Digital Profiles: Q&A Time

Ok, you've learned an awful lot about Digital Profiles. I can tell that you're excited!

Now it is time for your questions. What made sense, what didn't make sense? Any thoughts on how you might use the methodology in your marketing strategy? Catalogers --- this stuff works REALLY WELL when trying to figure out which customers to no longer mail catalogs to!!!!!

So, use the comments section to ask your own questions about Digital Profiles.

Or contact me for a Digital Profile project!

Digital Profiles: E-Mail Targeting And Predicting Demand

Digital Profiles aren't some geeky, theoretical methodology. They are a way to better understand your business, a way to predict future demand.

Take this example. I took our example, and created a simple regression model (ignore the fact that many variables aren't significant, I'm doing this to prove a point) that predicts future e-mail demand, based on last year's demand and last year's Digital Profile.

There are several Digital Profiles that are big predictors of future e-mail demand. Digital Dudes (men who shop on the internet) and Happy Holidays (customers who purchase in Nov/Dec) are the "most valuable" to the e-mail marketing program.

Now, if you know that Digital Dudes are the most valuable Digital Profile, might you create a version of your e-mail campaigns specifically for this Digital Profile?

And if you know that Summer Ladies are an unproductive Digital Profile, might you sometimes elect to not send them an e-mail campaign, just to test how that customer would perform without e-mail marketing ... or might you send that customer e-mail campaigns during May/June/July, when that customer is most likely to shop?

Use Digital Profiles to "amp" your e-mail marketing program. Get to know your customers, and then use the data to make actionable marketing decisions that improve sales and profit!

Digital Profiles: Web Analytics And Keywords

You have your Digital Profiles all set up. And you are able to tie the information to your Web Analytics solution.

Ohhhhhhhh Boy!

Each week, you update your Web Analytics solution with the Digital Profile of each customer who logs in to your site.

Now go ahead and analyze performance by Digital Profile. For instance, this table illustrates conversion rates by Digital Profile for customers who are searching for any keyword within the "Handbags" category.

Not surprisingly, the Digital Profiles attracting the most customers include Summer Ladies, Show Me A Deal, Bargain Betty, Web Sale Susan, and Bonnie Big Ticket. Conversely, you don't see a lot of "Multichannel Men" searching for Handbags, do you?!!!

Predictably, conversion rates are highest among Digital Profiles that appear to be populated by Women.

This is a level of business intelligence that is available to any analyst using any of the major Web Analytics software packages. It does require a bit of "elbow grease" ... you have to be able to score your web visitors at a point in time using advanced logic.

Ok, your turn. What stops you from using Digital Profiles to learn more about various customer habits?

Digital Profiles: New Customers

Any analysis of Digital Profiles should include a distribution of new customers by Digital Profile.

Remember, our best customers were Multichannel Men, Digital Dudes, Winter Wear, and Happy Holidays.

Now take a look at the Digital Profiles that are capturing the most new customers.

• Last Minute Guy
• Buy A Jacket, Man!
• Tech Men
• Show Me A Deal!
• Bonnie Big Ticket

In other words, this business is clearly in transition. The high-value Digital Profiles are not growing, while a different set of Digital Profiles, those that possess infrequent buyers, are taking over. Now, granted, this is part of what online marketing or catalog marketing is all about --- you don't acquire the best customers, you acquire customers with some migrating to "best" status.

Still, as a business leader, you want to know where your business is headed. Digital Profiles explain these trends in a way that lay people can understand.

Digital Profiles: Online Marketing Simulations

I really enjoy using Digital Profiles in my Online Marketing Simulation algorithm!

Here's a five year simulation, illustrating customer counts by Digital Profile.

Remember, there are a handful of Digital Profiles that exhibit high performance, including Multichannel Men, Digital Dudes, Winter Wear, and Happy Holidays.

When we run all of our Digital Profiles through the simulation, we see trends that, in this case, are not necessarily favorable.

Multichannel Men are projected to decline from 5,075 to 2,880 over time.

Digital Dudes are hanging in there, from 5,104 to 4,843.

Winter Wear show a significant drop, from 5,262 to 4,289.

Happy Holidays, however, are on the increase, from 5,089 to 6,116.

Clearly, there is a trend away from Multichannel Men, and a growth area among the Happy Holidays segment. But overall, the best segments are losing customers, and that is never a good thing.

Which segments are increasing over time?

• Happy Holidays
• Tech Men
• Last Minute Guy

You see a trend away from good customers, and you see increases in online Digital Profiles and Digital Profiles that represent customers shopping for a need during the Holiday Season.

In other words, Digital Profiles provide a way for the business leader to understand the psychographic trends that underlie the business.

Digital Profiles: Performance Analysis

Digital Profiles have value if and only if they do a credible job of describing customer behavior.

Here is a case where our Digital Profiles make a difference. We're analyzing customer behavior during January 2010, based on the Digital Profile the customer had as of the end of December 2009.

Which segments perform the best?

Winter Wear is the segment that performed the best, and for good reason, given that January is smack-dab in the middle of winter!

Notice that Multichannel Men perform well, too.

The bottom of the Digital Profile list didn't perform well. These tend to be sale customers who like to shop during the Holiday season.

How about by channel? Look at phone orders:

• Winter Wear
• Multichannel Men
• Happy Halloween

How about online orders?

