Did I Read The Profit Row Correctly?

Yes. Yes you did. Read the p&l portion of the mail/holdout test again.

The average catalog professional would look at the $6.40 per book metric and say "good enough"! Of course, matchback analytics lie to you (they're designed to lie to you - they protect your paper rep, your printer, your boutique agency that needs you to mail catalogs so they can make money). The $6.40 per book figure is not actually that ... you'll get $4.50 if you don't mail anything, meaning the catalog actually drove $1.90 of incremental revenue.

At the bottom of the table is a profit and loss statement. Look down the "True Gain" column ... this is a p&l of the incremental volume actually generated by the catalog. Tell me what you observe:

• It looks like the catalog generated $2.09 profit per catalog mailed.
• On an incremental basis, the catalog generated a loss of a penny - the catalog as a whole is a break-even proposition.

Can you imagine a world where all of that catalog work - the spreads, the planning, the inventory management, the creative, the photo shoots ... all of that work ... does not generate profit?

Is it any wonder that when presented with findings like this catalog professionals refuse to believe me? The catalog professional would have to confront an inconvenient truth ... that all of the hard work and knowledge ... and some guy with a testing methodology says it doesn't matter? Would you believe your vendor-centric matchback reporting (which makes you look like an invaluable employee) or would you believe a mail/holdout test (even though the mail/holdout test is communicating the truth)?

Tomorrow we'll talk about what it means when a mail/holdout test suggests the catalog doesn't have value anymore (hint - it likely does have value to some customers, just not most customers).

Here's What Test Results Often Look Like

It's 2025. You don't "need" to mail a catalog anymore. It might be a smart idea, yes. In 1990, you had no choice ... no catalog ... no sales. Today, you have a veritable plethora of marketing options to get a customer to your website. If you cater to a 33 year old customer, your situation is very different than catering to a 73 year old customer.

Anybody willing to execute a mail/holdout test is on the good side of the ledger.

This is essentially an average of what I see when I analyze mail/holdout tests in 2025. Ready?

The "mailed" segment received the catalog. Six weeks later (your mileage will vary) we sum all demand from all channels within the mailed segment. We do the same thing for the holdout segment ... this group of equal quality customers did not receive the catalog.

We're going to stop our discussion today with the blue number in the black box. 70%. If we take average demand in the mailed group and subtract average demand in the holdout group, we see what the catalog actually delivered ($1.90 ... not the $6.40 you'd see on a matchback report). Take the $4.50 that customers actually spent if not receiving a catalog, divide it by the $6.40 that the mailed segment spent, and we get 70%.

This figure (70%) is called the "Organic Percentage". It's the amount of business that is generated independent of a catalog. It's the most important percentage in catalog marketing ... the Organic Percentage dictates your future as a marketer.

In 1995, the Organic Percentage might have hovered around 3%.

In 2005, the Organic Percentage might have been around 20%.

In 2015, the Organic Percentage might have been around 40%.

Today, the Organic Percentage is frequently between 50% and 85% ... but there is a big "IF" ... if your digital marketing efforts are underdeveloped OR your customer is 65+ years old, your Organic Percentage will be much lower.

Let's assume your business is typical of a modern catalog-centric brand. If 50% to 85% of your sales will happen anyway, what does that mean for your business?

• Homework Assignment:  Take your catalog matchback reporting, and pretend that 60% of the results will happen anyway. Subtract 60% of reported results at a segment level, then run a p&l for each segment. Tell me what you observe: (kevinh@minethatdata.com).

Mail / Holdout Tests

There isn't one topic that has earned me more universal hate in my career than the topic of mail / holdout test results.

I've deleted more words in this post than nearly any other post written in the past five years. Version after version written, saved, re-read, then deleted.

Pre-COVID, a person reached out to me to mention that they executed holdout tests for a full year. The holdout tests showed that the catalog had minimal impact on top-line sales. They discontinued catalog mailings. Top-line sales didn't change much (if I remember, there was a 7% sales hit). Of course, with fifteen percent of the expense structure eliminated, profitability was at an all-time high.

I doubt (prove me wrong) that there is a professional who has executed / analyzed more catalog or email mail/holdout tests than me. I've been doing it for nearly four decades.

Regardless, I receive a lot of pushback. Pushback falls into the following categories.

