Let's Just Attack Those Lapsed Buyers!!

FYI - Hillstrom's Rebuild is available on Amazon - click here.

Alright.

Let's pretend you have a magic elixir that can get any customer > 12 months of recency to purchase at a rate 10% better than what you currently experience. Ohhhhhhh Boy, one might think you've just changed the equation, right?

Let's pop that into my Lifetime Value Simulation (#LTVS) and see if it makes a difference.

In my initial run, 66,917 of 100,000 simulated customers did not repurchase after four years.

But by #targeting lapsed buyers (those with recency > 12 months), a magic 10% increase in response yields a scenario where 66,197 of 100,000 simulated customers did not repurchase after four years ... I'll spare you the details.

Let's just round the results.

• By working our tails off to #reactivate lapsed buyers we prevent a case where 67,000 of 100,000 customers do not purchase again during a four year period of time ... and instead, we face a scenario where 66,000 of 100,000 customers do not purchase again during a four year period of time.

If you want copies of the two tables, send me an email message (kevinh@minethatdata.com) and I'll share the results with you.

The problem, my friends, is we wait until it is nearly too late to do anything. Why wait until the customer lapses before acting?

Why not act when the customer is active ... in the first three months after a purchase?

More on that tomorrow.

An FAQ On Developing Customers

FYI - Hillstrom's Rebuild is available on Amazon .. click here to purchase.

You read the post yesterday. You think I'm an idiot. You have questions. I have answers.

Question: Kevin, we all know it costs eight times as much to acquire a new customer as it costs to retain an existing customer. Your data has to be wrong. Why is your data wrong?

• We are looking at the data differently. You cherry-picked your good customers and looked at all the profit they generate and concluded that you shouldn't spend money on new customers. Point of inquiry ... how do you suppose you found those "good customers"? Oh, at one time, you had to "acquire" them, didn't you? In other words, you don't get "good customers" unless you have a robust "customer acquisition" program ... you have to spend the money to acquire the customer to enjoy having a good customer.

Question: Kevin, I looked at my own customer file, and we have a ton of loyal customers. Why does your simulation show so few loyal customers?

• I am looking at what happens to a cohort of customers. Almost none of the customers become loyal. You acquire customers every month ... and have been for thirty or forty years. So yes, you have loyal customers. But divide the number of loyal customers you have by the number of customers you've ever acquired over forty years, and tears will come out of your eyes.

Question: Kevin, if we just work harder at email marketing and get customers to open more emails, or if we overlay web data on our file, or if we #engage with customers on social media, or if we do a better job with paid search, or if we do a better job with natural/organic search, or if we partner with Pinterest, or if we pay Facebook more, or if we can get more than 7 in 10,000 to click on a retargeting ad, or if Amazon stopped hurting us, then the data you are sharing would look different, right?

• You have thousands of channels today ... you had dozens of channels in 1995. How has your annual repurchase rate changed in 21 years? It's similar? Or worse? Then what did all of those channels do for you in the first place? The channels made vendors money, that's for sure!! Maybe that's why you hear vendors constantly talking about channels. Think about it for a moment.

Question: Kevin, if we really dig into lapsed buyers, then we can change the dynamic ... we can flip the script, can't we?

• Tomorrow, I will show you that it's in the first three months after a customer purchases that we can make improvements ... by the time the customer lapses, you aren't able to fundamentally change the long-term trajectory of the simulation ... you can't change it, because the probability of subsequent purchase is simply too low.

Question: Kevin, your data makes it seem like we should almost ignore trying to get customers to become more loyal and instead focus all our efforts on finding new customers. Am I reading you correctly?

• Yes. Yes you are. 

Question: Follow-up question, Kevin. What you are suggesting goes against all best practices in our industry. You have to be wrong. Wisdom of the crowds, Man!!

• Did you watch the movie "The Big Short"? A handful of people bet against the housing market. They analyzed the situation, they realized that everybody was wrong, and when the housing market crashed, they got no credit for being right. I have hundreds of client data points, billions of purchase records spanning two decades, and more than a million lines of computer code I've written to study this dynamic. The story is consistent. If you double your customer acquisition counts, you'll eventually double your loyal buyer counts. Now, if you have a million+ lines of code you can forward me that prove the opposite, by all means, I'll publish your findings.

Question: Another follow-up question, Kevin. Customer acquisition is terribly expensive. We'll go broke finding new customers before we double the number of new customers.

• That's not a question, that's a statement. More important - I've spent the past year asking you to implement low-cost / no-cost customer acquisition programs. I've given you hundreds of examples. How many of the examples have you implemented? I'm more than twenty-eight years into my career ... there are two truths I've learned over time ... first is that low-cost / no-cost customer acquisition programs are critical to business success ... and second is that new merchandise development is critical to business success. Yup, nail those two and you can retire.