• Digital Dudes
• Happy Holidays (these must be good, cross-channel buyers).
• Winter Wear

And then there's e-mail:

• Happy Holidays
• Digital Dudes
• Multichannel Mavens

Last, we look at search:

• Multichannel Men
• Happy Holidays
• Digital Dudes

We repeatedly see the same Digital Profiles performing well in January. This will happen, over and over and over again. There will be maybe four Digital Profiles that consistently outperform all others. These are your "best" customers, and the labels associated with the Digital Profiles tell you who your best customers are.

For this brand, the Digital Profiles that perform well are:

• Winter Wear
• Multichannel Men
• Happy Holidays
• Digital Dudes

These are the Digital Profiles to pay attention to. If the business is going to be successful, then these Digital Profiles need to grow, not contract.

We will look at this issue in our next edition of Digital Profiles, next week.

Digital Profiles: Naming Conventions

Ok, you've come up with sixteen Digital Profiles. Now you need to name them.

This might be the hardest part. The names should be catchy, and should directionally describe what the customers in that Digital Profile have done, or are likely to do.

For instance, here are sixteen names, one for each segment in an analysis:

• Multichannel Men (Men who buy often, and buy via catalogs, e-mail, search, online, etc.)
• Digital Dudes (Men who love to shop online).
• Summer Ladies (Women who purchase merchandise in May/June/July).
• Show Me A Deal! (Women who buy lots of merchandise on sale).
• Winter Wear (Customers who buy in the winter).
• Happy Holidays (Customers who limit shopping to the Holiday Season).
• Autumn Online (Women who prefer the Fall assortment, in September or October).
• J. Peterman (A male shopper with taste and attention to detail).
• Tech Men (Think of a male shopper in the Bay Area, one who hates seeing trees cut down and converted into advertising catalogs).
• Bargain Betty (A rural female shopper who buys inexpensive items and sale items).
• Web Sale Susan (A suburban female shopper looking for the best deal online).
• Buy A Jacket, Man! (A male shopper who only looks for outerwear).
• Last Minute Guy (A male shopper purchasing gifts with expedited shipping).
• Bonnie Big Ticket (A female shopper with a discerning taste for expensive items).
• Gift For My Girl (A mom shopping for her daughter).

You simply analyze all of the variables that comprise each Digital Persona, looking for trends that illustrate what this customer is all about.

Up Next: A sample analysis or two!

Digital Profiles: Creating Each Profile

You have your data in a "spreadsheet", if you will, one row per customer, each column telling us something about that customer during the past year.

Now, it is time to generate each Digital Profile.

You are free to use whatever methodology you wish to use. I personally adore a methodology known as a "Factor Analysis", because the methodology is elegant, reducing the dimensionality of a complex dataset to a series of "factors".

Here's what I do:

1. I calculate the mean and standard deviation of each variable, for later use.
2. I run a Factor Analysis.
3. I extract three or four "Factors".
4. I run a frequency distribution, to determine the median value for each Factor.
5. I re-code each Factor, 0 = below 50th percentile, 1 = above 50th percentile.
6. I score the customer file at multiple points in time, so that I know the "Digital Profile" of the customer at many different times. This information is saved for analysis purposes.

Personally, like using sixteen Digital Profiles (four factors split into two groups each).

Enough for today. Next week, we'll dive into naming strategies, analysis and reporting, and Multichannel Forensics / Online Marketing Simulation examples.

Digital Profiles: What Data Do You Need?

For the next few weeks, we're going to talk about what I call "Digital Profiles". These are segments of customers that exhibit similar behavior. You might have 'Golden Girls', a segment of customers who are likely age 60+ purchasing via traditional channels. Or maybe you have "Robert Scobles", technology fans who eschew old-school marketing tactics.

Now how the heck do you create "Digital Profiles"? I mean, you don't have the lifestyle/psychographic/demographic data you need to do this the right way, correct?

Well, let's keep things as simple as possible.

Let's take all purchases that happened in the past twelve months. Among these orders, you know several things.

1. The method the customer used to pay for merchandise (Cash, Check, Visa, MasterCard, Amex, Gift Card, etc.) Hint --- cash/check are proxies for a 60+ year old customer.
2. The day the customer purchased merchandise (Sunday - Saturday ... yes, this matters).
3. The time of year the purchase happened (Valentines Day vs. July 4 vs. Cyber Monday ... you get the picture).
4. The merchandise divisions the customer purchased from (think about the tabs running across the top of your website ... yes, this matters, too!).
5. Average number of items per order.
6. Average price per item (hint, this is important). Items 5 and 6 yield AOV.
7. Total number of annual orders.
8. Sale customer.
9. Promotional customer (free shipping, % off).
10. Zip Code Forensics (Urban, Suburban, Rural customer).
11. Physical Channel the customer purchased from (Phone, Web, Store).
12. Advertising Channel that influenced the order (Catalog, Paid Search, E-Mail, Affiliates, Social, Mobile).
13. Did customer return more than 40% of merchandise purchased?
14. Does customer pay for expedited shipping?

This is not a finite list, use your imagination.

Create a "spreadsheet", with one row per customer ... the fourteen characteristics mentioned above are columns in the spreadsheet.

We'll stop here. If you can collect this kind of information, you have a fighting chance to create interesting Digital Profiles.