February 28, 2024

Lone Wolf?

I just went to the Safeway website to look for Tempura batter.

Odds are that most people don't go buy Tempura batter ... and that's it. Maybe they're making a black bean burger, and they need seven or eight ingredients.

If an item sells by itself ... and the item has negative Merchandise Residual Value ... well, that's no bueno.

For every item you sell, measure if the item is a Lone Wolf ... does it sell by itself without any pairing of any other items, or does the item sell with other items?

Even an item that has negative Merchandise Residual Value is valuable if it sells with other items that have positive Merchandise Residual Value.

More on this topic tomorrow.

February 27, 2024

How Can A Product Or Category "Cause" A Customer To Become More Valuable?

I'll go back to my time at Nordstrom ... two decades in the rear view mirror these days.

Our Accessories merchant was a force of nature. When hand bag price points were around $200 she pushed them up to $400, and did not care one tiny bit what people thought (people thought she was nuts).

She was not nuts.

She understood back then that customer wanted a handbag for $400 .... and four months later they wanted ANOTHER handbag for $400. Those additional handbags were purchases that wouldn't have otherwise happened, incremental to the customer and to the business. Perform a Merchandise Residual Value analysis on her handbags and you'd see that her handbags "caused" customers to become more valuable.

That doesn't mean you should keep selling the same handbag in perpetuity.

It means somebody with merchandising smarts works with somebody with marketing smarts to exaggerate the importance of handbags.

That's a solid use of Merchandise Residual Value.

February 26, 2024

Merchandise Residual Value in Practice

Here are the top ten selling items for the past year for a brand ... in this example (for the sake of simplicity, all items existed the year prior as well), I illustrate total sales for each item, then list MRV (merchandise residual value) for each item (i.e. the amount of future value that is added to each customer who buys that item because the customer buys that item), and then a sum of last year's sales and total MRV for the next year. Yes, you want to sell items that sell well and add value to customers in the future.


This is the look you should have on your face when you look at Item #6.



Item #6 is a problem child. It is a low-priced item ($9.99) that sells the second-most units of any of the top ten sellers. But for every $9.99 unit sold, MRV is a negative $4.75, meaning the customers who purchase the $9.99 item spend $4.75 LESS in the next year as a consequence.

In other words, the item is a loyalty killer.

If you take last year's sales and what you lose in the next year by selling that item into account and take the sum of the two metrics down to profit, you will learn that this best selling item is actually UNPROFITABLE to your brand.

I'm guessing you want to know these facts for your business, right?

February 25, 2024

Oh, It Adds Up

Let's look at an example.

  • You have a winning item.
    • It sold 50,000 units last year at an average of $30.00 each, for $1,500,000 in sales.
  • However, this item has negative Merchandise Residual Value ... negative $10.00 in the next year.
    • This means the item will cost your brand $10.00*50,000 = $500,000 next year.

Is the item truly a winning item? That's a good question! It generated $1,500,000 in sales, but it will cost your brand $500,000 next year. In other words, it's actually a $1,000,000 item.

Every single one of you manages a brand where this happens. Every single one of you.

You can't stop customers from buying that item.

But as a marketer, you are under no obligation to FEATURE that item, now are you?

If you aren't already performing this style of work, contact me now (kevinh@minethatdata.com) and we'll get started, ok?

February 22, 2024

Lots of Choices!

I updated my product offering (click on the image below or click here to visit my Hire Me page).



There are several updates here.

  • I re-introduced my CATALOG MODELING project, given how popular it was over the winter when I wasn't even offering it. Many (MANY) catalog brands are going through a dramatic transformation, forced upon them by rising paper/printing/postage costs. In these projects, I calculate the exact number of catalogs to mail a customer annually. You'll save a significant amount of ad cost, you'll increase profit in the process.
  • MERCHANDISE DYNAMICS:  This is going to be a project that I roll out over the next few weeks. There is a significant shift among e-commerce brands to fully understand the interaction between customers and what the brand sells. You've seen some of my work over the past 1-2 weeks. Lots more to follow.
  • HILLSTROM'S CHECK-UP:  You asked for this. You want a quick-and-tidy evaluation of your business that can be executed within 1-3 days. That's what this project is. I'll combine a few aspects of Merchandise Dynamics with a full Elite Program Run, helping you understand within a few days some high-level issues that are holding your brand back.

