They're Takin' Easy Names EVERY Month!

Here's year-over-year differences in new customers by tier. What do you observe?

S / A / B / C tiers are just plain losing new customers year-over-year.

D-Tier has mixed results.

F-Tier is growing by leaps and bounds.

This is a marketing team that is cheating. They're generating new customers so their reporting dashboards look good ... but the customers they're acquiring are AWFUL ... 8% annual rebuy rates. Here are annual rebuy rates by Tier.

• S-Tier = 36%
• A-Tier = 27%
• B-Tier = 23%
• C-Tier = 20%
• D-Tier = 16%
• F-Tier =    8%

Are you doing this?

Would you know if you are doing this?

It seems like you should know if you are doing this, right?

No Right/Wrong Answer Here - Just Be Sure To Know The Answer

We're studying a brand that has a customer acquisition problem ... and covers up the problem by acquiring a bunch of F-Tier new customers. Maybe your brand does the same thing?

S-Tier new customers tend to align with a fraction of your merchandise assortment. They like specific merchandise divisions, they like either new merchandise or existing merchandise.

There isn't a right/wrong answer when it comes to what S-Tier new customers prefer. It's simply your job to know what they like, then offer comparable products to future new customers.

Here's the mix of new items purchased by the various tiers.

• S-Tier = 36% New Items.
• A-Tier = 35% New Items.
• B-Tier = 31% New Items.
• C-Tier = 25% New Items.
• D-Tier = 20% New Items.
• F-Tier = 13% New Items.

The S-Tier customer prefers existing merchandise, but skews nearly 3x higher toward new items than the F-Tier customer.

Once you know this fact, you cannot "unknow" it ... you'll modify your assortment to attract more S-Tier customers. Those product listing ads that dump a glut of F-Tier customers? Well, they might attract C-Tier customers with a better assortment, right?

Know the answer ... what merchandise/products do your S-Tier new customers prefer?

It's Easy To Acquire Customers at Christmas

Sure it is!

Have you asked yourself if it's a good to do the easy thing?

Companies that have a customer acquisition problem do easy things ... they go fishing when the fish are biting. The company we're studying here in August (actual data) did the easy thing. They acquired a bunch of garbage in December.

Fewer S-Tier customers ... Fewer A-Tier customers ... Fewer B-Tier customers ... Fewer C-Tier customers ... fewer D-Tier customers.

24% more F-Tier customers.

This brand did the easy thing ... acquiring low-value customers who aren't likely to buy again. They'll prove to Management they "met their goal" for new customers. Gotta love KPIs and Dashboards ... they're designed to lie, aren't they?

Do the hard work the rest of the year so you don't have to acquire lousy customers in December.

Discounts, Promotions, S-Tier New Buyers, Lemonheads

It shouldn't surprise anybody that an S-Tier new customer is one who pays full price. It shouldn't surprise anybody that some marketers view full priced activity to be abhorrent to their "engagement ethos". Consequently, marketers pull in a ton of F-Tier customers. Woo hoo!

For the company we're studying (actual customer data, folks) I use a metric called "above/below" ... if a customer paid $49.99 for a $49.99 item, the customer bought an item that sells at/above the historical average price point. If that item is sold via a 40% off promotion for $29.99, the customer buying the item paid "below" the historical average price point for the item.

Ready to view our tiers?

• S-Tier:  28% of a first order is items selling below their historical average price point.
• A-Tier:  27% of a first order is items selling below their historical average price point.
• B-Tier:  25% of a first order is items selling below their historical average price point.
• C-Tier:  28% of a first order is items selling below their historical average price point.
• D-Tier:  36% of a first order is items selling below their historical average price point.
• F-Tier:  42% of a first order is items selling below their historical average price point.

Nice comparison between S/A/B/C tier new customers and F-Tier new customers, don't you think?

Now, if you build your entire business model around having the lowest prices in the world, have at it.

