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.