September 30, 2010

Dear Marketers: Faux Metrics

Dear Marketers:

Ever heard a quote that sounds something like this?

"Facebook users were 2.7 times more likely to click on calls-to-action in targeted and personalized e-mail campaigns, yielding users who were 1.6 times more engaged than the average e-mail subscriber."

The online realm is increasingly polluted with "faux metrics". Faux metrics, for your information, are metrics that sound impressive and convincing, are wrapped in beautiful language, but have minimal correlation with what matters most ... profit.

"Today's thrifty consumer will spend one percent less this Holiday season, requiring forward-thinking marketers to capitalize on a seventy-six billion dollar opportunity by offering value, selection, and deep discounts."

One wonders why a one percent drop in spend must be fought with value, selection, and deep discounts?

Faux metrics sound great, but lack an actionable outcome.

"Sixty-three percent of iPad users were male, and they were eight percent more likely to purchase after seeing an ad, creating a unique opportunity for the savvy advertiser."

Oh. In other words, if 2% of the population owns an iPad, and those customers are 8% more likely to purchase something after seeing an ad, then my $50,000,000 business is likely to grow by 0.02*0.08 = 0.0016 = 0.0016*50,000,000 = $80,000, but only if I act on this unique opportunity.

"Engagement" is probably the biggest faux metric in our industry. Engagement is seductive, because it implies that the customer is so in love with our business that the customer must want to transact with us. We suspend all other facts as we pursue the truthiness of engagement.

"E-Mail subscribers, in this highly competitive retail environment, were 9% more engaged with retail brands that offered personalized, trigger-based messages."

"Those who clicked on a bit.ly url on Twitter were thirty-seven percent more likely to respond to landing page calls-to-action, yielding engaged users that can be monetized by forward-thinking digital marketers."

The most interesting metric I track is "repurchase rate". I simply count the number of customers who purchased in 2008, then I track the percentage of 2008 customers who purchase again in 2009. I record the values of this metric over time. The metric is highly correlated with total company profit. The metric is not correlated with "engagement", or a myriad of faux-metrics promoted by the online digerati.

In fact, "repurchase rate" hasn't moved much at all in the past decade ... small increases in the 2003-2006 time frame are more than offset by large drops in 2007-2009. For all of the digital magic thrust upon us by the online punditocracy, customers are not purchasing any more today than they purchased back in 2000, or 1995, or 1990. We offer the customer more and more and more and more ... and the customer still repurchases at a 38% rate, generating 1.6 orders per year.

But to read the "faux metrics", you'd think that today's highly-engaged social shopper is spending 294% more than the moribund customer of days gone by.

Your turn: What are examples of faux metrics that you recently uncovered?

September 29, 2010

Digital Profiles Part J: Personalization And Marketing Plans

Let's say that you are the Merchandise Executive for Merchandise Division #8. It is your job to grow sales within your Merchandise Division, period. Nobody is going to give you credit for creating high-value, multi-channel, multi-merchandise customers, are they?

When we look at all sixteen Digital Profiles, we immediately observe that there are eight Digital Profiles that are pre-disposed to your Merchandise Division.
  • Gold Mine!
  • Multi-Channel Mavens.
  • Shop Online, Buy In Store.
  • Direct Newbies Looking For Eight.
  • Retail Fanatics.
  • Retail 4/5/6.
  • Retail 1/2/3.
  • Retail Newbies Looking For Eight.
So here's a fun analysis to do. In the 30 days after we determine the Digital Profile that a customer belongs to, I calculated the percentage of customers in each Digital Profile who purchased from Merchandise Division #8 online (website, e-mail, search, mobile, social).
  • Gold Mine! = 1.15% purchase rate.
  • Multi-Channel Mavens = 1.06% purchase rate.
  • The Digerati = 0.46% purchase rate.
  • A Long Drive = 0.33% purchase rate.
  • Shop Online, Buy In Store = 0.72% purchase rate.
  • Direct Newbies Looking For Eight = 0.51% purchase rate.
  • Online One/Twos = 0.30% purchase rate.
  • Ship It To Me = 0.22% purchase rate.
  • Retail Fanatics = 0.15% purchase rate.
  • Retail 4/5/6 = 0.05% purchase rate.
  • Retail 1/2/4 = 0.09% purchase rate.
  • In And Out = 0.05% purchase rate.
  • Retail 1/2/3 = 0.07% purchase rate.
  • Retail Newbies Looking For Eight = 0.05% purchase rate.
  • Retail 1/2 = 0.01% purchase rate.
  • Rotary Phone = 0.12% purchase rate.
Well, clearly, there are a handful of Digital Profiles that you'd target online if a visitor arrived at your website, correct? Heck, even "The Digerati" and "A Long Drive", two profiles that weren't really fond of this merchandise, they are more productive because of their online disposition than are any of the retail segments.

