Kevin Hillstrom: MineThatData

Retention: What Should I Do?

When I tell you to do something different to retain specific customers, I receive similar and common feedback:

"What do you want me to do differently? We email customers every day. We post on Facebook. We pay Google to snare our customers. We're on Amazon. We're everywhere!"

Oh come on. You do all the easy stuff. It is not hard to stick $0.60 into Google's trillion dollar bank account. It is easy to send eleven email campaigns per week featuring 40% off and 70% off clearance.

Your ESP provides software that allows you to separate any customer with recency = 8 months. Set those customers aside, and create specialized campaigns that override normal campaigns. Teach every customer getting ready to be "active" again what you're selling in three months. Create events for those customers ... three months out of course, but create the events regardless. Do some work! Educate the customers, prime the pump, create some excitement. Be exclusive, unique, differentiate yourself from the lemmings preparing to jump off of the Cyber Monday cliff.

Again ... your ESP provides you with software to perform customized Action Streams for any customer at any status in the Customer Life Stage. Use the tools you are already paying for.

Retention: When The Fish Are Biting

Here is one of my retention tables for a Home brand. The table illustrates incremental monthly rebuy rates. Tell me what you observe.

Did you miss it? Look at the row with twelve-months since a prior purchase. Heck, look at 10/11/12 months since a prior purchase. Those customers are MORE RESPONSIVE than are customers who bought 4/5/6/7/8/9 months ago.

This happens all the time ... run the table above for your brand and tell me what you see.

If you care about retention, and from your emails and messages on the socials it is clear you value retention a lot more than you value customer acquisition, why aren't you capitalizing on this window of opportunity? Why do you treat these people the same as you treat everybody else? Please, tell me the reason why you don't do anything different with seasonal buyers?

Access To Your Own Customers = 7% Fee

So many of you use Shopify. I've spoken at length about the day when Shopify is an AI-infused marketplace. Those days inch closer (click here).

In my work, there is something called the "Profit Factor" ... it is the percentage of sales that flow-through to profit/contribution prior to fixed costs. I have clients where the percentage might be in the upper teens ... I have other clients where the percentage is 45%.

The goal of any service provider / vendor is to take as much of your post-profit-factor contribution as possible without compromising your business ... if they take too much, you lose too much money and stop working with them (which harms them) ... if they take too little, they aren't profitable enough and they cannot "#scale" their businesses to the level they desire. There is this teeter-totter effect that is constantly being balanced.

Let's pretend you are one of my clients with a 35% profit factor. You generate a $100 order, and with a profit factor of 35% this means you generate $35 of contribution. Shopify/ChatGPT believe they deserve $100*0.07 (it's 6.9% plus a few pennies of a fixed fee) = $7 of the $35 of profit you just earned.

You have 200 employees for your $60,000,000 business ... you work so darn hard ... and then a third party is entitled to 7/35 = 20% of all profit/contribution dollars you generate.

Of course, this means nothing today ... ChatGPT generated orders are a rounding error. In five years? In five years, all of your search-centric orders will be AI-generated orders, all with a toll applied ... the days of organic search will be no more.

From a vendor perspective, tolls are best collected under three circumstances.

Tolls are variable, they apply to every order. Fixed charges are limiting.
Tolls are necessary, meaning the vendor makes the case that without the toll the order doesn't happen, even though the order likely happens without the toll.
Tolls are easy, they require very little additional work per order.

From a "brand" perspective, tolls are necessary when the brand is unable to generate orders organically via storytelling and/or merchandise the customer must have. The best "brands" don't have to discount, and they don't need a third party interjecting themselves into the customer relationship.

So that's the balance we're all fighting. Some of us do it better than others. Don't get stuck on the high end of a teeter-totter.

Here's What Saks Needs To Do To Fix Their Business

That's the way these LinkedIn posts start ... a "transformational retail expert" will talk about "products customers actually want to buy" ... Good Lord, do we think the folks at Saks read that and say "Yeah, let's try that, let's try to sell something customers actually want to buy, let's get in the war room on Monday and workshop it a bit, all hands on deck for this one!" "Treat the customers better." Great advice. When I worked at Nordstrom we paid sales floor staff 7% of the order ... not surprisingly, customers were treated well, and customers actually wanted to buy what we sold. Are you willing to give non-salaried employees 7% of everything? No, of course not. And that's part of your problem.

Here's something you don't ever read about ... how do you get 16,000 Saks employees to all move in lock-step in a new direction at the same time, harmoniously?

That one is tougher, isn't it?

A half decade ago I received an email from a client that they client mistakenly/accidentally sent to me (this happens all the time, and woo-boy, the stuff I get to read ... Midland Paper comes to mind). The Marketing Executive was lambasting the Management Team because I (Kevin) authored recommendations that the Marketing Executive absolutely did not want to implement. "We're not actually listening to this idiot, are we?" she bellowed to her Management Team. She was right ... they chose to not listen to the idiot. Ideas? Dead. Ideas killed off by one (1) employee. I mean, you can't get your family to agree upon Panda Express or Sonic for a quick, salty, calorie-laden dinner ... so explain to me, LinkedIn gurus, how exactly you will get 16,000 employees to embrace your ideas? Be specific. Show your work.

Early in my time at Eddie Bauer, the CEO passed a note from the company that owned us to my Sr. Vice President. He opened the note, read it, waited for the CEO to leave, looked at me, and said "Yeah, we're never going to do that", crumpled up the note, threw it in the garbage can, then continued his discussion with me. That happens a thousand times a day at big companies. It's happening "at scale" as the pundits say.

There are a small number of dynamic leaders who pull this stuff off. But my goodness, it takes interpersonal gifts of an order of magnitude to get 16,000 people to execute the way you want your company to execute.

P.S.: Do you want your salaried employees to do what you want? Try creating a reasonable bonus structure ... salaried staff get a 20% annual bonus, managers 30%, directors 40%, vice presidents 60%, Sr. Management get more (they always get more). Divide the bonus into three pieces ... a third based on sales gains, a third based on profit improvement, a third based on department goals (like personalizing all email campaigns, for instance). When a manager is earning $100,000 a year and another $30,000 is at risk and $10,000 of the $30,000 is dependent upon personalizing all email campaigns, I promise you from personal experience, things get done.

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!