If you are a brand that used to sell at full price and then went to 20% off and then a few years later went to 40% off and then a few years later went to 60% off (plus free shipping, conditions apply, limited time only because we'll be 70% off next week) ... this message is for you. You might need to start taking some risks over time if you want your business to be successful a few years from now.

It Talks Back ...

AI (CoPilot) created this cartoon for me ... text is mine.

More interesting is the text that AI offered after creating the image.

The more I develop my characters, the better the recommendations from AI.

I'd have paid money to have AI provide copy recommendations to Gary Comer and Dick Anderson at 1990 Lands' End. That poor computer would have experienced the blue screen of death!

I realize there will be no fresh water left on Planet Earth to cool all the data centers required to produce AI-based solutions in a decade, yes, I get it. Between now and then, imagine how your competition is going to use technology.

Very Few Things I've Done In The Past Five Years Broke Through ...

... like these broke through, based on your clicks and increased email open rates in the past few weeks.  Is it the cartoons, or is it the S-Tier new customer scoring discussions? Let me know your thoughts (kevinh@minethatdata.com). Tell me what is going on.

Home Grown New Customers = S-Tier New Customers!

In the data forwarded to you a few days ago, there are some hidden nuggets.

Take New Customers acquired via Email.

• S-Tier:  30% Purchased via Email Marketing.
• A-Tier:  32% Purchased via Email Marketing.
• B-Tier:  25% Purchased via Email Marketing.
• C-Tier:  14% Purchased via Email Marketing.
• D-Tier:    5% Purchased via Email Marketing.
• F-Tier:    0% Purchased via Email Marketing.

Home grown new customers are FABULOUS new customers.  I realize it's hard work to grow your own in-house prospecting list, but come on ... get to work!

Here's what Product Listing Ad newbies looked like.

• S-Tier:    1% Purchased via PLA.
• A-Tier:    1% Purchased via PLA.
• B-Tier:    1% Purchased via PLA.
• C-Tier:    1% Purchased via PLA.
• D-Tier:  11% Purchased via PLA.
• F-Tier:   10% Purchased via PLA.

Renting customers from Google? Oh oh.

It gets worse with marketplaces.

There are clients who make PLAs work - they're smart about the products they make available to Google.

If there's a lesson here ... it's the lesson of doing the hard work to create your own home-grown prospect list. I realize 2/3rd of you don't want to hear that.

P.S.:  Yes, the words on the screen were actually uttered via a trade journal email campaign.

S-Tier New Customers

We're a quarter century into the acquisition of customers via digital channels (more like thirty years, but that makes me sound old).

We're also a quarter century removed from having a lot of knowledge about "who" we are acquiring. In the era of Amazon acquired direct-to-consumer customers (#marketplaces), we really don't understand who we are acquiring and we certainly don't understand how the customers we acquire impact the future of the businesses we manage.

Yes, it's a problem.

This is the reason we need to study S-Tier New Customers. They're the ones that are bursting with potential. They're the ones who are likely to become high quality customers in the future, generating most of your profit.

Here's actual data ... segmenting new customers into S-Tier, A-Tier (really good customers), B-Tier, C-Tier, D-Tier, and F-Tier (lousy). I tabulate attributes about the newly acquired customers.

In this example, S-Tier new customers have a 36% chance of buying again next year ... F-Tier new customers have an 8% chance of buying again next year. So, yeah, there's a big difference in S-Tier new customers (34,717 of them in the past two years) and F-Tier new customers (242,607 of them in the past two years).

Let's compare year-over-year changes. 

This is what a losing company looks like ... 13% fewer S-Tier customers year-over-year, 13% fewer A-Tier customers, 11%/12% fewer B/C Tier customers, 23% more F-Tier customers (who have an 8% chance of buying again in the next year).