Now, let's turn the tables here. What are the purchase rates at retail during this timeframe?
  • Gold Mine! = 4.06% purchase rate.
  • Multi-Channel Mavens = 1.75% purchase rate.
  • The Digerati = 0.29% purchase rate.
  • A Long Drive = 0.32% purchase rate.
  • Shop Online, Buy In Store = 1.85% purchase rate.
  • Direct Newbies Looking For Eight = 0.33% purchase rate.
  • Online One/Twos = 0.16% purchase rate.
  • Ship It To Me = 0.12% purchase rate.
  • Retail Fanatics = 3.38% purchase rate.
  • Retail 4/5/6 = 1.80% purchase rate.
  • Retail 1/2/4 = 0.88% purchase rate.
  • In And Out = 0.71% purchase rate.
  • Retail 1/2/3 = 1.54% purchase rate.
  • Retail Newbies Looking For Eight = 0.98% purchase rate.
  • Retail 1/2 = 0.50% purchase rate.
  • Rotary Phone = 0.36% purchase rate.
Can you see why those "Gold Mine!" customers are who they are? They buy and buy and buy, regardless of channel! Also, pay attention to "Shop Online, Buy In Store" ... a high rate within retail, a not-so-good purchase rate online ... clearly, these customers are pre-disposed to shop a certain way, so when that customer visits online, don't expect to convert that customer on that visit. Instead, present merchandise to that customer that is congruent with what that customer will buy when the customer visits the store.

Finally, we'll look at total purchase rates (web, stores, phone) during the next thirty days:
  • Gold Mine! = 5.22% purchase rate.
  • Multi-Channel Mavens = 2.94% purchase rate.
  • The Digerati = 0.88% purchase rate.
  • A Long Drive = 0.74% purchase rate.
  • Shop Online, Buy In Store = 2.57% purchase rate.
  • Direct Newbies Looking For Eight = 0.89% purchase rate.
  • Online One/Twos = 0.51% purchase rate.
  • Ship It To Me = 0.38% purchase rate.
  • Retail Fanatics = 3.58% purchase rate.
  • Retail 4/5/6 = 1.89% purchase rate.
  • Retail 1/2/4 = 1.02% purchase rate.
  • In And Out = 0.79% purchase rate.
  • Retail 1/2/3 = 1.62% purchase rate.
  • Retail Newbies Looking For Eight = 1.05% purchase rate.
  • Retail 1/2 = 0.57% purchase rate.
  • Rotary Phone = 0.66% purchase rate.
Clearly, there are a set of Digital Profiles that are responsive to Merchandise Division #8. And within the Digital Profiles that are responsive, several are responsive online, several are responsive in stores, and two ("Gold Mine!" and "Multi-Channel Mavens") that are responsive in both the online and store channel.

So, what should the Executive responsible for Merchandise Division #8 ask of the marketing department?
  • A marketing plan for "Gold Mine!" and "Multi-Channel Mavens" that is channel-agnostic.
  • A marketing plan for "Shop Online, Buy In Store" that is research-based online and is commerce-based in stores.
  • A marketing plan for "The Digerati", "A Long Drive", and "Direct Newbies Looking For Eight" that is focused on online channels, including personalization opportunities across the home page and key landing pages for visitors with these Digital Profiles. E-mail campaigns might be fruitful for this audience as well.
  • A marketing plan for "Retail Fanatics" and "Retail Newbies Looking For Eight" that encourages additional purchases within this merchandise division.
  • A marketing plan for "Retail 4/5/6", "Retail 1/2/4", "In And Out", and "Retail 1/2/3" that seeks to cross the customer into a new merchandise division.
The marketer (that's you) creates a marketing plan that helps the Merchandising Executive achieve her vision of global domination.

