem I see all the time is inconsistency. Maybe everyone has a vanilla-flavored chapstick, but one writes “Vanilla” with a capital V, another writes “vanilla” lowercase, another writes “VANLLLA” with an extra L, and one writes it in Spanish.

Chantal: [laughs] Yes!

Aaron: And an eCommerce system does exactly what you tell it to do.

So if those are all separate values in your data, guess what? You now have five different filter options for vanilla.

Chantal: Exactly. This is why normalizing data is so important.

Aaron: Right. Because if a buyer can’t find the product, they can’t buy the product.

And I’ve seen suppliers with great products—competitive pricing, high quality—but they lose sales because no one can find them in the search results. Meanwhile, the companies that do invest in product data keep winning market share.

Chantal: Yep. Having complete, structured, and standardized data is a competitive advantage.

Aaron: Perfect.

Oh, and speaking of competitive advantages—this reminds me of something wild I heard.

I was talking to someone recently, and they mentioned that a company was literally tracking Taylor Swift’s concert schedule to plan their inventory.

They’d increase stock in stores near her tour stops, and apparently, it worked.

Chantal: That’s actually one of our clients! They set up their stocking mechanisms to follow Taylor Swift’s concerts.

Wherever her concerts were happening, they focused on those locations and increased stock.

Aaron: That is brilliant.

My SEO team thanks you so much for mentioning Taylor Swift, by the way.

We’ve been trying to get her name into this podcast for two years, and now it finally happened.

I know this episode is going to perform amazingly well when the Swifties find it.

Chantal: [laughs] Glad I could help!

Aaron: My 13-year-old might actually listen to this episode now.

Aaron: Okay, I want to ask you one final serious question, and then I have a surprise non-serious question to wrap things up.

Chantal: Sounds good.

Aaron: So earlier, before we started recording, you used a word that I recognized but didn’t fully understand in the context you were using it.

You said ontology. Now, I know that word from philosophy class—like, the study of being or whatever. But you were using it differently, and I was like, what is she talking about?

So, when you say ontology, what do you mean?

Chantal: An ontology is really a collection of taxonomies that link together.

So, you have different taxonomies that classify different entities.

For example, in my wildflower project, I had a taxonomy that classified the flowers and another that categorized the nature preserve sites.

An ontology would link those together—so you could say, this nature preserve blooms these specific wildflowers at this time of year.

If you expand that concept, you can link even more data points together and start seeing relationships you wouldn’t otherwise.

And if you take it even further, the next level after an ontology is a knowledge graph.

Aaron: Okay, hold up.

I’ve heard knowledge graph a lot, but I’ve never been 100% sure what it actually means.

I think I know, but I might be wrong. So, for anyone listening who has no idea—what is a knowledge graph?

Chantal: A knowledge graph is basically an ontology plus external connections.

So, in my wildflower example, let’s say I wanted to connect my data to weather data.

A knowledge graph would bring in real-time weather feeds, letting me see how rainfall affects bloom cycles.

It can also pull from external sources like Google’s search data, a library database, or anything else that helps enrich the dataset.

Aaron: Got it. So, it’s like an ontology, but smarter—it connects to outside knowledge, instead of just structuring internal data.

Chantal: Exactly!

Aaron: Okay, so let’s break this down using the wildflower example.

A taxonomy is just a structured classification system—for example, the classification of species.

An ontology connects multiple taxonomies—so, linking flowers to their bloom locations.

And a knowledge graph adds external data—like weather reports—to give you deeper insights.

Chantal: Yep, that’s it!

Aaron: Love it. Okay, next question.

You mentioned earlier that you won Taxonomist of the Year. That’s a real award?

Chantal: It is! It’s actually the only award for taxonomists, and it’s based out of London.

Aaron: That is awesome. How do you win?

Chantal: Someone nominates you, and then a panel of judges selects the winner.

Aaron: That’s so cool. We really buried the lede here—we should have started with Award-Winning Taxonomist at the beginning!

Chantal: [laughs]

Aaron: Okay, now for my last question—the one I ask every guest at the end of the show.

What’s one piece of media—something you’ve read, watched, listened to—that really stood out to you recently?

It can be serious, like a great book on ontologies, or not serious at all.

Chantal: Oh, that’s easy. Two things.

First, on the data side, I’d recommend following Sam Russo on LinkedIn. She’s an automotive data expert, and her posts are super engaging.

She makes complex data topics really accessible and uses great visuals.

And second—totally unrelated to data—I’ve been obsessed with Love Is Blind.

The new season just wrapped up, and I’ve been binge-watching it and watching all the recaps.

Aaron: Love it. So, Sam Russo for great data insights and Love Is Blind for peak reality TV.

Chantal: Exactly.

Aaron: Awesome. Chantal, this was such a great conversation.

It was the perfect mix of deep product data knowledge, wildflower science, and Love Is Blind.

Thank you so much for being here. Where can people find you if they want to connect?

Chantal: Best place is LinkedIn—just search Chantal Schweizer, or if you search Chantal Taxonomy, I’ll definitely pop up.

I also have a YouTube channel where I talk all things taxonomy—best practices, bootcamps, deep dives into data structuring.

Aaron: Perfect. Send me the links, and I’ll make sure they’re in the show notes.

Thank you again!

Chantal: Thanks for having me—this was fun!