Next week? More on Merchandise Dynamics.



February 21, 2024

MRV (Merchandise Residual Value): A Jewelry Example

Yesterday I showed you how Home merchandise negatively impacted how the customer behaved in the future.

Here's Jewelry for the same brand. Tell me what you observe:


Jewelry is generally purchased by customers in the middle of the life stage of the customer. Notice that 35 winners negatively impact customer spend in the future, while 64 winners positively impact customer spend in the future.

For this brand, Jewelry is a positive (on average) on future customer spend.



February 20, 2024

MRV: An Example With Home Merchandise

Home is one of those categories that Executives love, customers allegedly love, and analysts don't love.

Why?

Because buying Home product stalls customer evolution.

Here's a grid for Home products within this brand.


This table is just plain scrumptious.

Who buys Home Merchandise? In this case, it's generally customers in the middle of their life stage with the brand, but mostly among loyal customers.

So, our best customers are generally buying Home merchandise.

Do winning Home items help/hurt future customer spend?

  • Winning Home Items HURT future customer spend.

I see this happen a lot - have been watching this for more than three decades. Once that customer buys sheets or furniture, the customer doesn't need to buy again for awhile - even though you offer other products that the customer could purchase regularly.

As a marketer, you have to be very careful WHO you offer Home products to. You might think you are helping the customer cross-shop your entire brand. If you cause your customer to spend $50 less next year because of it, you've accomplished little and likely hurt the customer relationship.



February 19, 2024

MRV (Merchandise Residual Value): Placing All Those Dots In A Grid

You saw the graph yesterday.


 
The average business professional will glaze over when seeing the data presented in that manner.

Simply the presentation to a 3x3 grid, and you have something.


In total, yes, the counts should be reasonably equal.

Look at a merchandise category, and the story changes. Let's evaluate one such story tomorrow, ok?

February 18, 2024

MRV (Merchandise Residual Value): The Residual Plot

Remember, MRV (Merchandise Residual Value) is the amount of customer spend in the next year caused by each winning item, after controlling for where the customer is in the life stage of the customer when buying the item.

If an item adds value to the customer, you want to feature the item.

If an item causes a customer to purchase less often in the future, you want to de-emphasize the item, right?

Are you calculating this metric?

Here's a residual plot for a brand with more than 70,000 items sold per year and more than 2,000 winners per year.



Items purchased by a customer early in the life stage (i.e. a 2nd or 3rd purchase) don't have a lot of MRV variability ... customer spend in the future is changed by +/- $20, which isn't nothing, but it isn't what we see later.

When an item is purchased by customers buying for the 8th time, on average, future customer spend is impacted by +/- $50. In other words, if a customer is expected to spend $300 in the future and the customer buys an item that harms MRV, the customer trajectory is changed by the item and the customer now spends $250 in the future instead of $300.

The items we sell help dictate how the customer will behave in the future. Why are we repeatedly featuring items that cause customers to spend less in the future?



February 15, 2024

MRV Yields "Some" Bad Items and "Some" Good Items

This is the way MRV (Merchandise Residual Value) looked for the Winning Items at one brand:



Most items act "normally" ... but there are 20+ items that perform terribly (i.e. they deliver customers that, after controlling for the quality of the customer, deliver customers with reduced future value). Conversely, there are 20+ items that perform really well ... the customers who purchase those items are worth more in the future BECAUSE they purchased those items.

It's really, really important to know which items deliver customers that BECAUSE they bought the item are now worth MORE than they would have been previously.

This knowledge changes how the marketer does every aspect of his/her job.

It's that important.

Are you performing this analysis?

If not, contact me immediately (kevinh@minethatdata.com) and we'll get busy.



February 14, 2024

MRV: Merchandise Residual Value

"MRV" is what I call "Merchandise Residual Value". It's the amount of incremental value the item generates on an annual basis, per customer purchasing the item, above-and-beyond expectations.

We control for the LIFESTAGE of the customer buying the item. Here's a sample dataset of six items. For each item, I calculated the LIFESTAGE of the customers buying the item, and I calculated the FUTURE spend customers who purchased the item in 2022 spent in 2023.

