Median Time to Loyalty

Once we know the probability of a purchase at any recency/frequency combination, we can calculate the median time that elapses between a first purchase an loyal status (i.e. 5th purchase): 

Look at the 36-month recency level. After thirty-six months, 36.6% of customers who bought for a first time repurchased. Now read up the column until you see the median ... the period where 18.3% of customers repurchased. That's at the approximate 7 month level.

Perform the same exercise for 2x buyers, 3x buyers, and 4x buyers. You see that the median repurchase interval is about 4.5 months, 4.5 months, and 3.5 months. As the customer moves through the Emergence phase, the customer becomes more likely to repurchase, and tends to repurchase faster.

Add up the median repurchase intervals ... 7.0 + 4.5 + 4.5 + 3.5 = 19.5 months.

It takes about 19.5 months for the customer to achieve loyalty ... if the customer is willing to make the journey. Median Time to Loyalty ... MTL ... is 19.5 months.

Now multiply the percentages at recency = 36 months ... multiply the 1x-2x rebuy rate by 2x-3x by 3x-4x by 4x-5x. What percentage of the customers (based on 36 months of repurchase activity) become loyal (i.e. make it to five purchases)?

The answer is 11%.

You've got a lot of hard work to do, don't you?

• 11% of first-time buyers will become loyal.
• MTL ... Median Time to Loyalty ... is 19.5 months.

Instead of focusing on loyalty programs, why not focus on Welcome/Emergence efforts that reduce MTL and increase the fraction of customers who become loyal?

Big Differences

Yesterday we evaluated "Brand A", with an approximate 30% annual repurchase rate.

Here is "Brand B", with an approximate 45% annual repurchase rate.

These tables are so darn "delicious".

Read across the Recency = 1 rows in both tables. Customer Development in each table (meaning each brand) is essentially the same.

Read across the Recency = 3 rows in both tables. Customer Development in each table (meaning each brand) is different! "Brand A" is losing customers, "Brand B" is doing a better job of keeping customers.

Read across the Recency = 12 rows in both tables. The differences are stark.

Read across the Recency = 24 rows in both tables. The differences are like night and day.

By the way ... both brands are in the same category, so it's not like one brand sells something that customers want less often. Nope. This is what it looks like when one brand is really good at Customer Development. And this leads to an annual rebuy rate that is fundamentally better at "Brand B". That leads to a powerful file that leads to much more profit, so much more profit that "Brand B" can spend more on marketing programs, causing more repurchase activity, making Customer Development look even better.

If you are "Brand A" and you don't have money to do what you want to do, get clever.

• Hire some people who know how to make money on Instagram.
• Hire some people who know how to create entertaining videos.
• Hire a vendor who knows website personalization inside and out.
• Hire a vendor who knows email personalization inside and out.
• Create your daily podcast where you interview your merchants who crave the products they are selling.
• Develop in-house influencers and unleash them on your "community".

The big differences between each brand represent the fundamental difference in marketing strategy between "Brand A" and "Brand B". Marketing matters.

Are you "Brand A", or are you "Brand B"?

Do you know how to evaluate whether you are "Brand A" or "Brand B"? Email me for details and I'll help you figure it out.

Comparing Two Companies

Here's the Average Monthly Grid for Brand "A" (this brand has an approximate 30% annual rebuy rate).

And here is the Average Monthly Grid for Brand "B" (this brand has an approximate 45% annual rebuy rate).

Tell me what you see here?

I'll give you a day to find the key difference between the two brand. Talk to you tomorrow!

Loyal Buyers Decay, Too

Why do you need a loyalty program (and FYI, before I spent so much time on Customer Development in 2020-2021 I wasn't a fan of 'em, but I'm warming to them now)?

Let's look at our table from yesterday, what I call the "Average Monthly Grid", showing average twelve-month rebuy rates by recency/frequency combinations.

Read down the "Freq = 7" column. 

Once the customer purchases for the seventh time, the customer has a 73.4% chance of buying again in the next year. That customer is "loyal" based on my definition (>= 60% chance of buying again in the next year).