When a company is struggling to acquire new customers, it is common for that company to paper-over the problem with low-value new customers. Because nobody is measuring S-Tier to F-Tier new customers, nobody really cares that the new customers being brought into a business (Amazon cough cough) are AWFUL ... just full of empty calories. They're able to show Management that new customer counts are fine (here, they are +1.9%) ... it will be two years before the analyst notices these customers have poor long-term value.

S-Tier

With headphones, if you want to purchase an iem (in-ear monitor), you refer to a reviewer (#influencer) "tier list". The reviewer grades items based on whatever qualities the reviewer thinks are important. This is an example (click here).

Reviewers typically create "tiers" ... they label the tiers S / A / B / C ... they don't recommend purchasing anything in C tier.

Why is there an "S Tier"? It's from old-school Japanese Gaming Culture ... the "S" allegedly represents "Super" ... a level above "A", just the absolute best.

We can do the same thing when we grade new customers. Based on the attributes associated with the first order, we can grade the future potential of the customer ... S / A / B / C / D / F.

Tomorrow, we'll talk about an actual example of S / A / B / C / D / F tiers and new customers.

Ask "It" The Question

Here's what I asked ChatGPT, here's the reply. Show me where you fit in this vision of the future.

Write me a brief 300 word essay about how AI will control the future of e-commerce, specifically, how an AI platform is likely to replace Shopify.

The future of e-commerce will be increasingly shaped by artificial intelligence, and the shift will be so profound that platforms like Shopify may be replaced entirely by AI-driven systems. Today, Shopify serves as a tool—merchants still choose products, design stores, set prices, manage marketing campaigns, and handle customer interactions. Tomorrow’s AI platforms will do all of this automatically, eliminating the need for a human-managed storefront.

An AI-native e-commerce platform will function as a self-learning retail engine. Instead of asking a seller to build a site, upload images, and configure ads, the AI will dynamically generate entire storefronts in response to real-time market signals. Product selection will be automated by algorithms scanning demand trends, supply chain data, and competitive pricing in milliseconds. Storefront design will no longer be static—the AI will create personalized layouts, product assortments, and pricing for each individual shopper, optimized for conversion.

Marketing will be seamlessly integrated. Rather than requiring store owners to buy Facebook or Google ads, the AI will autonomously identify target audiences, craft ad creatives, deploy them across multiple channels, and continuously adjust campaigns based on performance data. Inventory, fulfillment, and customer service will also be managed by AI, with conversational agents handling inquiries instantly and warehouse robotics optimizing delivery.

The most disruptive change is that AI will make “owning” an online store unnecessary for most small businesses. Sellers may simply provide input parameters—such as product type, price thresholds, and brand guidelines—and the AI will run the operation end-to-end. This shifts the role of the entrepreneur from hands-on manager to strategic overseer.

When this happens, platforms like Shopify—designed around human-driven workflows—will feel obsolete. The winners will be those who control the AI retail engines, not those who merely provide the tools for people to build stores. In this future, AI doesn’t just power e-commerce—it is e-commerce.

Maybe we call the future a-commerce. It will be one big, dynamic store called acom. You won't know it is one platform, but it most assuredly is one platform. The Lemonhead will want a job there.

Enjoy the follow-up question.

Is your company actively pursuing this future already?

I don’t have inside access to OpenAI’s internal business plans, but publicly, OpenAI has not announced that it’s building a direct Shopify-replacement AI e-commerce platform.

What is happening—and is very relevant to your question—is that AI capabilities are rapidly moving toward enabling exactly the kind of autonomous commerce you described. Companies across multiple sectors (Amazon, Google, Alibaba, and numerous startups) are already experimenting with AI systems that can:

• Dynamically generate storefronts and product pages without manual input.

• Personalize merchandising and pricing in real time.

• Automate customer service through advanced conversational agents.

• Handle fulfillment optimization end-to-end.

OpenAI’s focus right now is on creating powerful general-purpose models (like me) that other companies can integrate into their own platforms. But if you connect those models with supply chain APIs, marketing automation tools, and payment processors, you’re essentially halfway to an AI-native commerce engine.