  • Winner #1:  Lifestage = 3.00.  Future = $200.00.
  • Winner #2:  Lifestage = 3.50.  Future = $235.00.
  • Winner #3:  Lifestage = 4.00.  Future = $180.00.
  • Winner #4:  Lifestage = 4.50.  Future = $305.00.
  • Winner #5:  Lifestage = 5.00.  Future = $285.00.
  • Winner #6.  Lifestage = 7.00.  Future = $400.00.
Let's graph the relationship.


There's a clear relationship - if items are purchased by customers later in the LIFECYCLE, those items produce better Merchandise Value.

That's not what we are looking for.

We went Merchandise Residual Value (MRV) ... the difference between the points on the graph and the orange line.

The equation is a straight linear Ordinary Least Squares Regression equation.
  • $34.63 + $51.75*(LIFESTAGE).

For each item, we know the FUTURE spend of customers buying the item, we know the predicted amount of future spend, so the residual is FUTURE - PREDICTION.

  • Winner #1:  Lifestage = 3.00.  Future = $200.00. Pred = $189.99. MRV =  $10.12.
  • Winner #2:  Lifestage = 3.50.  Future = $235.00. Pred = $215.76. MRV =  $19.25.
  • Winner #3:  Lifestage = 4.00.  Future = $180.00. Pred = $241.63. MRV = ($61.63).
  • Winner #4:  Lifestage = 4.50.  Future = $305.00. Pred = $267.51. MRV =  $37.50.
  • Winner #5:  Lifestage = 5.00.  Future = $285.00. Pred = $293.38. MRV =  ($8.38).
  • Winner #6.  Lifestage = 7.00.  Future = $400.00. Pred = $396.88. MRV =   $3.12.


Which item has a poor MRV?  Winner #3 does. At an average LIFESTAGE of 4.00 orders, the item should produce customers worth $241.63 ... instead it produced customers worth $180.00. MRV is negative ... -$61.63.

This is so ... darn ... neat!!

So ... darn ... neat!!

If you are going to feature merchandise in social or email or print, you might de-emphasize Winner #3 because it produces customers who spend less in the future, with an MRV of ($61.63).

Think about all the ways you'll use MRV in your business?

Now get busy using MRV!!

February 13, 2024

Once You've Defined A Winner ...

... it becomes your job to determine if the winning item helps or hurts your brand.

How can a winning item possibly hurt your brand?

Easy.

If the winning item attracts buyers who choose to reduce their spend in the future, then the winning item hurts your brand.

Here's what I want you to do.

Take every winning item from 2022, and perform two calculations for me.

  1. Calculate where in the customer life stage the customer was when the customer purchased the item. Was this the 14th purchase for that customer? If so, create a variable called "LIFESTAGE" and populate the column with a 14 in this instance.
  2. For every item sold in 2022, calculate how much the customer who bought that item in 2022 spent in 2023. Call this variable "FUTURE".

Now, for each winning item, calculate the average "LIFESTAGE" the customer resides in when placing the order. If an item was sold three times and the three customers were on the 1st, 3rd, and 8th order in their LIFESTAGE, the average LIFESTAGE for the item is (1+3+8)/3 = 4.00.

Next, for each winning item, calculate the average "FUTURE" amount the customer who buys the item spends. If the item was sold three times and the three customers spent $0, $0, and $300 in the next year, the average FUTURE for the item is (0+0+300)/100 = $100.00.

Things are about to get interesting!

Tomorrow, we'll get to the interesting part.

And To Think You Are Required To Implement Some Idealistic Omnichannel Strategy To Compete Against This

They've been around about a year-and-a-half ... they advertised during the Super Bowl, and they're going to run circles around the tactics your agency demands you implement ... they're Temu (click here).

February 12, 2024

Defining Winners

My thinking evolved a lot over the past decade. These days, to me, "Winners" or "Winning Items" are items that represent the top half of annual sales (or in many of my projects, I calculate winning items on a daily basis). In other cases, I just grade the top 20% (sales) of items as "A", next 20% as "B", next 20% as "C", next 20% as "D", and the remaining 20% (which is often 60% of items) as "F".

