Kevin Hillstrom: MineThatData

"I Just Don't Believe You"

That's how an Executive once responded when discussing the results of a mail/holdout test.

The results were really straightforward. Take an average of the twelve-month buyer file, select 50,000 customers randomly, then resample the group of 50,000 into two groups ... one receives the catalog, one does not receive it. In the six weeks the catalog was active, I measured total demand spent by the mailed group, I measured total demand spent by the control group.

$6.00 spent by the mailed group.
$4.00 spent by the control group.
$6.00 - $4.00 = $2.00 of incremental demand.
Organic Percentage = $4.00 / $6.00 = 67%.
Profit = $2.00 * 0.40 - $0.90 = ($0.10).
To The Twelve-Month Buyer File, On Average, The Mailing Was Unprofitable.

It was the last bullet-point that shook the Executive. The very thought that the heart and soul of his customer base was being managed in an unprofitable manner by ... the Executive himself ... the fact that the Executive was actually harming his business ... was unacceptable.

So he told the room ... "I just don't believe you".

The room moved forward, harming the business.

If you are unwilling to believe the most simple level of scientific inquiry, I'm sorry, you need to reflect inward.

Do you want to hear a story about a room deciding that they finally believed "facts"?

It's 2006 at Nordstrom. When we executed a mail/holdout test, we could measure sales on our proprietary credit card within Nordstrom ... and at other brands. We could see if customers used the Nordstrom card to spend more at Neiman Marcus, or Saks, or Amazon. Was our catalog taking business from Neiman Marcus or Saks? No. It didn't impact those brands, brands we believed we competed with.

Amazon? In 2006?

$10.00 spent at Amazon in the Mailed Group.
$7.00 spent at Amazon in the Control Group.
$3.00 per customer incrementally driven to Amazon.
15,000 credit customers per mail and holdout group.
Results were easily statistically significant.

Want to know something even worse? Within the mail and holdout groups.

$25.00 spent at Nordstrom in the Mailed Group.
$23.00 spent at Nordstrom in the Control Group.
$2.00 per customer incrementally driven to Nordstrom.
15,000 credit customers per mail and holdout group.
Results were easily statistically significant.

Do you see what happened there?

Our brand-centric catalog at Nordstrom, in 2006 (we ended the ecommerce catalog program in mid-2005), DROVE MORE BUSINESS TO AMAZON THAN TO NORDSTROM. Imagine if that was happening to most catalog-centric brands (hint - it probably was)?

In the my early consulting days (2007 - 2012), I'd share this fact with catalog clients. The Executives would just sit there, dumbfounded, thinking about the consequences of driving more business to Amazon than to their own brand. Then they'd move on, pretending they never heard what I shared, because having to face the data was just a bit too much for the recipient of the message to accept.

However, when I shared this fact at Nordstrom in 2006, the data was immediately embraced. Nobody questioned it (though they frequently questioned mail/holdout results). They immediately understood the strategic implications illustrated by the data.

Nordstrom is still here.

Most catalog brands went out of business or were gobbled up by parent companies, thereby losing their identity.

I understand why you don't want to believe mail/holdout results. It's an outcome that requires sober reflection.

I understand why your paper rep doesn't believe the results. They "can't" be allowed to believe them.

I understand why your printer doesn't believe the results. They "can't" be allowed to believe them.

I understand why your boutique agency doesn't believe the results. They "can't" be allowed to believe them.

I understand why so many consultants don't believe the results.

But it doesn't mean the results are wrong.

There's always a group of individuals who chose to not believe data. Modern politics comes to mind. You don't convince people with "more data". You work around them. Because you are reading this, you're willing to accept what your customers are telling you - that's a good thing!

At Bats

I watched this podcast from Orita about email marketing (and other stuff - click here).

One of the speakers talked about "at bats" ... this isn't entirely what he meant, I'm converting his topic into my world, but he was essentially describing the importance of the customer "doing something".

I've told you about this previously ... I worked with a company where the email marketers were very interested in open rates and total sales. The data suggested something interesting.