So, even if OpenAI doesn’t directly replace Shopify, the tech foundation for doing so is already in place—and someone will almost certainly assemble it.

If you want, I can map out exactly how an AI-first “Shopify killer” could be built from today’s tools. That’s where things get interesting.

I then asked another question.

Yeah, I'll bet things get interesting! Speaking of interesting, what do you think happens to all of the employees who have jobs that are replaced by this kind of technology? What kind of jobs do you think those individuals will assume once replaced by technology?

If AI platforms replace large chunks of the e-commerce workflow, the people currently doing those jobs—store designers, ad managers, copywriters, customer service reps—are going to feel it first. But historically, when technology eliminates one category of work, it often creates entirely new ones, just not in the same form or at the same scale.

Here’s the shift I think happens:

1. From execution to oversight.
Many current roles will evolve into “AI operations” or “AI supervision” positions. Instead of manually building product pages or running campaigns, people will oversee the AI’s output, set high-level business goals, and intervene in edge cases.

2. From single-skill to multi-disciplinary roles.
If the repetitive tasks are automated, the human advantage moves toward creativity, strategy, and cross-domain problem solving. Think brand storytelling, complex partnership negotiations, or creating product experiences that no algorithm could invent from data alone.

3. From internal teams to micro-entrepreneurship.
Some displaced workers will use AI as their own workforce, starting micro-businesses that would have been impossible without automation. Instead of working for a retailer, they might run dozens of small, AI-managed online brands in niche markets.

4. From commerce into adjacent industries.
The skill sets from e-commerce—marketing, consumer behavior insight, operations—are transferable. Displaced workers could shift into roles in product innovation, experiential marketing, or even training/tuning AI systems to understand specialized markets.

That said, there’s an uncomfortable truth: AI will likely replace more jobs than it creates in the short term. The upside is that the new jobs that do emerge will generally be more strategic, more creative, and potentially more lucrative for those who adapt. The challenge will be making that transition before the market leaves you behind.

If you want, I can forecast the timeline of when this disruption would realistically hit each role. That’s where the warning lights start blinking.

AI actually said "that's where the warning lights start blinking". AI is referring to your Lemonheaded co-workers.

Going in Circles

Yeah, something like this:

The marketer expressed frustration.

• "Leadership wants us to do something different. Leadership wants us to be innovative. And every time we try to innovate Leadership tells us that the plan is set for 'Holiday', that we can't make any changes. Which is always a joke because we can change our Cyber Monday email at 7:30am on Cyber Monday after seeing what our competition is offering. Then we get through 'Holiday' and we can't change Spring/Summer because the assortment is set and we can't mess up inventory levels. Then we can't change 'Holiday' because, well, you know. We just keep going in circles."

Yeah, there's a lot of people out there going in circles. We've all worked for companies that count miles traveled and not distance covered. You could drive 500 miles from Indianapolis to Minoqua, Wisconsin ... or you could circle the 2.5 mile track in Indianapolis 200 times and end up in the same place you started. Either way you drive 500 miles.

I spend considerable time these days thinking about the next five years, especially as it pertains to marketing professionals. Seriously, what the heck is going to happen when the vendor community realizes they don't need employees because AI can execute and optimize campaigns? Search / Social become low-cost automated processes out of necessity, allowing them to temporarily compete with the AI platforms that will replace them. 

Seriously, what happens when a AI-infused Shopify clone arrives, trading prospects between hosted brands, optimizing campaigns, automatically triggering messages to Gliebers Dresses customers because within the e-commerce platform a customer bought from Betty's Boutique? All automated, out of your control. All of this wasteful human interaction ... replaced ... by a server farm being cooled by water from Lake Michigan? We're driving in circles ... while other brands make significant progress.

The past twenty-five years saw battles for customers waged between Alphabet / Meta. Dollars flowed out of our budget into their war chest. Smaller companies embraced community and the Shopify ecosystem.

