Kevin Hillstrom: MineThatData

Case Study: Dispelling Legends

At "Beans: The Internet's Only Variety Store" there is a legend ... that a broad assortment "holds the brand together".

It's been my opinion that Leadership is violating this "legend", assuming the legend is true.

So ... I performed a classic Factor Analysis to demonstrate categories that customers like to purchase from. If dots on the image below are close together, customers like to purchase from the categories "near each other" on the image. If categories are far apart, it means different customers prefer the categories.

Here's the image from the Factor Analysis.

There appear to be three reasons why customers buy from this brand.

Apparel Tops: We know this is a high-volume category, and it is all by itself meaning that many customers ONLY purchase Apparel Tops.
Home / Outside: There are clearly customers who view this brand as a Home / Outside brand. There's some bad news here ... some customers clearly come for Home / Outside, but a few weeks ago I showed you that Home decreased from $6.0 million to $2.9 million over four years. No bueno.
Everything Else: There are customers who view this brand as an eclectic mix of categories (the right/middle side of the image).

Of interest ... Apparel Bottoms does not align with Apparel Tops ... it aligns with "everything else".

I have to analyze new customers by these three groupings ... if the future value of a customer who aligns with "Everything Else" is higher than other categories, we have a lot of freedom as marketers to be clever. If "Apparel Tops" drives new customers, we have to concede that those customers might not appreciate the "entire assortment". Messaging to Leadership doesn't always go well on that topic.

P.S.: If you like what you are seeing here and are interested in a Category/Customer centric analysis, send me an email (kevinh@minethatdata.com).

Case Study: Customer Response to a Dying Category

We talked about Apparel Tops yesterday. We've previously mentioned that Fashion is a dying category, largely because the merchandising team appears to be killing the category. How does customer response change when a category is being killed off?

Similar to Apparel Tops, most demand comes from new/reactivated category buyers (80%). Again, the marketer has to know this, because the marketing plan has to include a lot of $$$ and attention in awareness (organic social) and search (product listing ads). If the marketer doesn't acknowledge this fact and act upon it, well, the marketer is equally culpable with the merchant at killing off the category.

This likely applies to your business as well. Most of your categories offer products that largely appeal to new/reactivated buyers and/or prospects. A marketing department that does not understand this dynamic is a marketing department that sub-optimizes the potential of the category/business.

Ok, what have the merchandising team done with their assortment-contraction initiative?

Rebuy Rates over time.

12-Month Fashion Buyers = 2.0% to 3.0% to 2.8% to 1.5%.
All Other 12-Month Buyers = 1.6% to 2.5% to 2.2% to 1.2%.

What happened in the past year is telling ... 40% or greater decreases in rebuy rates (albeit very low rebuy rates). With less merchandise available, existing buyers become less likely to repurchase.

The astute reader should say "Hey, Goober, you just told us that almost all demand comes from new/reactivated buyers, please tell me how many new/reactivated buyers the category had over time".

I can do that.

36,236 to 54,438 to 52,418 to 27,308.

We see the same (ugly) trend with new/reactivated buyers ... counts are down nearly 50%.

This comes up repeatedly in my work ... if you trim the assortment, you harm demand/sales. If you grow the assortment, you increase demand/sales but introduce other challenges (inventory / liquidations / margin erosion).

I'm going to hold off on communicating this fact to Paisley Ingram (the owner) until have a few more data points. I need to find a simple way to tell a complicated story.

Case Study: Customer Response To Merchandising Changes

Let's approach this discussion in bite-sized pieces.

This table reviews repurchase activity for Apparel Tops ... the best-selling category that Beans: The Internet's Only Variety Store sells.

Yes, there's a lot going on here.

An introductory tidbit ... in the past year, 79% of demand in Apparel Tops comes from customers who haven't bought in at least a year or are first-time buyers. Only 16% of demand came from last year's Apparel Tops customers and just 5% of demand came from other twelve-month buyers.

If you are the marketer trying to grow Apparel Tops (your best-selling category), what might your approach be?

Awareness (organic social) and Search (product listing ads).

You could try to squeeze more out of existing buyers, but what is the point? If your efforts were positive and increased sales among existing Apparel Tops buyers by 10%, total demand would grow by 16%*10% = 1.6% ... meaningless.

When a merchant tells you that the marketer is not getting her product in front of the "right customers", the merchants is both wrong and right at the same time.

Wrong, in that the merchant usually has minimal experience with marketing techniques.
Right, in that most categories benefit from exposure to prospects who haven't bought the product previously.

Pay attention to rebuy rates over time.

16% to 15% to 14% to 11% for existing buyers.
7% to 6% to 6% to 4% among all other twelve-month buyers.

This is a mystery that requires professional levels of communication.