In one project, there were about 2,600 winners out of 72,000 items sold in the past year. In other words, 2,600 out of 72,000 items accounted for half of annual sales.

And so it goes with most of the brands we manage. Apple has the iPhone, then generates less sales from Macs and Watches and Airpods etc. The iPhone is the Winner.

Winners (and sales of most items, for that matter) follow a law of diminishing returns. You can't just keep expanding your assortment and expect sales to grow linearly.

All of us eventually max out our assortment. We do not come up with creative solutions, and our ability to grow via product ends. This is where marketers are called into the equation. Marketers help in two different ways.

  • New Markets/Channels.
  • Improved Perception of the Brand.

Modern marketers could help in a third way.
  • Developing new/existing items into winning items.

Modern marketers work hard to develop customers.

Modern marketers have time to develop winners as well.

Winning items generate the profit that pays our salaries, bonuses, health insurance, 401k, you name it, they're responsible for it. When we have more winning items, we have more of the things we need to be successful.

February 11, 2024

Diminishing Returns In Action

Maybe you don't believe all of the diminishing returns curves you see me present, in one form or another. So let's take a one-day detour into the world of Major League Baseball.

The image below evaluates data for MLB team spend and number of wins the team earned during the 2023 season.



The orange line represents a fitted power function (a*x^b). On the bottom right of the image, you can see how much money you have to spend to buy wins.
  • Going from $100 million in payroll to $150 million yields about 4.4 wins per season ... the difference between a team going 80-82 vs. 76-86, approximately.
  • Going from $150 million to $200 million yields 3.3 wins per season.
  • Going from $200 million to $250 million yields 2.6 wins per season.
  • Going from $250 million to $300 million yields 2.2 wins per season.
So a $100 million payroll team can expect to go 75-87. A $300 million payroll team can expect to go 88-74.

If you were an executive running a baseball team, you'd probably ask yourself if you truly had to spend $200 million to purchase 12 additional wins ... or was there "something else" in the data that could be used to manufacture 12 additional wins???

But notice diminishing returns happening in baseball. Each additional fifty million dollars buys you fewer wins.

The same things happens in your business ... quite literally everywhere. Each additional customer is generally worth less. Each additional product you offer is generally worth less. Each additional marketing dollar is generally worth less.

February 08, 2024

A Small Number of Items Make A Big Difference

Here, we see cumulative annual demand (y-axis) vs. the fraction of the total number of items sold in the past year (x-axis).



Half of all sales (in this case) come from 6% of the assortment.



If you are an analytics professional, you dive into that small number of items. In the case of this brand, it's around 400 items, total.

Four hundred.

How do customers who buy those items behave subsequently?



February 07, 2024

What Your Assortment Looks Like

Have you ever asked your marketing team to describe the merchandise assortment sold by your brand?

You'll get a different opinion than what your merchandising team says, that's for sure.

Now, have you ever asked your analytics team that question?

Your analytics team is not going to have product knowledge, and they sure won't understand how the assortment "communicates the myriad values of the brand proposition" that you'll hear from your marketing team.

But they might start off with a graph of annual sales by item, ranked by 1%-tiles.


This brand sold 72,000 different items last year. Only a handful of the items generated meaningful volume. Somebody on the inventory team has to know that the 55,000th best selling item sold 9 units at $20 each, and has to have the ability to get those items and have them shipped quickly.

Meanwhile, almost all customers have a relationship with about 12% of the items being sold ... about 8,000 items in total.

The analyst, therefore, is going to take a different view of this brand. The analyst is going to focus on +/- 8,000 items ... probably more ... and will make judgements about how those items contribute to the success of the brand.



More on this topic tomorrow.



February 06, 2024

Which Item Do You Prefer As A Marketer?

Item #1:

  • Annual Sales = $100,000.
  • Future Value of Customers Buying The Item = $12.00 Profit Next Year After Controlling For Differences In Customer Quality.

Item #2:
  • Annual Sales = $50,000.
  • Future Value of Customers Buying The Item = $27.00 Profit Next Year After Controlling For Differences In Customer Quality.

I should probably ask if you measure the future value of comparable customers buying each item in your assortment?