Customers who clicked-through two email campaigns per year were the customers who mattered to their email marketing program.
Any customer who clicked-through an email campaign in the past thirty days was likely to shop via any marketing/physical channel.
Clicks were very important (opens, not so much, conversion was good but was random while clicks were reliable). Specific clicks were even more important ... merchandise-centric clicks mattered more than promotional clicks.

A simple segmentation plan was created (something I've created variants of for the past fifteen years - your mileage will vary).

Segment 1 = 2+ Email Click-Throughs Per Year.
Segment 2 = Anybody With An Email Click In The Past 30 Days.
Segment 3 = 1 Email Click In The Past Year, Not In The Past 30 Days.
Segment 4 = All Other Email Subscribers

It shouldn't surprise you that the vast majority of attributed email sales in the future came from customers in Segment 1/2/3.

Also of interest ... getting a customer to click once on positive content (i.e. merchandise) caused the customer to become more likely to click in the future ... compound interest. I bring this up because this aligns with the podcast commentary at the start of this post. It's the concept of getting more "at bats". Email marketing functions on "at bats" as the podcaster said or "compound interest" as I'd say. If you build a program that encourages more clicks, you have more customers in Segment 1/2/3 above. If you have more customers in Segment 1/2/3 above, you will have more sales in the future. If you have more sales in the future, you'll increase customer lifetime value, generating more profit for your business in the future.

And if you have more "bad clicks" (i.e. clicking on sale messages or free shipping), you'll find your email list is biased (in a bad way) when compared to your overall customer file.

A sadness of the modern era is the inability for all of this technology to point out compound interest in a meaningful manner. Many of you overcome this sadness and are able to grow your businesses. Good job!!

Subscriptions / Continuity Programs / Merchandise Categories

My Pickleball Mathlete subscription (click here) is an interesting way for me to practice being a marketer.

You can subscribe for free, or you can pay me $30 a year. I make sure free subscribers can see "some" of each article I send to subscribers. About 1/7th of my subscribers pay me. I'm sure some of you would see that as a failure. Who could blame you?

Those with real marketing chops go steps further than I do. They're managing Continuity Programs. These are the smart people. They know what one of their customers is likely to prefer "next". They enroll their customers in these programs ... if the customer doesn't like the program, the customer opts out and the brand moves on. Each purchase, each category purchased from, moves the customer into a different stream of content and commerce. Sure, the business model was popular in the 80s/90s (get twelve movies on VHS free, then buy one movie per month for a year). But the business model is completely applicable today.

Stitch Fix leveraged elements of continuity marketing ... they knew what you bought and what you'd likely prefer next. They called it a "subscription" because the Silicon Valley folks like modern terms for old ideas. Make no mistake ... at a simple level, they knew what you bought, they knew your purchase categories aligned with what you were likely to prefer next, then they sent that stuff to you ... that was a 2015 version of a continuity program.

The secret to ecommerce success is to identify like-minded cohorts, anticipate what they'd like "next", then give it to the customer before the customer knows the customer wants it next. Take a look at continuity program brands ... those who sell collectibles, or coins, or stamps, and think carefully about what they're doing and how it relates to you ... especially if your customer base shops infrequently.

Happy Birthday!

America turns 250 in a few hours!

Enjoy the long weekend!

The World Cup

It was fun for the first 2+ weeks. People taking over Times Square, Scots drinking Boston dry, somebody from Australia eating at a Waffle House, that kind of thing.

We're in the knockout stage now, and fans are going through the blender.

You are probably being told by somebody with a keyboard that your "brand" must capitalize on World Cup momentum.

Apparently AI thinks it is 37 degrees in Houston at the game The Lemonhead is attending.

All of the reasons the World Cup is special (and it isn't special if you are not a fan of soccer/football) work against your CFO's need to deliver a net sales increase of 5% this month AND when the World Cup is over. What does the CFO ask everybody to do next year to "comp" World Cup success?

You know you have something powerful if you can convince a family to spend $10,000 attending a World Cup match in a foreign country.

You might also have something powerful if a customer is willing to pay $49 for a widget. But it is a different kind of "special", and amazingly, it likely requires harder work to move the widget than to sell tickets to a World Cup game.