1. I don't believe the results of tests.
2. We can't lose sales because somebody wants to test something.
3. Our culture won't do anything with the results of a test, so why test?
4. I don't want to change.

I used to have contempt for #4 ... that has "changed" over time. Seven years ago a 58 year old told me that he was "just trying to get to the finish line" ... meaning he didn't want to overhaul how an entire company did everything as the final act in his career. I get that. Still means you need to do the work, but I understand why some Baby Boomers vetoed knowledge.

Tomorrow we'll talk about the impact of a mail/holdout test.

MRV Distribution

Here's a recently analyzed dataset, showing MRV values for items with enough sales to produce reliable predictions. Look at the distribution of values.

In this case, "S-Tier" MRV items are those with MRV > 1.20 ... adding a 20% chance of the customer repurchasing in the future. Twenty percent is a big deal!

"F-Tier" MRV items are those with MRV < 0.80 ... harming customer repurchase rates by at least 20% in the future. That's no bueno.

If I were your Chief Marketing Officer, I'd have an Action Stream for all first-time buyers, featuring a half-dozen or dozen items that possess S-Tier MRV "and" are Winning Items. Immediately put the best performing products in front of new customers, and immediately put best performing items that generate customers more likely to repurchase in the future.

It's Time!

Yup, happens every four months. It's time for the latest run of the MineThatData Elite Program! In this run ($1,000 for existing participants, $1,800 for first-time participants / then $1,000 per run thereafter), I will briefly analyze your Cyber Monday customers from last year to see if all of those marketing dollars you spend and margin dollars you give up provide meaningful benefit to your business. Hint - results are always mixed.

• Contact me (kevinh@minethatdata.com) to participate.
• Payment Due Prior to October 15, 2025.
• Data Due by October 15, 2025. Five Years of Purchase History at Line Item Level.
• Analysis Delivered by October 31, 2025.

Fat Bear Week!

In 2014 (the first Fat Bear event), 1,700 people voted for the fattest bear at Katmai National Park.

In 2024, 1,200,000 votes were cast for the fattest bear.

How many votes will be cast in 2025? Who will win?

Who do you have winning it all this year?! Join the live chat on September 30 (Fat Bear Tuesday) and learn who is the "big" winner.

How many of you would have quit in 2014 when 1,700 people voted on something and somebody would have told you "what a waste of company resources". Most of you would quit. You'd instead spend money on product listing ads or you'd pay Facebook for engagement. "We can prove that works!"

Going from 1,700 votes to 1,200,000 votes in eleven years means they grew their event by an annualized factor of 1.81. Not breathtaking growth, just a bunch of hard work that compounds over time.

If you are an old-school / traditional "direct marketer", by now you must realize that Instagram and YouTube and a myriad of social channels are "your prospect list". Do the hard work to populate your prospect list.

What A Two-Handed Bowler Can Teach You About Daring To Look Stupid

Ok, this is from a year ago, but YouTube decided to push it to me this afternoon ... I spent 45 minutes with it, and will do so again. This is another "Moneyball" moment for me. Click here for a video about (checks notes) ... bowling. This video is right up my alley.

More specifically, it's a video about Jason Belmonte, who might be the greatest bowler of all time. He doesn't put his right thumb in the ball and bowl one-handed, as is the "best practice". Nope. He bowls with two hands. And he's the best. All sorts of mocking, misinformation and misunderstanding happen as he rises to the top.

• He's told (as a child) by adults that he'll never be accepted by the establishment ... that if he wants to bowl for his national team he'll have to change. Pay attention to what he tells "the establishment" after he wins their tournament.
• He comes to America to prove himself ... the time-honored story so many have followed.
• He's mocked.
• Fans taunt him while he tries to win televised matches.
• Competitors mock him.
• He's accused of cheating.
• He's bowling for a championship and his opponent is interviewed before the match and his opponent insinuates Jason is "doing it wrong" and the opponent is "doing it the right way".
• He's called a "cancer on the sport" by a leading individual in the sport.

Let's just say kids now bowl with two hands, as do younger professionals entering the sport.

Maybe a third of you work in what used to be called the "catalog industry". Can you identify parallels between the content in the video (two-handed bowling = using other techniques to drive sales/profit) and your business? And when presented with facts (i.e. mail/holdout tests) the former "industry" discredits the results or derides the approach of testing (i.e. ignoring the scoreboard) to protect the old-school approach.