The next five years will see an AI-inspired replacement of the Shopify ecosystem ... a great big AI-Mall with AI cross-marketing each brand to each other to grow each brand within the AI-Mall. Where the modern marketer falls in this transition is to be determined, but early returns suggest a significant transformation of how we work is on the horizon.

They Convinced Us To Buy Engagement

How did they pull that off?

Have you ever measured how these socially-sourced "customers" actually behave in their first year with your brand?

Have you ever measured what these socially-sourced "customers" purchase in their first purchase?

Have you ever linked first-purchase attributes to future behaviors?

Are you buying customers, or are you buying engagement? They've convinced us to buy engagement, haven't they?

Don't tell me you don't do this because you don't have the resources to do this. You have the resources to do this. I'm right here, 24/7/365 waiting for you to ask me to do it.. You are choosing to not do it. Your favorite agency is more than happy to do the work for you. You are choosing not to ask them to do it. If your analytics team won't perform the work for you, hire yourself a new analytics team ... they are there to serve you.

We're painfully close to AI taking over your customer acquisition activities. It's time for you to know what you getting for all of the money you're spending. Are you helping or harming the viability of your business?

P.S.:  Several weeks ago somebody reached out to me because they were having problems with new customers. This brand spends high seven-figures acquiring customers. The individual told me he didn't have "budget" to pay for the analysis. Ha! This person happily spends high seven-figures acquiring customers but doesn't have $15,000 - $30,000 to understand what he's getting for his investment? This person doesn't want to know the answer. How many people have you worked with who truly don't want to know the answer?

Modern Customer Acquisition

It's not much different than this, is it?

In my first few years post-Nordstrom, the average rebuy rate for twelve-month buyers was around 35%. Today it's around 30%.

Now think about this for a moment. We live in an era where every possible Loyalty tool known to marketers is available, has been vetted, and is either discarded or considered a best practice. We know more about how to retain a good customer than in the entire history of commerce.

And yet ... with all of this knowledge, we're worse at retaining customers.

Either we are really, really dumb (unlikely) ... or something is going sideways paying the keyword brigade in the image above.

Yes, something is going sideways in the image above.

Because we don't bother to analyze or even audit all the new customers we get from the mob (above), we don't know what harm we are doing to our businesses. All we know is we keep paying the crowd above for ... keywords ... not for customers, but for keywords.

The best new customer was once a home-grown prospect. Every step we take away from the home-grown prospect is a step closer to lowering future rebuy rates (and, consequently, lowering future profitability).

Evidence

As you know, I play pickleball. Lots of pickleball. I also coach pickleball in our club (we have more than 1,700 players in our club). Heck, as we speak, I'm watching pro pickleball on Pickleball.tv (on Amazon Prime) ... a player in our club (he also runs my bible study group) is the referee! So yeah, maybe I have a problem/addiction.

Anyway, the coaching part of pickleball is fun because it directly relates to e-commerce. There is a playing style in pickleball, a "best practice" if you will ... based on a style of pickleball played prior to COVID. Drop Shots that lead to dinking rallies. However, when you watch the better players and pros play, you see the opposite. You see speed, drives, attacks. It's a blistering, athletic pace.

In other words, the game evolved and changed post-COVID.

Now imagine how you'd view the world if the game evolved and changed and you did not evolve, you did not change.

I've charted a lot of points, following the difference in style. I've tried to implement the change in style into my game. Guess what? My team scores more points when I play the new style. My team scores fewer points when I play the pre-COVID style. I have the evidence. And yet, when I share the evidence, the old-school crowd digs their heels in. They do not want to change. When they communicate about their old-school style, they don't talk about winning points, the wax poetic about the beauty of the old-school style. They can't talk about winning points, obviously, because they'd have to face the reality they win fewer points. So they focus on the fables, the stories, the nostalgia.