You can do this, using Digital Profiles to help guide the process!

Digital Profiles: Reader Contribution

Dave sent us this nice Tableau Public image of customer migration from last year's Digital Profile to next year's Digital Profile.

Take a peek at the nice work that Dave did!

September 28, 2010

Digital Profiles Part I: Mobile Buyers

Within every Digital Profile analysis, there are subsets of customers that are interesting to analyze, from a marketing standpoint.

Take a look at Mobile buyers. Within our Digital Profiles, Mobile buyers are represented by the following Digital Profiles:
  • Gold Mine! = 1,007 customers.
  • Multi-Channel Mavens = 654 customers.
  • The Digerati = 293 customers.
  • A Long Drive = 471 customers.
  • Shop Online, Buy In Store = 417 customers.
  • Direct Newbies Looking For Eight = 323 customers.
  • Online One/Twos = 515 customers.
  • Ship It To Me = 207 customers.
  • Retail Fanatics = 0 customers.
  • Retail 4/5/6 = 0 customers.
  • Retail 1/2/4 = 0 customers.
  • In And Out = 0 customers.
  • Retail 1/2/3 = 0 customers.
  • Retail Newbies Looking For Eight = 0 customers.
  • Retail 1/2 = 0 customers.
  • Rotary Phone = 0 customers.
The bottom eight Digital Profiles do not possess a single Mobile shopper.

Now, you are a marketer, and by most accounts, a darn good marketer, right? So what would your marketing message be to the first eight Digital Profiles, and what would your marketing message be to the final eight Digital Profiles, should you wish to promote your Mobile experience to your customer base?

Here's another interesting way to slice/dice Digital Profile data. Let's look at everybody who shopped via Mobile in the next month. We'll list which Digital Profile the customer came from. Here we go!
  • Gold Mine! = 74 customers.
  • Multi-Channel Mavens = 54 customers.
  • The Digerati = 26 customers.
  • A Long Drive = 34 customers.
  • Shop Online, Buy In Store = 12 customers.
  • Direct Newbies Looking For Eight = 13 customers.
  • Online One/Twos = 28 customers.
  • Ship It To Me = 8 customers.
  • Retail Fanatics = 2 customers.
  • Retail 4/5/6 = 3 customers.
  • Retail 1/2/4 = 2 customers.
  • In And Out = 4 customers.
  • Retail 1/2/3 = 6 customers.
  • Retail Newbies Looking For Eight = 3 customers.
  • Retail 1/2 = 1 customers.
  • Rotary Phone = 1 customers.
  • No Purchase Last Year = 91 customers.
  • Total = 362 customers.
Only 22 of the 362 Mobile customers came from the final eight Digital Profiles.

91 of the 362 Mobile customers came from outside the twelve month file.

249 of the 362 Mobile customers came from the eight Digital Profiles that are "Mobile-Shopper Friendly"!

It seems intuitive to think that Mobile marketing activities should be targeted at the first eight Digital Profiles, and should be touted to non-buyers/prospects as well. The final eight Digital Profiles are not very "Mobile-Responsive", if you will ... so we probably shouldn't focus conversion efforts in this area.

Clearly, some customer types have a pre-disposition to shop in certain ways, while other customers are pre-disposed to shop in other ways. It's pretty darn important to understand customer pre-disposition, if we want to use marketing dollars in an effective manner.

September 27, 2010

Join Me At ECMOD!

Are you in London next week?

If so, join us at ECMOD on October 6-7, a well-done conference in London for marketers who use many different channels to grow a direct marketing business. Or follow ECMOD on Twitter.