What happens if the customer lapses just two months, to a recency = three months?

Well, the repurchase rate is down to 59.6%.

According to my definition of "loyal" (>= 60% chance of buying again in the next year), the customer is no longer loyal.

I mean, seriously, the customer just bought for a seventh time and already after just two months the customer is in a serious state of decay.

The loyalty program isn't necessary for a month or two ... it's the opposite of the Welcome situation (where marketing is badly needed in months 1-3 after a first purchase). The customer is quite likely to buy within months 1-2 following a seventh purchase. The loyalty program should be designed to reign-back-in the customer who has just fallen out of a state of loyalty.

That's what a loyalty program should do, quite honestly. It should not interfere with normal buying. It should help bring the formerly loyal buyer who lapses a few months back into the fold, because loyal buyers decay, too.

Utterly Beautiful

I could look at this image all day long (to which you might say, "Get a Life, Kevin"):

The image highlights a bit of knowledge that is Central to Customer Development ... "Customer Value Decays".

It's easy to see it when looking at a first-time buyer. Immediately after a first purchase, the customer has a 24.4% chance of buying again in the next year. If the customer doesn't buy that month, the probability of a purchase in the next year drops to 18.6%. Nearly 25% of the future buying potential just disappeared ... after only one month. That's why the Welcome Program is so darn important.

From there, the customer just keeps on decaying. Merchandise will keep the customer interested, no doubt about it. But Marketing plays a huge role in keeping the customer interested. There's the operational stuff like search, there's the owned stuff like email marketing and website personalization, there's content-related entertainment on Instagram and YouTube. All of it matters. You matter!

To me, the table is utterly beautiful.

Tomorrow we'll look at this table again, reviewing loyal buyers.

The day after? We'll compare this company to another company.

Running An Entirely Different System

If you like basketball, then this is a must-read (click here).

Notice in the article that the coaches running "The System" are not following best practices.

I ran a poll on Twitter, asking readers to tell me how confident they are when a vendor says that the vendor is sharing "best practices" that the vendor is actually sharing best practices (click here). The results were not in favor of vendor suggestions of best practices.

One way of winning in business is to do something different. When everybody is doing the same thing, you go down a different path. How do you think Stitch Fix got to over a billion in sales in under a decade? I mean, the TOLD YOU WHAT YOU WERE GOING TO WEAR. Their "system" is fundamentally different than you having an omnichannel assortment of sameness available in all channels, available to everybody at the same time.

Keep these ideas in mind as you consider your Customer Development tactics.

It's Here: Hillstrom's Customer Development

It took several months to organize the ideas, but here we are ... Hillstrom's Customer Development is now available in Paperback and Kindle (click here).

We have COVID to thank for this booklet and the #1 project request I currently field from clients. Maybe 75% of my client base had gigantic gains in Customer Acquisition since the start of COVID. Clients kept asking a simple question ... "Are these customers converting to loyal buyers??"

I use Twitter to flesh out ideas / hypotheses. It was clear from those following on that platform that very few Professionals had an understanding of how customers migrate from prospect to a purchase to loyalty, if the customer migrates at all. Hence, we have this text, to explain the process.

Now, if you don't want to buy a book and don't like being "sold to" go ahead and click here for a PDF outlining the concepts in the text ... it won't cost you a thing.

If you are in the camp that needs help understanding if your Customer Development efforts in the age of COVID are bearing fruit, click here for project pricing and file layouts.

Whether you want free content, or to pay a small price for a booklet, or you want me to run the analysis for you, rest assured that if you don't understand if you do a good job of Developing Customers, you won't be as profitable over the next few years as you should be. And based on the feedback I've received over the past nine months, very few professionals understand if they do a good job at Customer Development.

How Do I Know If I Have A Customer Development Problem?

For one thing, you look at annual repurchase rates. They're generally low, and that tells you that you have a Customer Development problem.

Your repurchase rates might be more than acceptable, but all of your marketing is generic, the same to every customer. That's a problem.

If your customer ages one year for every year that passes, you have a Customer Development problem.