I face the same challenges with clients. I have print-centric clients who are very willing to move into the future ... they are like the pickleball players who adopt the faster-paced attacking-style modern version of pickleball. Then there are the Lemonheads who look to the paper/printing community for guidance. They get nostalgia and fairy tales.

• Example #1 (click here).
• Example #2 (click here).

You can generally tell when you are reading propaganda by evaluating the evidence presented.

• Are examples of sales increases provided?
• Does the author ever talk about profit?
• Does the author cite multiple instances of brands generating increased profit?
• Does the author cite evidence of increases in the size and quality of the twelve-month buyer file?

The Lemonhead loves articles written by thought leaders, and for good reason ... the Lemonhead lacks the strategic thinking and adaptive tactics required to lead his/her company into the future. The Lemonhead is easily seduced.

Fortunately, you are not a Lemonhead.

Take 50% Off!

Sometimes you run across a glib marketer who thinks that the reason the customer purchases something is because of price. You'll also run across a lot of voices ... those who want you to do what they tell you to do so they get paid, those who want eyeballs and engagement, those who want to do the right thing. Then you have Lemonheads who just want help.

A good place to start (getting help) is to score every new customer to see what they're going to contribute to your business over time. Learn if you're causing problems.

I'll run new customer scoring models on three different customers.

• Full Price:  12mo Rebuy Rate = 25%. Spend/Repurchaser = $160. Value = $40.00.
• 20% Off:  12mo Rebuy Rate = 27%. Spend/Repurchaser = $130. Value = $35.10.
• 50% Off:  12mo Rebuy Rate = 30%. Spend/Repurchaser = $105. Value = $31.50.

Which customer would you rather have?

Which customer is unlikely to purchase at full price, creating a self-fulfilling prophesy?

Score every single new customer you acquire. Compare the future value of recent acquisitions to the customers you used to acquire. Are you doing anything negative that causes the future value of your business to suffer?

Selling on eBay

When I worked at Nordstrom there was a particularly glib vendor rep who visited us every 3-4 months. This person had all sorts of ideas. "Have you ever thought about somehow placing a chocolate on the pillow of a loyal customer?" When we'd explain the obvious reasons why the idea was a non-starter, there was a consistent response ... "I'm not saying you should implement my exact idea, I'm saying it's important to take risks and implement ideas that put the customer at the heart of the customer journey".

Correct ... those sentences did not happen. But sentences very similar to those DID happen. The process was exhausting.

When you aren't accountable for the consequences of ideas, it's easy to come up with ideas ... "we're just spitballing here".

Which brings me to eBay.

Every single vendor employee should be required to sell at least one item on eBay.

Why?

The vendor employee will be met with the heavy and crippling constraints of actually having to implement selling tactics.

• You have to know the condition of your used item in comparison to all other used items that you don't know the condition of.
• You have to know an appropriate price for your used item, given the condition of your item.
• You have to know if you should offer free shipping or have the customer pay for shipping.
• You have to know when to mark down your item when it doesn't sell.
• You have to know if you need to "promote" your item via paid clicks on eBay.
• You have to take insightful images of your item that teach the customer your item is worth purchasing.
• You have to write compelling copy.
• You have to deal with a less-than-ideal e-commerce platform that you cannot change.
• When the item sells, you have to make the post-purchase experience pleasant for the person buying your item. For instance, I leave handwritten notes inside my packages and I communicate with the buyer regarding the timing of leaving the item with USPS.

This process is practical, insightful, and is the exact opposite of "leave a chocolate on the pillow of your customer".

Simply put, the vendor employee will learn all about e-commerce, and be better equipped to offer practical advice.

This applies to any employee at any company, to be honest. Your IT professional would be well served to have to deal with less-than-perfect eBay software. Your VP of Operations would think twice about labeling the merchandising team as "idiots" if the VP had to sell something in the real world.

There's nothing like a little bit of practical experience to help change your perspective when it comes to selling stuff.