Kevin Hillstrom: MineThatData

The Ecommerce Middle Class

Not "middle class" in terms of income. "Middle Class" in terms of value to a client of mine.

If we go back twenty years, your customer file was comprised of a lot of "middle class" customers. These were customers who generally purchased two times. This was a good thing for ecommerce brands, because these customers had a fighting chance of becoming loyal.

Whether it is Amazon taking all the middle class customers away or the Shopify empire encouraging purchases from a million smaller brands, the middle class has generally disappeared in ecommerce.

Twenty years ago, a first-time buyer had a 35% +/- chance of purchasing again in the next year. This accelerated the rate of customers moving to a second purchase, where the customer had a 45% chance of placing a third order. There were many 2x/3x buyers as a consequence, customers in the middle of the customer value bell curve.

When first-time buyers have a <= 25% chance of buying again in the next year, the middle class gets hollowed out. Not nearly enough customers develop to 2x/3x status, which means eventually not enough customers become loyal.

You've heard me talk frequently about "Welcome Programs". These are separate email marketing streams that target customers who just purchased for the first time. The goal of the programs are simple:

If the customer needs to return something, make it easy for the customer to return something - and then find something else for the customer to purchase.
Teach the customer about who you are. Get the customer hooked up with Instagram or any of the social media platforms.
Based on the merchandise purchased in order #1, cross-shop the customer into the next most likely category that the customer wants to purchase from.
If you have a loyalty program, get the customer enrolled in the program.
Rapidly convert the customer to a second order - get this customer into your "middle class".

How many businesses do an excellent job at quickly converting a first-time buyer to a second order?

Very few.

I'm convinced few are good at this because few possess the creativity to be good at this. It's one thing to create a beautiful image that you put up on Instagram. It's quite another thing to put together a dozen different triggered email programs based on "what" a customer purchased in a first order. It's not easy for a campaign-centric creative professional to adapt to the reality presented to us in late 2026.

New Customers: Does Quality Matter?

Maybe your data looks like this:

All Other New Customers: $20.00 profit in the next year.
Google Shopping New Customer: $12 profit in the next year.
Paid Social (Facebook): $10 profit in the next year.

When you email me your questions, you'll ask if you should avoid the Google Shopping / Paid Social new customer?

What you should avoid is wasting a lot of money marketing to those new customers. If you can earn a profit acquiring those customers, market to them in the "welcome window" ... the first 90 days after acquiring the customer, then be gentile with your marketing dollars. Don't waste them chasing a customer that is not interested.

I know, this is the point where you suggest it isn't worth acquiring a customer that doesn't want to have a relationship with you. Almost no customers want to have a relationship with any of us. The customer has a need at a point in time. The customer wants that need met.

The quality of a new customer matters in relation to how you manage the downstream relationship. If you can earn $25 profit acquiring the customer and then earn another $10 in the future (whereas a different new customer pays you back $25 / $20), just back off your downstream marketing efforts to that customer.

The average ecommerce brand doesn't have this problem, FYI ... they don't have outbound marketing programs that cost $15/year. This is a decidedly old-school problem.

Old School Geography

I think this goes back to 1991 or 1992. I was asked to explain what was happening to swimwear sales at Lands' End. I mapped sales by state by month.

In January/February, sales boomed in the Midwest and Northeast.
In March, sales thrived in the Southeast United States.
In April, the eastern seaboard performed well.
By May/June, sales moved back to the Midwest and Northeast.

In other words, sales followed the weather. Vacationers bought merch in the heart of winter. From there, the weather improved most in the south, then southeast, then eastern seaboard, then in the midwest/northeast.

The relationship was consistent. In fact, when I left for Eddie Bauer, the relationship at Eddie Bauer was nearly the same.

But when swimwear business was not great, it was largely a weather issue.