I'll give two presentations ... one on modern Matchbacks and Attribution, one on 35 tips for Direct Marketers to grow a direct marketing business tomorrow. We'll share some neat secrets about how search marketing is influenced by e-mail and catalog marketing!

See you next week in London!

Digital Profiles Part H: Migration

We marketers like to believe that we "make" the customer do things. We believe that we cause customers to become "multi-channel", or that we "cross-sell" and "up-sell".

On a small scale, yes, absolutely, it is possible to influence customer behavior.

On a large scale, customer behavior is something that is nearly pre-determined, and is outside of our control.

We can use Digital Profiles to understand how customers evolve and change over time.

Click on the image above to enlarge it. Now read down the "Gold Mine Column". Here, you get to see what happens in 2010 to customers who were "Gold Mine" Digital Profile members in 2009.

As we already know, 83% of these customers purchase again in the next twelve months. Just as interesting is the Digital Profile that the customer will belong to in the next twelve months. 33% of customers will be "Gold Mine", 8% will be "Multi-Channel Mavens", 8% will be "Retail Fanatics", with the remaining 34% distributing across the remaining 13 Digital Profiles. In other words, "Gold Mine" customers are not likely to stay in the "Gold Mine" profile, the move and evolve, but not terribly far from their original destination. Notice that 16% either become "Multi-Channel Mavens" or "Retail Fanatics", Digital Profiles that are congruent with the "Gold Mine" Digital Profile.

Next, take a look at "Ship It To Me", a direct-channel-centric Digital Profile. Only 32% of these customers are likely to purchase in the next twelve months ... with half of those who repurchase migrating to "A Long Drive" or staying in "Ship It To Me". Clearly, these customers are not likely to be encouraged to become "multi-channel" buyers, and are not likely to become strong retail customers. Heck, why even bother with spending energy trying to get these customers to shop retail stores when it is not their natural pre-disposition!

Let's look at one final Digital Profile ... "Retail Fanatics". 67% of these customers are likely to repurchase in the next twelve months. These customers do seem somewhat likely to migrate to other Digital Profiles, including "Retail Fanatics", "Retail 4/5/6", "Retail 1/2/4", "In And Out", and "Retail 1/2/3". This is not a good sign! In other words, the "Retail Fanatic" is not likely to become a "Multi-Channel Maven" or a "Gold Mine" shopper (these are Digital Profiles that exhibit retail and online behavior). Instead, these customers are likely to "drop down" to a lower-value Digital Profile, one that is retail-focused, one that prefers fewer (not more) merchandise divisions. It has been my experience that you do not "force" a customer pre-disposed to stay within one channel to shop other channels (like the web) ... in this case, the majority of customers in this Digital Profile are going to stay in the retail channel.

We use Digital Profiles to understand how customers behave and evolve, we seek to understand how we might better market to customers, given unique behaviors outlined in the "Digital Profile" framework!

September 26, 2010

Dear Catalog CEOs: More Pages

Dear Catalog CEOs:

Have you ever been in a meeting where the head merchant asked you to add pages to an upcoming catalog?

These are painful situations if more than half of your direct channel sales happen online. Painful, because customers aren't shopping in the manner that the merchant thinks the customer shops.

Over the past decade, my "page curves" have become sharper and sharper. In other words, ten years ago, I'd get 75% of the demand on 50% of the catalog pages ... if a test were executed comparing a 48 page catalog to a 96 page catalog, I'd get 75% of the demand in a 48 page catalog that I'd get in a 96 page catalog, at the same quality of circulation.

The economics of that situation dictated larger page counts.

Since 2004 or 2005, the curve increasingly became sharper. Today, I repeatedly see situations where folks get 88% of the demand in a 48 page catalog that they get in a 96 page catalog.

This means that additional pages are largely irrelevant to the customer ... especially when the customer has access to your entire merchandise assortment online.

The merchant struggles to think in terms of merchandising online, because the whole process is so algorithmic, so mechanical, lacking inspiration or creativity. The merchant easily conceives an additional four pages in a catalog, she can see what these pages might look like and can easily see how the customer might respond to those pages.

Except that, increasingly, customers don't respond to additional pages.