If you have a low annual repurchase rate and a high number of skus, you have a Customer Development problem.

If you have a highly seasonal business where 40% of sales happen in November/December, you have a Customer Development problem.

If customers are purchasing via paid search in a 10th order from your brand, you have a Customer Development problem.

If email marketing isn't responsible for a minimum of 20% of your online sales, you have a Customer Development problem.

If you aren't generating at least $0.20 per email delivered, you have a Customer Development problem.

If you don't do something unique and personalized to motivate a first-time buyer to purchase for a second time quickly, you have a Customer Development problem.

If a customer is highly responsive at recency of 11/12/13 months (due to strong seasonality), and you aren't priming the pump to recognize this behavior at a customer level and then act upon it, you have a Customer Development problem.

If a customer has purchased five times and all purchases are from one merchandise category, you have a Customer Development problem.

If a customer is offered a promotion, purchases at 30% off, and then continues to buy only when promotions are offered, you have a Customer Development problem.

If a customer buys in stores and is marketed to via email and is constantly offered messaging that demands the customer buy online right now and the customer listens and buys online and greatly slows in-store purchases, you have a Customer Development problem.

I could go on. But now you know what you should start looking at in your data/practices.

Each Audience

In a Customer Development project, each customer is scored. I predict how likely the customer is to purchase in the next year. I predict how much the customer will spend if the customer buys next year. Multiplied together, we have "12-Month Sales Value" or "TMSV".

There are many ways clients use this information.

Among first-time buyers, companies quickly personalize the experience of the first-time buyer.

Let's say a customer bought one item on a first order, from an affiliate, from category 05. My model might predict a probability of purchase in the next year of 17.3%.

Let's say a customer bought four items on a first order, from email marketing, with category 08 as the primary category. The model might suggest that the customer has a 30.3% chance of buying in the next year.

Are you going to treat each customer the same?

For catalogers - should each customer get the same catalog?

Email marketers - should each customer above get the same email frequency? Same content? Same creative presentation? Same messaging?

Discounts / Promotions - should they be the same? Different?

I could go on.

If you are working through your Customer Development efforts, you have different tactics designed to stimulate a second purchase for each customer above. Each audience has different needs, requiring different tactics.

Customer Development Index

When you begin the Customer Development process, you chart how you are performing.

This brand could leverage website personalization, email personalization, notifications, print marketing, discounts and promotions, and a variety of other tactics.

The key to Customer Development is to have unique tactics for unique audiences. If every customer gets 40% off, you aren't Developing Customers, you are managing campaigns.

If you offer some customers a discount and not other customers, you are implementing a Customer Development tactic. We can debate whether this is right/wrong. But you are Developing Customers.

In the image above, I color cells blue if there isn't a personalized Customer Development strategy for the intersection between a customer audience and a marketing tactic. If the brand executes specific email campaigns as part of a Welcome Program, the intersection of Welcome (row) and email (column) are yellow.

There are five rows above, and six columns, yielding thirty (30) cells. In the image above, three cells are colored yellow. The Customer Development Index (CDI) is therefore 3/30 = 0.100.

In other words, you aren't doing much to foster Customer Development.

Customer Development + QuickScores

My project work really changed over the past six months. Heck, the running of your business really changed over the past eleven months.

So many of my e-commerce clients generated a ton of new customers. They wanted to know how those customers were going to Develop over time. Customer Development (click here). Who would become loyal? Which first-time buyers had the potential to emerge into good customers? Are there attributes that lead to better customers? How does one act against the information?

Current projects are fusions of Customer Development and QuickScores.

• Determine if new customers truly have a three-month window to convert before decaying.
• Identify attributes associated with first-time buyers that lead to a second purchase.
• Associate merchandise and channels that combine with 2x-4x buyers to lead to additional purchases.
• Quantify how many customers become loyal.
• Outline the attributes of a loyal buyer.
• Tell the story of how a Customer Develops.
• Simulate sales gains associated with better Customer Development.
• Generate QuickScores to enable action against what was learned.
• Brand QuickScore.
• Welcome Program QuickScore.
• Email Marketing QuickScore.
• Print/Catalog QuickScore (where applicable).
• Merchandise Category Preference QuickScores.
• Marketing Channel Preference QuickScores.