P.S.:  I'm willing to bet nobody has evaluated more catalog mail/holdout tests in history than I have. I'm sure there is somebody, but go try to find the individual. The mail/holdout tests tell the truth. Pure and simple. Ask Orvis. They are the scoreboard that our bowler (above) could point to when he won his matches. And yet, people discredited him - ignoring the truth. The parallels between what used to be the "catalog industry" and this 45 minute story are deliciously intertwined.

Does Your Dataset Have What Is Needed To Actually Identify Answers?

One of those vendor emails came out on Saturday. This is what was said.

"They just ran better offers".

"The data backs this up."

I believe the author.

I believe the author because vendors and digital experts do not live in a world where merchandise is properly analyzed. It's seldom included in the datasets needed to answer questions.

Allow me to give you two examples, real data.

Example #1:  I am working with a company that has +10% comp segment merchandise productivity and is contracting by 10% per year. If you have access to merchandising info, you quickly learn the company is failing from a customer acquisition standpoint but has great merchandise that is being bought at higher levels than last year. If you don't have access to merchandise performance, you'd try to correlate the -10% top-line issue to the data you actually collected, and you'd correlate an issue that is not "the" actual issue ... the actual issue is a structural customer acquisition issue (i.e. death of catalog co-ops).

Example #2:  I performed a Merchandise Dynamics analysis for a business. This business saw a 10% sales decline. Marketing spend was the same as prior year. No change in marketing tactics. Website traffic was the same. Conversion rates were down 10%. Digging into the data, it became clear that this brand killed off a bunch of winning items ... replacing them with? Fewer new items (i.e. nothing). Self-inflicted wounds. Now, if you take away the "carryover rate" of winning items ... if you remove it from the analysis and you don't have merchandising data available, you'd likely conclude that the website is failing the brand and you'd launch a sophomoric redesign that wouldn't solve the core problem.

I know, the digital folks are going to start howling and unsubscribing ... "he doesn't get it, offers matter and we aren't responsible for merchandise and even if we were we don't have access to the data, so don't listen to him". I hear this stuff all the time.

It doesn't mean I'm wrong.

Analyze "what" you sell. Incorporate "what" you sell into your digital marketing reporting efforts. The customer is not buying your offer, the customer is buying the merchandise you sell.

Three Men in Trench Coats

In some of the AI-drawn cartoons I ask AI to create three shadowy men in trench coats.

Interestingly, the AI platform I use told me I couldn't reference the three men in trench coats as having any connection to the "mafia" or the "mob", going so far as to say that the AI platform has been programmed in a way to prevent use of the terms. Ok. We'll go with trench coats.

This has been an interesting week on the blog. Lots of communication from readers, some of it quite inspiring, most of it random criticism.

This brings me back to the shadowy figures above.

The last time I attended the NEMOA conference (which, back in the day, was more influential than the Catalog Conference that was disbanded after the Great Recession) was either 2014 or 2015. I paid my own way, one of the myriad joys of running your own business. As Kramer once said, "it's a write off"!

I sat in on a session on attribution ("the catalog drives the whole process, if you don't have the catalog, you don't have search performance"). Somebody from Google sat next to the speaker, nodding (probably nodding because they didn't have to pay their own way to sit there). Anyway, I finish taking notes, the room empties, and I get up to leave. At the door are three men. Oh, I know who the men are. They work for one of those boutique catalog agencies. One of the men closes the door. I realize there are only four of us in the room.

The three men issue a demand.

• "Come work for us, or we will destroy your business."

Um. No. Great offer. Really appreciate the earnest nature of the request. But, no. No!!

One of the men proceeds to tell me that my message does not follow the mainstream messaging the industry needs to thrive. One of the men informs me that the best way to monetize the industry is to have all parties unifying around a shared message.

Yeah, that's offensive. Am I not allowed free speech? You want me to lie to benefit you? No. No!!

Of course, there are consequences for freely saying whatever you want to say, even when speaking the truth. In this case, three shadowy men vowed to put me out of business. Three months later, a fourth man (working for a co-op-style organization) called me and told me he was going to put me out of business unless I joined his organization. A year later, a woman from a boutique agency yelled at me on the phone for a half-hour because my A/B Mail/Holdout tests that proved that catalogs mailed to retail shoppers were unprofitable was a message that was "dumb" and "illogical". She informed me she "won" the account with the retailer and fired me on the spot.

