Monday Musings

Sports might be on hiatus (and for good reason) but marble racing is not (click here).

Orangutans and Otters ... we always knew they'd get along but then there's this (click here).

Washington Area Man faces charges after a high-speed chase ends with revelation that his dog was in the driver's seat and may have been driving the vehicle at times during the chase (click here).

Goodnight with Dolly (click here) ... every Thursday night.

You think you've got it bad? This is from 1991 (click here).

Apparel Catalogers ... go scour Etsy (click here) and then figure out how to push a small catalog out the door with face masks.

One-Page Summary Of What You Need To Provide For A QuickScore Project

I'm getting A LOT of inquiries about performing a QuickScore analysis ... so I thought I'd summarize the requirements for a project so you know exactly what you need to provide. When you contact me about starting a project, you'll already know what to do and you can just tell me to "go" and get busy.

Again, I will provide QuickScores at the ridiculously low cost of $8,000 for the duration of our COVID-19 stay-at-home situation ... and likely well into next year if issues continue. You deserve to get something at a very low cost right now, don't you?

And if you need the standard Contact Strategy project (if you are a catalog brand), read the bottom of this page for details. In the standard Contact Strategy project, you will learn exactly how many times you should mail a customer on an annual basis to optimize company profit.

Why QuickScores?  Your feedback suggests you will need to cut back on marketing expenses, and you need a tool to quickly and easily manage email marketing (targeting best email subscribers with the merchandise they love - personalized at a customer level) RIGHT NOW!!  Catalogers are telling me they want a simple score they can apply to their database to reduce catalog mailings to save money while better leveraging email marketing to target customers properly at a low cost, helping make up sales volume lost via print reductions.

QuickScores Project Outline

File #1 (Required): One row per item purchased, 5 years of purchase history, .csv file format required.

• Household_ID
• Order Date (20200402).
• Merchandise Category.
• Item Number.
• Quantity Purchased.
• Amount Spent.
• Physical Channel Purchased From (Mail, Phone, Online, Retail).
• Marketing Channel Purchased From (Paid Search, Email Marketing, Catalogs, Affiliates, Television, Radio etc).

 File #2 (Optional): One row per household, .csv file format required.

• Household_ID
• Email Subscription Date (20171118).

 File #3 (Optional):  One row per household, .csv file format required.

• Household_ID
• Date Customer Clicked-Through an Email Campaign (20190614).

File #4 (Optional / Catalog Brands Only):  One row per household, .csv file format required.

• Household_ID
• Number of catalogs mailed, past 12 months.
• Total cost of mailings at a customer level. 

Profit Data:  Percentage of demand that flows-through to profit (i.e. 42%).

What do you receive?

• Brand QuickScore:  Each customer is scored based on overall brand responsiveness, 0 to 100 (100 = best).
• Catalog QuickScore:  Each customer is scored based on catalog responsiveness, 0 to 100 (100 = best).
• Email QuickScore:  Each customer is scored based on likelihood of clicking through an email campaign (or buying via an email campaign if email clicks are not available), 0 to 100 (100 = best)..
• Primary Merchandise Category QuickScore:  Based on a weighting scheme, I tell you the merchandise category the customer prefers most, for email targeting purposes. You personalize email merchandise strategy via a combination of the Email QuickScore and Primary Merchandise QuickScore.
• Secondary Merchandise Category Quickscore:  Based on a weighting scheme, I tell you the merchandise category the customer prefers second-most, for email targeting purposes. You personalize email merchandise strategy via a combination of the Email QuickScore and Secondary Merchandise QuickScore.

 Project Cost:

• $8,000, held firm during the COVID-19 outbreak, due within 15 days of project completion.

 Project Length:

• Once data is received, the analysis should take 5-7 business days.

 Pseudo Code:

• A document will outline the pseudo code required to implement the scores in your customer data warehouse.
• You will also be sent the QuickScores, at a customer level, based on the most recent date in your database, so that you can act immediately upon the results.

 Project Add-On:

• If you are a catalog brand and you wish to know exactly how many catalogs to mail to customers on an annual basis to optimize annual profit, you can add the Contact Strategy component to your project.
• Annual Sales up to $30,000,000 = $8,000 QuickScores fee plus $9,000.
• Annual Sales $30,000,000 to $99,999,999 = $8,000 QuickScores fee plus $22,000.
• Annual Sales $100,000,000 to $999,999,999 = $8,000 QuickScores fee plus $32,000.
• Annual Sales $1,000,000,000 and above = $8,000 QuickScores fee plus $47,000.