This is the most popular project you've requested over the past six months. You get the intelligence of understanding how your Customers Develop, and you get QuickScores to allow you to act upon the information.

In 1-2 weeks, I will release a booklet on the subject.'

You can click here for file formats and project costs.

This week, we'll discuss using the information learned in a Customer Development project.

Customer Development Differs Significantly

Here is a brief summary by month of first purchase. Look in particular at the Emerging Customer row and the Loyal Customer row at the bottom of the table.

This company has no problem developing customers in July, August, and September, with 4.5% / 6.7% / 5.8% of the newbies in those months becoming loyal.

Now look at new customers acquired in October / November / December ... 1.2% / 0.9% / 0.9% of those newbies become loyal within a year.

Well, we know what the homework assignment is going to be, don't we?

And in terms of future value (demand per inventory on the slide above), Jul/Aug/Sep newbies average about $26.50 whereas Oct/Nov/Dec newbies average about $15.75.

This pattern repeats across my client base, but not to the level you see here when there isn't a credible Customer Development program in place. Acquire a customer prior to the peak purchasing season, and then those newbies Develop much faster than the average customer. It's simple math, and almost none of your peers bother to take advantage of basic math.

So go take advantage of basic math. Create a Customer Resume, and get busy Developing customers!!

The Customer Resume - More!

Here's an image from yesterday's Customer Resume of first-time buyers purchasing in December 2019.

In my work, I have nineteen channel-based variables, twenty-nine merchandise-category variables, and ten additional "attribute" variables. In your project, instead of it saying "Channel 01" it would say "Email" or "Social" ... instead of it saying "Merch 15" it would say "Widgets".

I like to look at "what" a customer buys in the future, and I like to look at "where" a customer purchases in the future.

In this case, customers distribute future volume across Categories 2/11/16. These categories are the "gravity" of the brand, they are the categories that customers generally prefer. The majority of your marketing efforts should focus on those products, of course.

Here's the Customer Resume for first-time buyers acquired in September.

December newbies spent 18% of their future dollars on Category 2, 18% on Category 11, and 11% on Category 16.

September newbies spent 16% of their future dollars on Category 2, 14% on Category 11, and 26% on Category 16.

That's a big difference ... look at Category 16. My goodness. September newbies have a fundamentally different product preference than September newbies had. Your Customer Development plan has to account for this factor. When you try hard to convert the customer to a second purchase in months 1/2/3, you are going to personalize the website to take this into account, and you are going to personalize email campaigns to this audience to capitalize on this effect.

By the way, did you look closely at the differences on the left side of the Customer Resume?

December Newbies Rebuy Rate = 19.7%

December Future 12-Month Spend = $14.69 per customer.

September Newbies Rebuy Rate = 26.9%

September Future 12-Month Spend = $26.00 per customer.

It's going to be a lot harder to Develop the December newbie, isn't it?

In terms of Loyal status after twelve months?

December Newbies = 0.9% become loyal within a year.

September Newbies = 5.8% become loyal within a year.

It is CLEARLY OBVIOUS that this company does not have a plan for Developing December newbies.

As a marketer/analyst, it is your job to figure out how to move these (December) customers along. The customers won't likely ever become as loyal as September newbies, but even moving 2.0% to 2.5% to loyal status within a year is not an unreasonable goal.

Use the Customer Resume to identify where your weaknesses are. This brand has a major December newbie weakness that must be addressed.  And then do something about it. Doing the same thing for every customer (i.e. mass email campaigns, Friends & Family postcards, same homepage presentation) doesn't cut it. "Develop" your customers.

The Actual Customer Resume

Ok, here's the table from yesterday, and below it isthe actual customer resume.

Let's look down the left side of the bottom table. You started with 85,476 customers in the cohort. Their rebuy rate was 19.7% (as mentioned yesterday). If they bought, they purchased 1.511 times, buying 3.947 items per order at an average price per item purchased of $12.54.