I'll never forget the glee in her voice. If you read her content today, you'll sense the fear in her message, as she tries to hold her business together in a changing world.

There is always a consequence for saying whatever you want to say. Always. If you suggest that digital marketers aren't always gifted with attention to detail or you praise an industry legend or you develop a math-centric analysis that rubs against the grain of an industry, yes, you get to say what you want but people are going to use their right to free speech to discredit you (or to prove they are right when they are wrong ... OR to prove they are right AND they are right!!).

Some come after you because they want an argument. Do not give those folks any oxygen. Ever.

Some come after you because you might be taking business away from them. I've had challenges with the paper industry for nearly twenty years because they believe that my analytics and mail/holdout tests cause companies to spend less with them. I freely say things, in return there is a consequence. Even when I'm right, which in this case I have a career of test results to prove my point ... there are consequences. It costs me money to be "right".

As this week draws to a close, it's important to remember to not let the shadowy figures in your life get to you. Don't give them oxygen.

The Merchant Prince

Preamble:  I mentioned yesterday (click here) how I'd expect unsubs to come because I dared question the attention to detail maintained by digital marketers. And that is what happened. It's always interesting to learn the hills some readers are willing to die on.

It's been eight years since Retail Dive gleefully cheered on the fact that the smartest merchant of a generation didn't perform to "their" lofty expectations (click here). Management and Customers make those judgments, and you already know.

Have you ever worked with a brilliant merchant?

If you have, it's an experience you'll never forget.

I recently met a brilliant merchant ... just walking with this person for fifteen minute was like standing next to greatness. This person "was" her craft. Sheer excellence.

I was on a video conference earlier this year. No merchant. A tech person, a handful of analytics experts, a couple of digital marketers. Not one of these individuals knew ANYTHING about what the company sold. They could tell you that the conversion rate on the website yesterday was 3.04% and the AOV was $77 and the number one source of traffic was an Instagram post from three weeks ago. They were clueless what their converting customers purchased. Worse, they didn't care. "Not our job" as one of the individuals said. It is absolutely your job to integrate with your merchandising team to do what is best for your business.

That team needed a Merchant Prince to come into the meeting and kick some butt.

Until we have strong digital marketers who are allowed to Lead companies (this is a big qualifier, because sometimes they're suppressed from leadership positions), we are at a point in time where we need a Merchant Prince or two to pull us out of 60% off (plus an additional 80% off clearance items) as gimmicks to boost "engagement".

Why Does MRV Matter?

I've told the story before, but 32 years ago at Lands' End I was invited to a Catalog Review meeting. On the walls of the room were each spread in the catalog, colored GOLD / GREEN / BLUE / RED based on the profitability of the spread (a spread is a two-page combination in the catalog ... like pages 14-15, for instance). 

The story of the catalog became apparent the moment you walked into the room ... you could see that the first twenty pages of the catalog STUNK and consequently the catalog didn't perform. You could see that somebody decided to put Home merchandise on pages 16-23, and woo-boy did that kill the momentum of the catalog (including killing the items on pages 24-25).

In other words, you learned why you business didn't perform to expectations. Yes, what you sold mattered. HOW you chose to present it to the customer also mattered ... a +/- 15% impact ... which is the difference between being profitable and not being profitable.

A great sadness of the rest of my career is the complete lack of attention to detail that followed by the companies I either worked with subsequently or observed. As the late Lori Liddle once said, "there's money just lying on the floor, waiting for somebody to pick it up, why aren't you picking it up?" Good question!

The emergence of e-commerce is largely responsible for the lack of attention to detail.

I know, here come the unsubs.

You know that 1,438 visits came from Pinterest. You know nothing else about those visits (oh, I get it, you know that the conversion rate was 1.33% and the AOV was $171). You generally don't know what those customers bought, and you most certainly don't know if those customers bought items that caused the customers to become more loyal in the future.

This is why MRV (Merchandise Residual Value) is so important. It's a modern day proxy for the value of an item. Would you rather have an item that generates $500,000 and causes customers to be 24% more loyal in the future or an item that generates $600,000 and causes customers to be 40% less loyal in the future? You'd prefer the former, not the latter.

Say you are Macy's and you feature this dress in an email campaign (as they did last week).