The category is stable because new/reactivated customers are buying the product.
The category is stable because new/existing items are being managed reasonably.
The housefile ... twelve-month buyers ... are increasingly less likely to buy this product, but their share of total demand is not sufficient for management to notice there is a problem.

Case Study: Hints of Discontinued Items

Remember our "Class Of" table for the Fashion category? I do.

There are "tells" in this table that help me understand how a merchant approaches the business. In this case, the merchant in charge of Fashion discontinued existing items. Yes, the merchant failed to introduce enough new items to grow the category. Acknowledged. But the merchant also decided to take the hatchet to existing items.

How do I know this?

Look at the Class from Three Years Ago. Demand went from $564k to $384k to $160k in Years 1/2/3. Demand trails off after the introduction year, then trails off faster.

Look at the Class from Two Years Ago. Here comes a "tell". Demand went from $860k to $276k to $75k. That doesn't happen because the items fall off faster ... that happens because a merchant says "I don't like those items". Not liking existing items and then not introducing enough new items is one of two things.

The merchant is killing the category on purpose (which happens all the time).
The merchant is committing professional malpractice.

Our job is to understand if (1) is happening or if (2) is happening.

If a merchant is trying to purposely kill a category, the merchant has to demonstrate that killing the category does not impact other categories. In other words, when I worked at Nordstrom, if we killed off Cosmetics we'd be killing off the business because fragrance on the ground floor of a store brought in new customers who shopped the entire store. There's nothing more alluring to customer acquisition than fragrance.

Tell me what you've learned so far (kevinh@minethatdata.com). Are you finding this valuable? My clients find it valuable.

Case Study: The Virus Is Spreading

Yesterday we talked about the fact that Fashion was being starved by a lack of new items. Because existing items either die off quickly or are discontinued quickly, new items must fill the gap if a category is going to thrive.

I summarized all categories - looking at new items by year, as well as demand across new items and existing items. Here we go.

What does the data tell us? A lot. Your category data will communicate to you as well.

Total Business = Starved of New Items.

Apparel Bottoms = Not Starved.

Apparel Tops = Not Starved.

Fashion = Starved.

Home = Starved 2 Years Ago, Lack of New Item Demand in the Past Year.

Jewelry = Starved.

Entertainment = Starved.

Workplace = Drastically Starved.

Outside = Not Starved.

Having Fun = Starved and a Lack of New Item Demand (might be a metaphor in this category).

Seasonal = Starved.

Decorations = Not Starved (many new items 1-2 years ago).

We have Apparel Bottoms, Apparel Tops, Outside, and Decorations that are being managed consistently. Everything else? Being cut to the bone.

Apparel Bottoms / Apparel Tops / Outside / Decorations?

1,099 new items three years ago, 1,055 new items today.
$7.1 million existing demand three years ago, $7.1 million today.
$6.7 million new item demand three years ago, $5.5 million today.

Everything else?

2,047 new items three years ago, 1,125 new items today.
$7.4 million existing demand three years ago, $5.5 million today.
$7.2 million new item demand three years ago, $2.6 million today.

In the four "preferred" categories, there is still degradation in new item demand on a comparable number of new items.

All other categories are simply imploding.

This business isn't failing. Management is likely failing the business. Four categories are performing acceptably. I emailed Paisley Ingram a paragraph about what I observed.

From: paisley.ingram@beans.com
Sent: Monday, May 4, 2026 11:13 PM
To: Kevin Hillstrom <kevinh@minethatdata.com>
Subject: RE: Four Key Categories

Kevin, we certainly struggled with our merchandising strategy the past four years. We know that Apparel is performing well, so we elected to prioritize Apparel. Sloane's arrival (our Chief Merchandising Officer) has been invigorating! She has us focused on products and categories that performed well historically. She uses actual performance data to show us "what works", then plans her assortment accordingly. She believes that new items are inherently risky, and any downside to not having enough new items can be overcome by a sound marketing strategy to attract customers or speak to loyal buyers. She strongly believes in a marketing/merchandising partnership that is led by product carryover tactics.

You keep asking interesting questions that stray from marketing, focusing on what we sell and how we sell it. Are you suggesting a gap between our current merchandising strategy and our results?

Best,

Paisley

______________________________________________________

From: Kevin Hillstrom <kevinh@minethatdata.com>
Sent: Monday, May 4, 2026 4:57 PM
To: paisley.ingram@beans.com
Subject: Four Key Categories

The data suggest that about half of your business is somewhat healthy ... Apparel Bottoms, Apparel Tops, Outside, and Decorations. All other categories show a disinvestment in new items, new items then fail to become existing items, causing those categories to perform poorly over time, with an acceleration in poor performance in the past year. Was there a plan to contract the business, or was there a plan to approach most categories differently and the results observed were unexpected?

Thanks,

Kevin