Kevin Hillstrom: MineThatData

Growing up, we had a Classic Rock station (it wasn't really "classic" yet, but you get the drift) that adored the month of October. They called October "Rocktober". Combine that with "Two for Tuesday" and you really had something ... the dulcet tones of the band Rainbow singing "Stone Cold" and "All Night Long".

I thought about stopping the post right there, but that wouldn't be fair to you, the loyal reader.

Every one of you has a marketing channel that is dying. Sort of like FM Radio is dying. My favorite station in the Pacific Northwest was KRXY ("Roxy") 94.5 from Shelton. You'd hear Rainbow's "Stone Cold" followed by "Cherish" by Madonna followed by "Loser" by Beck and then they'd wrap it up with something from Arianna Grande. Just random Rock/Pop nonsense.

There was nothing like it.

And then?

Spotify playlists.

On June 30, KRXY didn't sell to some soulless entity and become part of the "iHeart Radio Family". They just shut down. Over. Fin. Bankrupt. Their scraps were ultimately acquired by another local station.

Every one of you has a marketing channel that is dying. For some of you, you don't have the metrics to realize that the marketing channel is dying. You keep paying Facebook and they keep sending you customers eight years older than your average customer and you think things are fine. They're not fine. One of the signs of a dying marketing channel is that the customers delivered by the marketing channel are disproportionately older than the age of customers from other marketing channels. Catalogers know this all too well ... they ran their businesses into the ground trusting co-op marketing channels who delivered 74 year old customers instead of the 47 year old customers they needed.

Identify the average age of all customers acquired in the past year by marketing channel. Yeah yeah, I know, marketing attribution double-counting blah blah blah. Do the work anyway. If the nascent new customer from ChatGPT is 33 years old, the Google Search customer is 43 years old, and the paid social customer from Facebook is 58 years old, you have a pretty good idea which marketing channel is going to become KRXY.

        

When you increase prices (typically by discontinuing a product line and introducing a new/comparable line at a higher price), there are several things that can happen.

1. Nothing. This is the hope. You essentially increase prices and customers don't care. It's the kind of thing Apple gets away with.
2. Fewer Items per Order. This is the most common outcome. Prices increase from $20 to $25, customers in-kind decrease items per order from 2.5 to 2.0. Nothing has been accomplished with the notable exception of fewer items going through your distribution center.
3. Fewer Orders per Buyer: Happens less frequently than (2) above but it still happens. Some of your existing customers just say "no". AOV hawks will quickly point out that there's nothing wrong with AOV, and they're right, AOV often increases in these circumstances. And yet? Business is just "off" a few percentage points ... because customers are not buying as often.
4. Lower Repurchase Rate: Sort of a byproduct of Fewer Orders per Buyer, but more significant because some customers just say "no mas" and go find something cheaper on Amazon. Once you start trading in the dark arts of lower repurchase rates, all sorts of lousy outcomes transpire. Your buyer file decreases, you have to "make budget" so you acquire more new customers at ever-more-expensive rates to make up the difference harming the p&l in the process. This is a common scenario.
5. Fewer New/Reactivated Buyers: The counterpart to (4) above. (4) is a measure of what your existing customers think, (5) is a measure of what "the market" thinks about your pricing. It is common to see a disconnect here ... stable repurchase rates among existing buyers but a 20% drop in new/reactivated buyers. If your annual rebuy rate is 30%, the market dictates what you do. If your annual rebuy rate is 70%, you dictate what you do.

        

Here's lawyer-esque boilerplate commentary from Gap's 10K report for 2025, page 38.

A credit rating of BB+ isn't great.

For you and I, interest rates are going up.

And if you really want to terrify yourself, listen to this podcast (click here) and learn what happens when the AI bubble pops. Assuming the hosts are correct, which is an assumption.

You can slide on down to page 45 in the 10K document. Sales up two percent. Store sales +1%, Ecommerce sales +4%. A +2 comp in an inflationary environment isn't really a +2 comp, as you all know. Gross margin was down from 41.3% to 40.8%. Inventory was +7% (oh oh).

If this phrase for fixing the business doesn't signal a warning sign, I don't know what does. Again, from page 45.

• "... optimizing our platform to drive scale by advancing capabilities that amplify and enable our brands."

        

There's been all sorts of calamities, some caused by external issues, some self-inflicted.

I recall 1995, 1998, years where the companies I was with caused their own problems.

2001 - a hybrid of the end of the dotcom era and self-inflicted damage.

2007 - when you could feel that customers spent all their "free" money from their homes, there wasn't anything left to spend (as we'd soon find out).

2012 - the collapse of the omnichannel movement begins (accelerating into 2016/2017).

2021 (Fall) - the popping of the COVID bump.

The data I'm seeing these days is mixed, with some able to thrive, others battling external issues and internal mistakes.

When business signals are mixed, reassess your marketing spend, reassess your new item development plan. Run your Class Of Report (I'll do it for you if you don't have the resources to do it), and make sure you aren't -20% on demand from new items this year (which will cost you money on existing items next year) or -20% on existing items this year (meaning you discontinued stuff you shouldn't have).

        

Four months go by in the snap of a finger!

It's time for yet another run of the MineThatData Elite Program. Cost is $1,800 for first-time participants, $1,000 for those who previously participated. You receive the standard suite of metrics I run (rolling twelve-month analysis, comp segment, comp new/reactivated buyers, repurchase behavior, new/reactivated buyer trends, next twelve-month forecast). Very little cost, plenty of insights into where your business is headed.

Contact me now (kevinh@minethatdata.com) and let's get started.

• Data Due by 10/15/2026.
• Payment Due by 10/15/2026.
• Analysis Delivered by 10/31/2026.