Do you think that Macy's has any idea whatsoever if that dress increases customer loyalty among the customers who purchase the dress?

You already know the answer.

Now, it's perfectly fine to advertise/feature items that don't have great MRV, especially if they generate sales today. Every item matters. But it is far smarter to give "more real estate" to items that cause customers to become more loyal, right? It just makes common sense.

Common sense.

A Simple X1 / X2 / Y Model Gives Us MRV

As mentioned yesterday, MRV (Merchandise Residual Value) is a function of two independent variables and one dependent variable (yes, I'm about to tell you how to calculate the metric yourself ... something your favorite agency absolutely will not do).

Analysis Period:  All items sold in the past year.

First Independent Variable:  Time ... the average amount of time that has passed (months) since the average customer bought that item. A value of 1.4 means the average customer purchased that item 1.4 months ago.

Second Independent Variable:  Order ... the average order frequency for the customer buying the item. If the customer places her fifth life-to-date order buying that item, the value is five (5). A secondary benefit to calculating this metric is that not only do you get to derive MRV, you also get to know if your best customers buy an item or if new customers buy a specific item. You'd want to know the items that attract new customers, right?

Dependent Variable:  Rebuy ... do customers repurchase after buying the item being analyzed?

Select all items generating at least "$x" in the past year ... in the example I'm using, it's at least $10,000 sold in the past year.

Regression:  Time and Order variables used to predict Rebuy.

MRV:  Rebuy / Predicted Rebuy. You could also use Rebuy - Predicted Rebuy, the outcome is the same, I now find the ratio to be more important.

So, there. Go have your analyst run the analysis for you, Go have your analytics agency run the analysis for you. Seriously, what stops you from doing this right now?

Or, contact me right now (kevinh@minethatdata.com) and let's get busy (click here).

MRV

I call the metric "Merchandise Residual Value", and when you hire me for your S-Tier Analysis you get to see how MRV varies at an item level.

What is Merchandise Residual Value (MRV)?

• (The Probability of a Customer Who Bought "Item X" Repurchasing Again) / (Expected Probability of a Customer Who Bought "Item X" Repurchasing Again).

It sounds nerdy, sure. But it sure is effective.

In the dataset I'm analyzing (actual data), the brand has a new item. The item has been sold for about 1.364 months, and has been purchased by customers placing (on average) their 2.27th order.

• Based on this data (1.364 months, 2.27th order), we "expect" a rebuy rate of just 8.4% after buying this item.
• In reality, the rebuy rate is actually a whopping 36.4%.
• MRV, or Merchandise Residual Value, is (0.364) / (0.084) = 4.343.

In other words, customers buying this item are 4.343 times more likely to repurchase than customers buying an average item.

Wouldn't you want to know that?

Shouldn't you HAVE to know that?

Every item you sold in the past year has a MRV. You need to give extra marketing love to items with a high MRV.

Your S-Tier Analysis

I've been telling you about this all week ... here is your S-Tier Analysis Project!

What do you get?

I analyze all of your new/reactivated customers in recent years to determine if the "mix" of who you are acquiring is shifting ... shifting in a good way or bad way.

I will score all new customers for the next year, either on a monthly or quarterly basis, so we can understand if you are making progress regarding "who" you acquire.

I will measure the life cycle the customer possesses. Does the customer migrate to different channels as the customer matures? Does the customer purchase different merchandise as the customer becomes loyal? Does the customer ever become loyal? Are there key inflection points early in the customer lifecycle when you should apply Action Streams to convert the customer? Are there seasonal differences that need to be taken into consideration (i.e. Cyber Monday or Spring or Summer)?

I will analyze your merchandise assortment ... identifying items that skew to new/reactivated customers, while measuring the items that lead to high Merchandise Residual Value (MRV) ... items that cause customers to become more loyal. You'll know which portion of your assortment is valuable to acquiring new customers.

I will measure the effectiveness of new customers generated via Marketplaces. This has been a really big weak spot across clients post-COVID.

I will produce a five-year forecast that identifies S/A/B/C/D/F Tier new customers, showing the long-term impact of generating too many D/F Tier new customers.

Given all of the shifts in your business tactics post-COVID (especially shifts into Marketplaces), it's more important than ever to understand if you are making choices that harm your customer base over time. Most important - home grown prospects are the path to success. Growing your Instagram / YouTube / Email subscriber platforms means you don't have to ultimately pay somebody else for a one-and-done customer. One of the biggest problems in my post-COVID analyses are the acquisition of too many one-and-done customers.

