Kevin Hillstrom: MineThatData

It's Time!

Four months ... it went by quickly, didn't it?

If you'd like to participate in the next run of The MineThatData Elite Program, here are the particulars:

Five years of data from 2/1/2021 to 1/31/2026, one row per item purchased, ask me to send you details.
Data due by 2/15/2026.
Payment ($1,800 first-time participant, $1,000 for participants in their 2nd+ run) due by 2/15/206.
Analysis delivered by 2/28/2026.

I'm leaning toward adding more information from comp segment analytics in this run, looking at new/existing items ... items selling at full price vs. on sale ... category data ... price band data ... some high-level views I wouldn't normally provide.

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

Retaining Really Good Customers

Let's look at a problem for catalog advocates.

Matchbacks will give you a misleading outcome. You need mail/holdout tests and you need frequency testing to get to a "better" view of profitability. For instance, you might have a customer who receives 10 catalogs and spends $30.00 across the 10 catalogs. Each catalog would look "profitable" on the surface. When frequency testing is applied to the situation, 10 catalogs isn't best ... it's around 6.

The most profitable outcome is six catalogs. When it comes to retaining most customers, the catalog is not as optimal as it could be ... send six catalogs to this customer instead of ten and take the $3.60 per customer you save and spend it digitally ... something so many catalog professionals are loathe to do.

Here's another thing that so many catalog professionals are loathe to do. This customer received 10 catalogs and it a bit more than twice as productive as the customer above is. What is the optimal number of catalogs for this customer?

It's not ten, is it?

It's 25!

When I share this with catalog professionals, they ... don't ... like ... this ... outcome.

"We're not ever going to have 25 in-home dates, so this outcome doesn't have any meaning for us. Please recommend an actionable strategy."

This is where so many catalog professionals just choose to dump their head in the sand.

They won't mail a customer fewer times even though profit would increase.
They won't mail a loyal customer more times even though profit would increase.

Which, of course, means one thing.

They won't change.

It's 2026. If a catalog professional won't mail nearly every customer less often to generate profit and free up dollars for digital marketing while simultaneously mailing really good customers more often to retain more of 'em ... well ... the professional simply doesn't want to change / evolve / grow. It means that the catalog professional is holding his/her company back.

I realize your favorite catalog agency, your paper rep, your printer, your industry consultant ... few of them want to hear this message. It's 2026. It's time to hear this message.

Loyalty Gives You Options

There are probably at least two ways to define loyalty.

Customer Purchases At High Rates, Repeatedly.
Customer Pays Attention To You.

I can't even believe I'm typing the second bullet point. But 2026 is not 1996. Or 2006. Or 2016. Times change. Attention matters. In my website visitation models, each visit adds to future value, regardless whether the customer purchases anything or not. Which means each time the customer reads your Instagram Post or watches your YouTube Video or actually bothers to look at your Email Campaign, the "attention" is adding long-term value.

I've measured it.

Loyalty gives you options.

Want an example?

Do you know which post I wrote in 2025 that attracted the most eyeballs?

Are you ready?

It's this one.

I mean, are you kidding?

When your customer pays attention to you, you are able to steer your customer in interesting directions.

What are the four topics I write about that generate the most email communications from you to me?

Headphones, In-Ear Monitors in Particular.
Politics ... I don't generally write about politics, but some of you will somehow connect a topic to either "Trump is God" or "Trump is Satan".
The impact of "paper" in marketing. Just a stunning thought given it is 2026 and it has been 30 years since paper mattered.
Pickleball!

Almost none of the topics have anything to do with what pays the bills.

At least directly.

It's hard to ascertain what talking about Headphones and Pickleball does for business. It can't hurt. Someday I'll expand into another hobby (weather), and we'll see what happens.

Loyalty gives you options. You get to step outside of "TODAY 60% OFF PLUS 85% OFF CLEARANCE!" In 2026, we need to step outside of our standard messaging and start relating to our customer base.

Months to Loyalty

When you know how likely a customer is to repurchase in any given month, you can also derive a fun metric that I call "Months to Loyalty".

Months to Loyalty = How Many Months It Takes, On Average, For a First-Time Buyer To Achieve A Fifth Purchase.

Mind you, few customers ever achieve loyalty status, so the metric applies to the vast minority of customers who ever become loyal. The metric informs us of the long process the customer must go through to ever become highly profitable.

I reshaped yesterday's data, so I could calculate Months to Loyalty.

At the bottom of the table you see two metrics.

Months to 5x = Average # of Months Until the Customer Buys for the 5th Time.
Months to 10x = Average # of Months Until the Customer Buys for the 10th Time.

In this case, Months to Loyalty is 28.6 ... it takes the average customer who becomes loyal 28.6 months to get there ... about 2.5 years.

28.6 months. In this case, it means you have a lot of work to do for several years to nudge the customer to loyal (5x) status ... realizing of course that very few customers will make it there.

The more months required to generate loyal status (i.e. a fifth order), the more you focus actually needs to be on new customers. When the metric is lower (i.e. 12 months) you can leverage all of those wonderful loyalty tactics you've been craving to implement.

A Beautiful Retention Table

This table shows the conditional probability of a customer repurchasing, given that the customer in the frequency band has not repurchased in the previous "x" months. Look at this beautiful retention table!

The secrets of your business are outlined in this table.

For instance, you've heard me harp incessantly about the first three months following a first order being critically important in the development of a customer. That fact reveals itself in this table ... look down the "1x Buyer" column ... in this example, the customer has a 14.0% chance of buying again in the first month, then 7.8% in month two (conditional on the fact that the customer did not purchase again in month one), then 4.9% in month three. From there, the customer slowly fades away.

If you knew you had three months to make a difference with new buyers, would you do anything differently? If the answer is "yes", would the company you work for allow you to do anything differently? The answer is frequently "no", and that's why ecommerce (outside of marketplaces) is so darn hard ... customers churn too often, mostly due to a failure to imagine an experience for the new customer within three months of acquisition.