"Demand per Inventory" is a critical metric ... it's the amount of sales these customers generate in the year after the customer was acquired.

80% of what they bought was existing merchandise.

34% of what they bought were discounted items (those selling below the average historical price for the specific item).

These metrics, on their own, don't mean a whole lot. You have to compare the metrics to other cohorts to understand if they metrics are good/bad.

Tomorrow, we'll address the right-hand side of this table.

The Customer Resume

Part of the Customer Development process is "understanding" how customers behave.

Think about it this way. When you want to hire somebody, you have limited data about the candidate. You scour social media for clues, and the candidate prepares a resume for you. Based on the resume, you make a decision whether to invest in the candidate or not.

Customer Development is no different. Cohorts of customers prepare their own resume. When you acquire a cohort of customers, the customers evolve and change over time. This evolution/change is described via the Customer Resume.

Let's look at customers who bought for the first time in December 2019. We follow the customers for a full year. First, we review how the customers evolved from a Recency/Frequency standpoint.

I like to first look at the bottom row of the table. Notice that 80.3% of these customers didn't buy again in the next year. Oh oh. That's a problem. It's easy to acquire a customer in December, it's hard to Develop customers acquired in December.

How many customers achieve Loyal status (5+ life-to-date purchases)? A whopping 0.9%. Heck, only 18.8% of the customers achieve Emerging status.

How many of the customers were active in the past three months? Just 9.6% (sum the totals for recency = 1/2/3).

These customers likely DON'T WANT TO BE DEVELOPED. They had a Christmas need last year and the brand met that need, now the customer is gone. Well, that's where you have to fight against that trend. That's where the Welcome Program comes into play. You have to convert that customer in Jan/Feb/Mar before you end up with this outcome. I mean, virtually nobody acquired in this cohort is buying anything. And you aren't going to convince a customer who hasn't purchased in thirteen months to buy now. Be honest - name the third party you bought from on Amazon thirteen months ago?

Tomorrow we'll dig into this table ... the actual Customer Resume.

Other Tactics Work

Last week I showed you an example of a promotion that didn't pay off in the short-term, paid off better in the long-term, and overall didn't pay off.

Promotions are one way of stimulating the Customer Development process.

Personalization is another way of stimulating the Customer Development process. In the image below, we have one segment, with a 20% gain in response due to personalization. This has minimal impact on the development of the customer, because in the example below we only add four (4) customers to a second purchase.

But the relationship is profitable!

Now go apply this strategy to ALL SEGMENTS ... apply it to ALL PROSPECTS who haven't purchased (i.e. feature the merchandise that prospects love to buy), and apply the concepts ALL OF THE TIME.

You'll significantly grow your file over time.

Profitably.

Did The Promo Pay Off?

Yesterday we talked about giving a segment of customers (Recency = 3, Frequency = 1) a 30% off promotion. 

The key is to get enough customers to respond that even if you lose money on the promotion you make up profit downstream because you did a good job of developing your customer file.

Let's see if our example pays off.

Click on the image to see the details.

The details? You lost a fortune on the promotion, giving money away to all responders (not just the incremental 50% gain in response). However, you generated 32 responses (per thousand) instead of 21 (per thousand). Those customers now have Recency = 1, Frequency = 2) and are much more responsive. The table clearly shows that you generate a lot more profit downstream from 32 respondents than you generate from 21 respondents.

Look at the bottom row.

When you add up the short-term disaster of the promotion and the long-term gain of developing your customer file, you end up with less profit.

In other words, in this example (your mileage will vary ... but you're probably doing this type of math to understand if your customer development methods work ... right?) the discounting strategy is a failure.

But not all strategies are failures. More on that next time.

Is It Worth It To Do Something?

Oh yes, it is worth it!

Here's our dataset from yesterday.

Let's pretend that you let the customer lapse to a recency of three months. Look at the table. What is the probability of the customer purchasing in the next year?

It's 16.9%.

Now, look down the "Freq = 2" column. If you convert the customer to a second purchase, what is the probability of the customer purchasing again in the next year?