If you want to do a fascinating study, try this ... when your catalog is in-home on a Monday-Wednesday, randomly call 1,000 customers the following Monday-Wednesday. Ask those customers one question ... ask them what merchandise was featured on pages 8-9 of the catalog they just received?

Record how many customers answer this question accurately.

You see, what matters so much to the merchant is barely even noticed by the customer.

If more than half of your direct channel sales happen online (and that is the case with the majority of catalog brands with an audience under the age of 55), merchandising energy is better spent online.

Online merchandising, unfortunately, is not part of the DNA of modern merchants. This needs to change.

September 23, 2010

Gliebers Dresses: Stupid Marketers and Kewpon

As most of you already know, Gliebers Dresses and Fetzer's Footwear are fictional stories about catalog and online businesses. The stories are created to relate marketing concepts to you via parable. Today, I join an Executive Meeting at Gliebers Dresses.


Roger Morgan (Chief Operating Officer): "Ok, folks, let's get busy, we have a lot to address today. I asked Kevin to Skype-in, just so you know, we're not paying him for this, because budgets are tight, but I wanted an outside opinion as we discuss a couple of concepts today. Ok, first off, let's talk about the possible promotion we want to run with a company called 'Kewpon'. As you know, Kewpon is a fast-growing location-based service that gives their e-mail subscribers opportunities for fantastic discounts at leading retailers. Kewpon is looking to branch-out into national campaigns with online retailers, so I approached Kewpon about the opportunity to run a promotion with Gliebers Dresses."

Meredith Thompson (Chief Merchandising Officer): "Isn't Pepper responsible for Marketing?"

Roger Morgan: "Look, we need to explore bold ideas, and I don't think we should be married to the old-school concept that ideas only come from the folks accountable for Marketing. According to Woodside Research, by 2013, 49% of company employees will be accountable for a portion of the integrated, cross-channel Marketing calendar, so let's not be closed-minded about how we grow our business, ok? I don't want for us to get stuck in old-school thinking anymore."

Lois Gladstone (Chief Financial Officer): "How does a Kewpon promotion work?"

Roger Morgan: "It's really quite innovative, to be honest. We'll take one of our best items, maybe our sheath dress for $49, and we discount the living daylights out of it. Let's say we sell it for $22 as part of a one-day promotion."

Meredith Thompson: "We sell it for what?"

Roger Morgan: "We sell it for $22. Then, we do a rev-share with Kewpon."

Pepper Morgan (Chief Marketing Officer): "Like they get 8% or something similar to what we pay our affiliates when they generate a sale for us? I know most marketing folks hate having affiliates suck 8% out of the demand equation for doing so little actual work."

Roger Morgan: "Oh, no, nothing like that. They get 50%, and we get 50%."

Meredith Thompson: "They get what? Do you realize that the cost of goods on a sheath dress is $19, Roger?"

Roger Morgan: "Of course I understand that, Meredith. Look, I checked out your inventory reports, and we have plenty of sheath dresses to offer as part of this promotion."

Meredith Thompson: "But Roger, think about it for a minute. We get $11. Kewpon gets $11. The item costs us $19. How do we make money on this, Roger? Because, as I understand it, the objective of a business is to turn a handsome profit."

Roger Morgan: "We don't! But that's not why we do this, Meredith. We do this because we get access to Kewpon's subscriber list of 22,000,000 e-mail addresses. They will promote this to every single one of the 22,000,000 e-mail addresses, nation-wide. That's a honkin' boatload of eyeballs that we're about to monetize! Just think about the traffic, the traffic, Meredith!"

Meredith Thompson: "Can we limit the number of items we sell?"

Roger Morgan: "I'll make that a condition of the promotion ... only the first 10,000 buyers get to take advantage of the promotion, given our inventory situation."

Meredith Thompson: "Why do you get to pick the item? I'm responsible for merchandising, Roger."