Cost:  $24,000 ... half due up-front, half due within fifteen days of project completion.

Special Subscriber Opportunity:  As always with my subscribers, I give you an opportunity to help break in my code. If you respond by 11:59pm EDT on September 16, you can help break in my code for just $17,000. If you are a prior paying client, it's $15,000!

Contact me right now (kevinh@minethatdata.com) to get busy!

P.S.:  Need adaptations to this project for your unique business issue? We'll modify the scope, no worries.

It's Different Today

Not quite sure what AI had in mind as it blended a bunch of stuff in this image.

Maybe a decade ago I visited a retail brand - the CEO was ANGRY - we've all been in the meetings. HIs anger was obviously misplaced. His stores performed admirably! His share of e-commerce volume was low. He didn't like that. At all.

Ten years ago the pundits demanded full integration of all operations. If you sold something in a store, you HAD to sell it online at the same price and the same promotions. What a paralyzing thesis.

We learned that the customer could have cared less.

In 2025, it's different.

We've learned.

You do what is best for the customer, regardless of "channel". 

Sometimes there are popular items that you want to make available all the time, in every channel. One problem. The customer who buys those items doesn't come back and buy again. It's a bad way to acquire a customer.

This means you have to have a measurement method to know what items new customers purchase ... items that lead to lousy downstream performance.

This analysis will be part of what I announce to you, shortly. If you want additional details, let me know (kevinh@minethatdata.com).

Football Manager

This is a 46 minute video (click here) about the game Football Manager ... a full scale simulation of running a Football/Soccer team.

Have you ever been absolutely stunned by something? I'll assume you answered yes. I am stunned.

What we really, really, really need is a version of this called "Chief Marketing Officer". I'm half-heartedly thinking of doing it.

I'd adore a simulation where you are sitting in your physical office and you are told that the paper rep somehow snuck into the building and is sitting in the lobby, asking for just fifteen minutes of your time to discuss 100# cover stock options for 2026. Five minutes later he somehow escapes terminal velocity and interrupts your weekly staff meeting with a binder that says #printisback on it.

Wouldn't you want to get through a simulation where the Chief Financial Officer mistakenly includes you on a group thread where he says that you "have no clue" what is going on?

Shouldn't everybody have to deal with a merchant who thinks you are, and I quote, "sending emails to the wrong people" and that is why her handbag isn't selling?

You wake up at 4:30am and the CEO texts you that she wants to see you in her office at 7:00am ... what kind of excuse do you come up with to avoid that situation, and then how do you deal with the consequences of your actions?

I mean, if you want to truly vet whether a person is qualified to be your next CEO, wouldn't putting the candidate through a simulation ... watching to see how s/he handles the crazy situations that come up every day ... wouldn't that alone make production of the simulation worth it?

Thoughts? (kevinh@minethatdata.com)

Reader Mail

I recently received feedback from a professional ... "We're using AI to reactivate customers, it works well, we're seeing a 10% lift in performance!"

Yeah, it sounds reasonable to celebrate how these customers are performing in the short-term. I get it. 

I wonder if the dude even knows "who" exactly he is reactivating? Is he reactivating somebody who is going to pay him back over time, or is it yet another case of giving away 50% off to seduce a transactional customer via AI?

I'd think you want to know "who" is buying from you, right? The more digital our businesses get, the less y'all seem to know about "who" is choosing to buy from you.

Contact me (kevinh@minethatdata.com) if you want to learn more about the analysis project I'll be releasing later this week.

A Change In Outcome

Sometimes you read stuff that is just so awful you wonder how people keep their jobs?

Coming out of the Great Recession, it was common to see shortfalls in the number of new customers. I'd forecast the next five years, and it was obvious the brand being studied could not "keep up" ... sales were going to contract because the company could not find enough new customers.

Then COVID happened ... with customers trapped in their homes for a few months, new customers were plentiful ... and we got lazy. At the same time we got lazy, startups built new ways to find customers. We see it every day in 2025, but the roots of what we see in 2025 were planted in 2020. So many of us became movie-makers, building community in the process.