It's 43.0%.

Does that look like a fair trade / upgrade?

Maybe!

We need to do some math, don't we?

Ok, let's pretend that you decide to become promotional. You decide to give these customers 30% off. For a month, you get a 50% increase in response. That's a big deal. But you have to give 30% away to make the magic happen.

Tomorrow, we'll see if the magic pays off.

Fighting an Uphill Battle

Based on recency / frequency metrics for customers, I measured the probability of the customer purchasing again in the next year. The data is outlined below. Take a look.

Click on the image to blow it up a bit ... it makes it easier to read the content.

Every month a customer fails to purchase is a month where the customer "retires" just a bit more. Read down the "Freq = 1" column. These are first-time buyers. Each row represents the number of months since the last purchase.

If the customer just purchased for the first time, the customer has a 24.4% chance of buying again in the next year (in our example). In other words, three out of every four first-time buyers won't purchase again in the next year. This company has a huge problem. You want to see this metric be in the mid-30% range for a company to have a fighting chance.

Now let's say you don't have a Welcome Program in place, you just have your standard barrage of marketing programs and discounts and what not. If the customer does not purchase in the first month, the customer drops down a row, to Recency = 2 (two months since the last purchase).

At this point, the customer has an 18.6% chance of buying again in the next year. The customer is starting to fade ... already ... after buying for the first time just two months ago.

Let' assume the customer doesn't buy again, and lapses into a third month.

At this point, the customer has a 16.9% chance of buying again in the next year.

This is why Customer Development is so darn important. You have all of these COVID buyers that were blessed upon you while we close in on a half-million deaths from the disease. And within a few months, those buyers begin to lapse. Each month that passes makes your job even harder!

Get busy!

Craft a program to convert your first-time buyers quickly. Don't let them fade into oblivion. Do something!! You've got the smarts to do great things. Don't start fighting an uphill battle, ok?

Slides for You

Click here and download fifty (50) slides outlining an "Introduction to Customer Development".

The slides outlined the thought process and targeting opportunities to develop customers ... the slides also illustrate just how hard it is to move a customer from a first purchase to loyal status. You already know that fact, but it might help to have slides to reinforce your arguments.

A comparable post on LinkedIn (click here) garnered more than 1,300 views, so clearly somebody cares about the topic (i.e. your peers).

As The Year Progresses, The Customer SLOWLY Emerges

We continue our study from yesterday. Look at the series of tables below, which show how our cohort of new buyers evolves and changes over time - through the end of the first year the customer was on the file.

Follow the customers through the year, what do you see?

Look at rebuy rates as they accumulate through the year.

• 18.8% through November.
• 20.0% through December.
• 21.3% through January.
• 22.1% through February.
• 22.6% through March.
• 23.2% through April.
• 24.3% through May.
• 25.1% through June.

As discussed yesterday, much of the "action" in getting customers to a second purchase happens in the first three months after a first purchase (10% converted after three months in our example).

How many customers migrated to "Loyal" status (5 purchases)?

• 0.9% through November.
• 1.1% through December.
• 1.4% through January.
• 1.6% through February.
• 1.7% through March.
• 1.8% through April.
• 1.9% through May.
• 2.0% through June.

So here's the thing. You are constantly told to focus on your Loyal buyers, and who can blame you, that's where all of the fun is in your marketing efforts? But look at this example. Just look at it. After a year, only two (2) percent of the customers acquired become loyal (in my example, "Loyal" = five purchases).

And only 2.3% of the customers acquired last June had a recency of one month as of the following June. Heck, only 5.2% of those customers purchased in the past three months. That's a really, really tiny fraction of "active" customers.

Virtually nobody is loyal.

Virtually nobody is active.

THIS is why Customer Development is so important. The company I am citing here does a modest job of developing customers ... not great, but certainly not awful like we see across the board these days.

Focus on those first three months after a first purchase, and accelerate the customer as s/he migrates from a 2nd to 3rd to 4th to 5th purchase, where Loyal status is frequently achieved.