Roger Morgan: "Again, that's old-school thinking. We need to think like these modern digital marketing experts think. I recently read a Woodside Research report about digital marketers, and they said ... "

Meredith Thompson: "Lois, let's do a quick profit and loss statement here. We sell 10,000 items, we get $11 per item, we paid $19 per item, so we lose $8, so that's an up-front gross margin loss of $80,000. That kills the annual bonus for the merchant responsible for sheath dresses."

Roger Morgan: "We need to stop thinking about old-school incentives like bonuses, we need to start thinking like modern digital marketing experts. How many times do I have to say this?"

Meredith Thompson: "We lose $80,000 of gross margin. Roger, who pays to pick, pack, and ship the items, Kewpon or us?"

Roger Morgan: "We do. And we'll give the customer free shipping as our way of saying thanks. That should really sweeten the deal."

Meredith Thompson: "Ah. Now we get to lose another $10 of shipping and handling revenue, so thus far, we've lost $180,000 on the promotion. When the customer returns an item, and 20% of our sheath dresses are returned, does Kewpon refund us their share of the transaction?"

Roger Morgan: "No, they keep their money. See, that's not their fault, it's our fault the customer returned the item."

Meredith Thompson: "Oh, they keep their money. But we take the item back, so that means we refund $22 to 2,000 customers, so we lose another $44,000, right?"

Roger Morgan: "But we get the gross margin on those items back, because we get to re-sell them, so we really lose $3 per item, costing us just $6,000. Glass half full, Meredith."

Meredith Thompson: "Fine, we lose just $6,000. In total, we lose $186,000."

Roger Morgan: "Wrong again. Kewpon told us that 20% of the customers will buy an additional item, this item would be at full-price, so for 10,000 customers, we'll get 2,000 additional items at an average gross margin of $30, so we'll make up $60,000 there as well."

Meredith Thompson: "Oh, Kewpon told you that. What, did they access our database and mine the information themselves?"

Roger Morgan: "Meredith, you always want to block new ideas and modern strategies. I just found a way for us to acquire 10,000 new customers, and we'll only lose $126,000. That means we lose $12.60 per new customer. That's well within our new customer acquisition cost limits."

Pepper Morgan: "How do you know that all 10,000 customers are new customers?"

Roger Morgan: "Kewpon told me that the vast majority of customers who take advantage of Kewpon promotions are typically new customers."

Meredith Thompson: "So you want for us to lose $12.60 per new customer?"

Roger Morgan: "The math is really quite optimistic. I asked Kevin to join us today because he did all of that lifetime value work for us. We know that the average newly acquired customer generates $15 profit in the next twelve months, so quite honestly, we'll come out ahead on this one. I'm making us money, something like $20,000 in the next twelve months."

Kevin: "The average customer generates $15 profit ... full price customers generate $25 of future value, discount/promotion buyers generate $5 profit."

Roger Morgan: "And we really have no idea what a Kewpon customer will generate, because we've never tried a promotion like this, so let's assume it is $15. The math works! And Kevin, you're always telling folks on your blog that they have to be willing to test new ideas. Well, here I am, I'm willing to test a new idea, and I've got math to support my numbers. So we're going to do this."

Meredith Thompson: "Your math is based on rogue assumptions."

Roger Morgan: "Look, Meredith, I'm attending Shopper.com's annual conference next week. They invited the CEO of Kewpon to be a keynote speaker. I'm meeting with the CEO after her keynote. Shopper.com wouldn't invite her to be the keynote speaker unless there was something really intriguing about her business model."

Lois Gladstone: "The math is intriguing, Roger."

Meredith Thompson: "Lois, the math is crap. Kewpon gets 10,000 orders at $11 per order. They make $110,000, no risk, no picking, packing, or shipping, no processing of returns ... just money. We lose $126,000 and deal with all of the headaches.. You are the CFO. On any other planet, you'd refute this argument as being complete garbage. Why would we ever give a third-party all of our profit? Intermediaries are sucking the life out of this business. Google, Kewpon, the USPS, it never ends."

Lois Gladstone: "Meredith, have you ever seen our Response Shop bill? We pay co-ops $0.06 per name for one-time use, month-after-month-after month, and only one percent of the customers we rent from Response Shop purchase. That means we technically pay Response Shop $6.00 for every customer who purchases from us. In this case, we're paying Kewpon just $11.00 for every customer who purchases from us."