Those who missed out on this trend moved in the opposite direction. As new customers dried up (reverting to historical trends, which were not good trends), e-commerce folks did two things to attempt to make up the difference.

1. Ridiculous discounts/promotions.
2. Marketplaces.

Both accelerate a new problem ... the quality of the new customer is so bad that even if there is a modest increase in new customers there is frequently a modest decrease in future sales. This is something I didn't see back in 2010 ... but it's here now.

The best new customer is a previous prospect that you've nurtured. The prospect followed you on Instagram, subscribed to your YouTube content, opted-in to email campaigns. The prospect is on three lists ... and you old-school direct marketers are always looking for lists ... these are the "lists" you should care about, home-grown low-cost lists. When these prospects become new customers, you'll find they are frequently "S-Tier" new customers ... they care about you, they aren't there for a transaction, they believe in the relationship that has been developed.

Later this week I'll reveal a new direction in project work. As always, you're going to get the first opportunity to benefit. Contact me now (kevinh@minethatdata.com) if you want more details prior to the announcement.

Did You Know?

Did you know that you sell items ... popular stuff ... to first time buyers? And some of those items cause your new customers to repurchase at lower-than-average rates?

Before you click "delete" and continue paying Google a bunch of money for (checks notes) keywords, think carefully for a moment.

• Somewhere in your marketing department, you have somebody who is accountable for marketing to loyal customers to increase loyalty.
• That person is being hamstrung ... by you ... because you are promoting specific items that are attractive to prospects who will become new customers who won't buy again.

I wouldn't tell you about this fact unless it was a rampant issue in 2025.

I'll take you back to 1990. Back when the President lavished hate on broccoli, I worked at Lands' End. We had a 64 page "prospect catalog" with only the best selling items. By doing this, we could avoid all those vapid spreads that didn't generate enough profit.

We didn't look at whether the items we were selling in prospecting catalogs led to future customers bursting with potential. We finally did that in 1995. That's when we learned things ... things like getting a first time buyer to purchase bedding was a bad thing because the customer didn't want anything else in the future ... but getting a Mom to buy Kids product caused the Mom to buy Womens and Mens and Kids and Home and Tailored product in the future ... a good thing!

Do you think you have comparable items?

Don't you think you should know what items meet the criteria ... good or bad?

Next week I'll release a New Customer product. We'll look for S-Tier new customers, we'll look for S-Tier products that lead to S-Tier new customers. Actionable. And I'm not aware of anybody else doing this stuff.

Sometimes It's Really Simple: Multiple Items In A First Order

For the brand we're studying, customers who are S-Tier customers tend to be customers who buy multiple items in a first order (#duh).

I mean, there's been cross-shopping and upselling software for a quarter century or more, but it's just as important as ever to actually use the tools available to us. The data (seen above) is so overwhelmingly supportive of doing some hard work to encourage multiple-item purchases that you should ... encourage multiple-item purchases.

S-Tier Customers by Channel

Yup, marketers evaluate success/failure by channel. It's just how it goes. I can count on one hand the number of companies I'm familiar with that measure the "quality" of the customers they acquire.

But my goodness ... YOU NEED to measure the number of new customers by "Tier". 

Why?

Look at this table, measuring new customer counts by Channel and by Tier.

Where do the S-Tier new customers come from?

• Email.
• Call Center (hint, print)

Where do all of the lousy new customers come from?

• Product Listing Ads
• Amazon
• Marketplaces

Remember, in my example, here are twelve-month rebuy rates by Tier.

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

In other words, the channels many of you have switched to in recent years deliver AWFUL new customers (in this example). Meanwhile, the "hard work" channels (i.e. Email) deliver mostly S/A/B Tier new customers.

Too many of us harmed the companies we work for ... listening to Lemonheads parroting Thought Leader nonsense. If we want our businesses to be successful, we need to do some hard work. Yes, I realize you don't want to hear that ... you're already stretched for resources. You'll need to find resources.

Also - notice the dearth of social media here (and video, geez). Social Media and YouTube are your modern prospect lists ... to use terminology from the 1990s. This is where you reach out to prospects ... at minimal cost ... teaching them why they'll eventually purchase from you! In my projects, I can tell when a brand prioritizes easy (paying for clicks) vs. prioritizing hard work (email, social, community, video). Hint - most of you prioritize easy. When you prioritize easy, you get a lot of F-Tier new customers who never become loyal.