Back in 2005-2006 when we killed off the catalog business model at Nordstrom, it was my job to field angry calls from customers. For about four months, angry 45-75 year old women in Minot and Rapid City and Cheyenne and Bozeman let me hear their frustrations with no longer receiving a catalog. I recall telling my boss where these customers lived and she said "those aren't our core retail customers". The stirrup pants customer in Billings was not the core retail customer at The Galleria outside of Dallas. In fact, when we coded zip codes, those who bought from the catalog skewed disproportionately rural. One of my more successful consulting projects was from 2008 ... "Zip Code Forensics", where I created an algorithm to code zip codes, then showed how certain zip codes aligned with catalog marketing. That project lingered for eight years until ecommerce fully consumed the catalog business model.

A while back a CEO talked about the core customer being "native" to New England. I didn't think that was possible until I analyzed where his customers lived ... sure enough, they were skewed to zip codes starting with a 0 or 1.

There are things to be learned by analyzing business performance by geography.

Gap Year

There's an election tomorrow. Take advantage of your right to vote.

In sports, it is common for coaches to be fired. The LinkedIn crowd might appreciate that "accountability" comes for the coaches being fired.

Of course, if you were the coach, you know full-well how difficult your job was. It is highly unlikely that you didn't try your best ... you likely gave everything you had to be successful. It just didn't work out. Happens in ecommerce all the time, likely has happened to you at some point in your career.

When a head coach is fired, it is common for the coach to take a "gap year" ... a year off. The coach doesn't rush into a bad job (which likely harms the coach long-term), the coach studies how everybody else does their jobs. The coach looks for innovation, for blind spots in how s/he performs compared to others.

A year later, the coach reappears, wanting a new head coaching job or potentially taking a step back to become an offensive or defensive coordinator.

I was never smart enough to take a gap year. I moved on from Lands' End to Eddie Bauer in the course of a weekend. I moved from Eddie Bauer to Avenue A in the course of a weekend. I took one (1) week off moving from Avenue A to Nordstrom. I took a weekend off moving from Nordstrom to my consulting life. All of it was stupid. I should have studied how to be better as a professional, how to be better as a leader, how to be better as a co-worker, how to better help my new company be successful.

If (and only if) you can possibly afford to financially take "some" time off, do so when you are forced to change your path or you voluntarily change your path. Learn from somebody else. Shadow an acquaintance for a few weeks. Learn how other people with comparable skills approach your job.

If you're in this situation, contact me (kevinh@minethatdata.com). I'll share how I view things, you can take my advice and throw it away or you can apply it to your work - that's what you get to do if you take a gap-week, gap-month, or gap-year.

Share of AOV

Every business possesses a unique relationship between first-time buyer AOV and future value. The relationship dictates exactly how the brand must approach marketing for first-time buyers.

An example: "Brand A" has an AOV of $70 and in the first five years after a first purchase the customer spends $230. You wouldn't be surprised if this company offers free shipping, executes television / radio / billboards, hires expensive influencers, seemingly wastes money. This business "can" waste money because the customer pays the brand back over time. In fact, it is almost good that the brand wastes money, because the long-term payback of finding new customers is so worth is that the math all works out.

Another example: "Brand B" has an AOV of $70 and in the first five years after a first purchase the customer spends $22. This company has no choice but to be exceptionally disciplined in their marketing approach. No waste. None. They can only afford tactics that are guaranteed to work and deliver profit during the acquisition transaction. Their "Share of AOV" is poor ... just 31%.

There are times when you read stuff from the LinkedIn experts, telling "brands" what they "must do" to be successful. Well, "Share of AOV" dictates what a "brand" must do ... if the company wants to be profitable.

There is a reason we always end up with department stores, or Walmart, or Target, or Amazon, or Marketplaces. The extended merchandise assortment (which comes at a cost, by the way, but frequently at a cost that is lower than wasteful marketing spend) drives "Share of AOV" much higher, allowing the company to generate a ton of downstream profit.

There is reason most newer enterprises hit a ceiling. Their merchandise assortment is too narrow, meaning "Share of AOV" is too low and eventually the brand reaches all of the prospects it can possibly reach.