Thanks,

Kevin

Oh, It Matters!

Take a look at the graph below. The blue line represents the probability of a customer buying next month by recency / given the fact the customer clicked through at least one email campaign in the past month. The green line represents the probability of a customer buying next month by recency / given the fact that the customer did not click through any email campaigns in the past month.

Which customer would you rather bet your career on? The customer who clicked through at least one email campaign last month, correct? The difference in response is staggering ... at all levels of recency. This one (1) attribute equalizes customers ... a customer with 40 months of recency becomes like a customer with 10 months of recency once the customer clicks through an email campaign. You'd want to know that, right? Oh, it matters!!

Append a QuickScore (click here) to your customer file and TAKE ACTION!!

Email Clicks Impact Other Channels, Too

Yesterday I described the importance of overlaying a QuickScore (click here) against lapsed customers ... culling out the customers who clicked through an email campaign last month.

Recall that response in the next month increased from 1.20% to 3.52% based on this simple overlay. You'd want to know that, right? And if you knew it, you'd act upon it, right?

Each channel is impacted differently by this fact.

• Call Center = 0.26% response to 0.32% response after customer clicks through an email campaign.
• Online Orders Matched To Catalogs = 0.07% response to 0.16% response.
• Email Purchasing = 0.10% response to 2.02% response (oh).
• Search Purchasing = 0.33% response to 0.36% response.
• All Other Online Marketing = 0.03% response to 0.12% response.
• All Other Online Orders = 0.25% response to 0.48% response.
• Retail In-Store Orders = 0.24% to 0.28%.
• Non Email Channel = 1.10% response to 1.50% response.

The email click results in a 35% increase (1.50 vs. 1.10) in response in channels outside of email marketing.

You'd want to know that, right?

Click here for details regarding Hillstrom's QuickScores.

P.S.: This is one way of demonstrating that "engagement" is important. Draw a link between engagement metrics and actual purchases.

Leveraging QuickScores to Trigger Actions

Ok, this one came up in a recent project.

Assume the customer has Recency = 15 Months .. .the customer hasn't bought anything in a long time. Assume historical spend is between $175 and $225 ... equalizing historical spend.

Now we segment the customer based on very simple segmentation variable.

• Yes = Customer clicked through at least one email campaign in the past month.
• No = Customer did not click through one email campaign in the past month.

Simple enough, right? You've got this data in your omnichannel customer data warehouse, right?

Now measure rebuy rates by channel in the next month, as well as overall. Here's an example from a recent analysis, where we look at overall rebuy rates in the next month.

• 15 Months Recency, $175 - $225, Click = Yes:  Next Month Rebuy = 3.52%.
• 15 Months Recency, $175 - $225, Click = No:   Next Month Rebuy = 1.20%.

Oh ... look at that!

All things being equal, the customer who clicked through an email campaign last week was almost 3x as responsive as the customer who didn't, after controlling for recency & historical spend.

Seems like you'd want to know that.

Seems like you'd act upon that if you knew that, right?

Seems like if you were a catalog brand you'd kick out a hotline catalog if the catalog was not previously mailed, right?

You can perform a simple overlay to get this information. You can use Hillstrom's QuickScores (click here) to get this information. You can use your own in-house models to obtain this information. What is most important, of course, is that you do something with the information.

Monday Musings

Life is hard enough ... so on Monday we pursue a different vector:

Grocer H-E-B performed simulations prior to the outbreak, leaving them better prepared for everything that has happened since (click here). Enjoy the oral history style of storytelling.

Dinner with the Gaffigans (click here).

Medical masks are one of Animal Crossing's hottest looks (click here).

Kringle from O&H Danish Bakery in Racine ... perfect for our trying times (click here).

Stay tuned until at least the twenty-nine second mark of the video (click here).

Christmas Clicks

I recently built a QuickScore model (click here) ... in the email portion of the model I predicted who would be most likely to click through an email campaign in the next month.

That's where the fun started:

• Customers who clicked through email campaigns between Thanksgiving and Christmas were less likely to click through a campaign in Spring ... the variable possessed a negative coefficient.

Oh.

There is so darn much good stuff to research in email click-through activity ... and almost all of it has been ignored by an industry lusting over vanity metrics. The secrets to File Power are just lying there, waiting to be discovered. Go discover how your customers actually behave as long as you are trapped at home.