Developing Customers

FYI - my new booklet ... Hillstrom's Rebuild ... is now available on Amazon (click here).

When you run a Lifetime Value Simulation (#LTVS) of first-time buyers, you quickly face an uncomfortable reality. Here's the reality you face.

• "Wow, most of these customers just kind of drift off into oblivion."

Yes. Yes they do.

Now, if you work at Starbucks or Wal-Mart or Target or Nordstrom, different rules apply.

But you don't work for those companies. I know that to be true. I look at my visitor analytics and email metrics. Your company sells Widgets, and let me tell you, the rules are different if you sell Widgets.

Here are the results of a recent simulation for an average of businesses with a 37% annual repurchase rate ... the kind of repurchase rate that is AVERAGE across my entire client base. The simulation tracks the progression of first-time buyers for the next four years. I am going to present you with a series of tables of 100,000 simulated newbies, showing you how many end up in different recency/frequency cohorts.

Ready to be disappointed? Good!

Here's the table for the 0-12 month file, four years after 100,000 customers were newly acquired.

Look at the counts, my friends. Look at the number of customers with 10+ purchases four years after being acquired. The number is virtually zero!

Oh, I know, your are told there is a pot of gold with lapsed buyers. Here's your 13-24 month file.

Again ... there's a hundred thousand customers in this simulation.

Your 25-36 month file, enjoy!

#OhBoy.

Through the start of month 49.

Look at the lower right portion of the table ... read across LTD Orders = 1 ... read across to RECENCY = 49 (49 Months) ... what do you see?

You see 2/3rd of the original purchase cohort sitting there ... idle ... inactive.

You aren't going to #engage these customers ... you aren't going to execute a #campaign chocked full of #relevancy and #personalization either ... they're inactive. They bought, their need was met, and they've moved on.

How many life-to-date purchased did these customers place?

1. 66,917 (out of the original 100,000 mind you).
2. 14,756.
3. 5,980
4. 3,404
5. 2,431
6. 1,595
7. 1,249
8. 911
9. 724
10. 541
11. 412
12. 295
13. 253
14. 171
15. 122
16. 79
17. 45
18. 46
19. 20
20. 21
21. 17
22. 6
23. 3
24. 2

In fact, 2,033 of the original 100,000 first time buyer purchased 10+ times in four years.

That's it.

This is why I repeatedly talk about what a waste of time it is to give discounts and promotions to move customers up the loyalty chain. It is Fool's Gold. Given the opportunity to spend money, double the number of new customers - you'll have twice as many loyal buyers in five years.

Tomorrow, an FAQ on this topic.

The Life Table

In any Lifetime Value Simulation (#LTVS), the secret to measuring how customers develop is outlined in a Life Table analysis.

You can click here to learn more about how to perform your own Life Table Analysis (via Wikipedia).

The Life Table tells us the probability of a customer purchasing in the next month, given that the customer did not previously purchase in the past "x" months.

For instance, for a first-time buyer, the first three months of a life table might look something like this:

• Month 1 = 8%.
• Month 2 = 6%.
• Month 3 = 4%.

Let's pretend we have 1,000 new customers.

After one month, 1,000 * 0.08 = 80 customers purchase, meaning that 1000 - 80 = 920 customers do not purchase.

• 80 Customers Now Have Recency = 1 Month and Frequency = 2 Purchases.
• 920 Customers No Have Recency = 2 Months and Frequency = 1 Purchase.

Let's look at the 920 customers who have yet to purchase. In Month 2, they have a 6% chance of purchasing. 920 * 0.06 = 55 customers purchase, meaning that 920 - 55 = 865 customers do not purchase.

• Of the 1,000 first-time buyers, after two months, 865 have yet to purchase for a second time.

We move on to Month 3. The 865 customers who have yet to purchase have a 4% chance of purchasing. 865 * 0.04 = 35 customers purchase, meaning that 865 - 35 = 830 customers are still stuck on one order.

• Of the 1,000 first-time buyers, after three months, 830 have yet to purchase for a third time.

Similarly, for customers with 2 purchases, or 3 purchases, or 27 purchases, we can create life tables that allow us to measure how likely these customers are to purchase in any given month, given where the customer resides in the life cycle.

The life table makes a Lifetime Value Simulation (#LTVS) possible.

It's in the Lifetime Value Simulation that all of the magic of your business is revealed.

Customer Planning and Development

When I was hired at Lands' End, back in 1990 (those were the days when we questioned why a President didn't like to eat broccoli - good thing we moved on to more substantive issues when discussing the Commander in Chief), I was hired into a department called "Customer Planning and Development".

The name of the department says a lot about what the department was supposed to do.

• Plan.
• Develop.

It's really hard to manage a customer ecosystem when all you do is "target" customers with "discounts and promotions" via "campaigns".