February 05, 2024

Corn Yield Wheel

My first two years out of college were not in retail. Noooo. They were spent at the Garst Seed Company in Slater, IA. I analyzed corn and sorghum research plots, trying to identify the hybrids best adapted to different geographic regions. When it came to people and management and Dunder-Mifflin style office politics, I was an idiot. When it came to wanting to learn about maximizing yields, I was a sponge.

On a crisp October morning, one of our Agronomists took me to a meeting to discuss "density tests". Oooooh!

A handful of analysts and agronomists stood around what was called a "wheel test", if I remember correctly. On the inside of the wheel corn was planted at a high density - seeds planted close to each other. On the outside of the wheel corn was planed at a low density - fewer seeds were planted, but were planted farther apart from each other.

Naturally, corn on the outside of the wheel was tall, with huge cobs.

In the middle of the wheel, corn wasn't as tall, but the cobs were nice-sized.

In the inside of the wheel, corn was short, crowded, with tiny cobs. But there were a lot of tiny cobs.

The agronomists showed us that the secret to high yields was to plant the seeds as close as possible without pushing the spacing too close to each other. In a world where you could have 10 huge cobs, 20 large cobs, or 30 small cobs, you wanted 20 large cobs. That's how you maximized yield.

Standing outside on a chilly October morning in 1989, you'd never guess how important that concept would be to marketing strategy in 2024.

In 2024, you have a daily or twice-daily email "blast" ... a daily Facebook post, a daily Instagram post, TikTok videos, YouTube videos, Pinterest content, something on Snapchat because Heather says you need to do it, loyalty mailers, coupons, retargeting ads hounding the customer all across the internet ... it never ends.
  • And yet, your sales don't increase.

How is it even possible that you can speak to the customer a dozen times a day and the customer doesn't spend one penny more than the customer would have spent 20 years ago?

It's the corn wheel.

If you plant a dozen seeds a day right next to each other, the yield of each individual seed is terribly low. And yes, that's what all of us in this industry are doing.

Sometimes when I'm feeling sarcastic, I want to say something like "you have 20 communications with a customer daily and the customer still visits Google and is apparently performing some level of comparison shopping ... what was the point of all of the communications?"

The apparent goal of an "omnichannel communication strategy" is to yell at the customer with a bullhorn from multiple directions. "It all works together, fueling frictionless synergy!"

Alright.

If you could say one important thing or twenty unimportant things ... per day ... what tactic might you test?

February 04, 2024

Diminishing Returns

I am strongly leaning toward creating something regarding Diminishing Returns as a marketing/business concept. Might be a booklet, might be a video series, gonna be something.


This is one of those concepts that is popular but largely unprofitable for somebody like me:

  1. The concept is not well understood.
  2. The concept, when explained, becomes quite popular.
  3. The concept is terribly difficult to monetize in spite of the importance of the concept. In other words, I could have 100 inquiries about the concept and when I say that it would cost $5,000 for me to develop a solution for a client, I'll get zero (0) new projects as a consequence.

If you think a booklet or video series makes sense, let me know (kevinh@minethatdata.com).

If you think some sort of project offering makes sense, let me know and let me know what you think a fair price is for a project offering where I figure out your "diminishing returns" relationship.

Seem fair?




February 01, 2024

Aligning With The Business

Like this.


What happens is you become an overexposed technology leader, because you are trying to convince people to do things that a third party wants you to do - you don't understand the actual business issues your own company faces and you don't understand the technology a third party is trying to sell you.

It's not just the people reading trade journals for inspiration that fail to align with the business.

About a decade ago I sat in an Executive Boardroom. The marketing SVP was not even invited to the meeting - they brough him in at 2:00pm like a relief pitcher. The marketing SVP said a few things about efficient bidding on keywords, open rates on email campaigns, and the raw power of social media. By 2:30pm he left the room.

That's when the CFO looked at me and said "he has no idea what is going on around here".

I spent a couple of decades trying to do what I thought was right for a business. I should have spent that time trying to do what was right for the business.

Align your efforts with those paying your salary.








End of the Line for the Orvis Catalog

When a traditional catalog brand makes changes, I get a head's up via blog / email subscriber metrics. Blog unsubscribes happen from the...