In other words, modern marketing has nothing to do with "Customer Planning and Development".

Over the next several blog posts, I'm going to share the results of some of the Lifetime Value Simulations I run for clients (#LTVS). You wonder why I advocate developing a customer file through customer acquisition activities? You won't once you see how a customer file "develops" over time.

I realize this topic is alien to many. You won't find a Woodside Research report that provides the insights that are coming over the next several posts. Nor will you find the insights in Google Analytics. And you may disagree with everything I have to say. That's fine. But you are going to learn how a customer "develops" over time.

Ready?

Jump Around

Back in 1998, at the end of the third quarter of a game where Drew Brees would eventually throw 83 passes as a Sophomore at Purdue ... losing 31-24 to Wisconsin ... a tradition was born.

The raucous crowd ... Wisconsin head coach Barry Alvarez later said that "you couldn't hear yourself think" ... was treated to a song on the PA system at the end of the third quarter ... House of Pain's "Jump Around". The student section obliged.

The rest is history.

Watch any Badgers home game on TV, and you'll get to see 80,000 fans jump up and down to the song. If the score is 43-3, you'll see a third of the crowd leave after the song is over ... they're staying for the song.

"Jump Around" is now a tradition.

What traditions do you rally your customer base around?

Oh, I know, I know, I can hear it already. "We sell widgets, there are no traditions surrounding the sale of widgets. It's all about price and promotion, and you'd know that if you paid attention to the widget industry."

Nonsense.

What do you think "Cyber Monday" is, after all?

Cyber Monday is a silly tradition created by a trade organization. Nothing more, nothing less. But the trade organization did a great job of telling a story - of telling the story to the press (and to you) - of getting the press (and you) to believe in customer behavior that actually happened for two decades prior (at least) - all resulting in the need for your brand to give away profit on the Monday after Thanksgiving just so folks can sell advertising space in anticipation of a lot of page views and clicks. You give up profit, they make profit. Nice tradition, huh?

So don't for one minute tell me that you don't have any traditions. 

If you're willing to participate in Cyber Monday to make a trade organization and the "press" happy, might you be able to create your own tradition and keep all of the profit in the process?

If the answer is "no", maybe it is time to hire some creative marketing chops, right?

Cabela's

Here you go ... Bass Pro Shops purchased Cabelas (click here).

One of the biggest frustrations of my career has been the lack of understanding of the profit and loss statement by all but a handful of employees. You will have pundits tell you that you must "engage" your customers and "be relevant" or your brand will "die". These folks, clearly, do not know how to calculate profit.

After a recent bout of frustration, I decided to create an equation. I translated sales growth and pre-tax profit percentage into a Won/Lost record, and then explain the trajectory of a business based on the Won/Lost record. I decided to use NFL Won/Lost records, since an awful lot of people seem to be able to understand that the Arizona Cardinals are in trouble with a 1-3 record. The Won/Lost record is easily understood. Profit and Loss? Not so much.

Each year's Net Sales Growth Rate and Pre-Tax Profit level is translated into a Won/Lost record ... a 3-13 year or a 6-10 year gets the CEO fired (I'll share Lands' End data at some point and you'll see what I mean) ... several years of 7-9 or 8-8 or 9-7 cause changes to happen ... a bunch of 15-1 seasons and you are Steve Jobs.

Here is the fourteen year trajectory of Cabela's, using my new equation.

Now, I just ran a table like this for Chicos - not so good (i.e. three years with performance similar to a 6-10 NFL team). The table for Cabela's is pretty credible, don't you think?

Here is the Won/Lost record, depicted graphically.

The bars are set at an 8-8 record ... which direction are most of the bars pointed? In most cases, Cabela's posted what my methodology calls a "Winning Record" ... this business has been run like a winning NFL team is run. There are only two "losing seasons" in the bunch ... 2009 and 2010 ... coming out of The Great Recession.

When the record goes above ten wins, I consider the brand to be a "Playoff Team". In the NFL, you want to be in the playoffs, right?

When the record goes above twelve wins, I consider the brand to be a "Championship Caliber Team".

In the past fourteen years, what did Cabela's post, using my framework?

• 12 winning seasons in 14 years.
• 4 playoff caliber years.
• 2 championship caliber years.

So that's a pretty good run!

And you can see that there is a bit of a "quadratic" trend happening from 2011 - Current, right? There's risk involved - sales are growing, profit is in decline. Year-to-date, Cabelas posted a sales gain (about 7%) and a modest decline in operating profit percentage ... which will yield another 9-7 season when translating sales/profit trends to NFL won/lost records.

Yes, you are going to hear a lot more about this methodology ... give it a few weeks. There's even going to be a booklet about the methodology - linking the scoring method to the start of what I call a "rebuilding process" for companies I work with.

So stay tuned!