Pepper Morgan: "Of course, the Response Shop customer is paying full price, plus shipping and handling, so we usually make money on that transaction, and then get $25 of long-term value after making money on the transaction. That's why we work with Response Shop, we get profit on the front-end, and we get profit on the back-end."

Lois Gladstone: "In this new, complex, modern world of digital marketing, especially during challenging economic times, we have to be willing to adapt to new business models. With Response Shop and with any co-op, we violate the rights of the customer by shoving a catalog in her mailbox when she didn't ask to receive the catalog. That's technically a violation of her privacy, and the customer doesn't make any money in the transaction, only Response Shop makes money. With Kewpon, the customer saves $27, the customer decided which Kewpon promotions to accept, and Kewpon gives us access to 22,000,000 e-mail addresses nation-wide. Kewpon makes money, the customer saves money, and Gliebers Dresses has the opportunity to make up the losses on the back-end. Meredith, it's up to you to convert these customers into full-price customers, downstream. Honestly, you're the one who is accountable to make this work, long-term."

Roger Morgan: "That's some sweet action!"

Meredith Thompson: "When did we become stupid marketers, everybody?"

Roger Morgan: "We're not going to be stupid marketers next Friday when we sell 10,000 sheath dresses that we wouldn't have sold otherwise. Imagine the rush folks working in the building will feel when we're 100% over plan? When is the last time we were 100% over plan? It will be like one of those telethons from the 70s, with folks chanting 'light those lights ... light those lights ... light those lights'!"

Meredith Thompson: "No really, when did we become stupid marketers? Do you remember Glenn Glieber's favorite saying about marketing?"

Pepper Morgan: "I love free marketing!"

Meredith Thompson: "Yes, that's what he always said. He loved free marketing. He hated paying for anything. Worse, he hated to lose money."

Roger Morgan: "We're not losing money, we'll make it up on the back-end. You old-school folks really need to start thinking more like us digital marketing experts, you know, the folks that are truly driving the new economy in this post-recessionary environment."

Meredith Thompson: "How do you know that we'll make it up on the back-end, given we've never tried anything like this before? Yours is a theory, a hypothesis, not a reality. You are assuming that these customers will purchase again, and they will purchase at full price. Kevin just told us that the discount/promotion buyer has $5 of lifetime value, while the full-price customer has $25 of lifetime value. Why do we think we can convert this Kewpon customer into a Gliebers Dresses full-price customer?"

Roger Morgan: "That's your job, Meredith, your merchandising strategy has to be held accountable at some point. Your job is to provide such compelling merchandise that the Kewpon customer cannot help but buy merchandise at full price. That's your job, Meredith. In this case, it is my job to drive traffic. We Digital Marketers are all about driving traffic, it's up to the rest of the employees to turn that traffic into profit. Old-school marketers are afraid of this paradigm shift."

Meredith Thompson: "You don't take a discount/promo customer loyal to another platform and convert them into a loyal, full-price Gliebers Dresses customer."

Roger Morgan: "Figure it out, Meredith, that's your job. You are accountable here."

Lois Gladstone: "I'm willing to make the investment, then let's measure the results and see if it was a worthwhile endeavor."

Roger Morgan: "Woodside Research and Shopper.com seem to suggest this is a worthwhile endeavor. Look at all of the sales that Shopper.com created with their innovative post-Thanksgiving online holiday event? These organizations love to promote the concept of giving the customer a great deal at a low price, and their interests have to be aligned with ours, or they would go out of business. Woodside Research and Shopper.com would not lead us astray."

Meredith Thompson: "Kevin, can you help me here? Have we all become stupid marketers? I think we're just plain stupid. Since when is it a good idea to lose money?"

Roger Morgan: "I didn't ask Kevin to be here for free today just to beat us up, that's not in our best interests. I asked Kevin to be here so that he could provide us with the lifetime value of an average Gliebers Dresses customer, and he did that, so thank you Kevin, you can leave now."

Meredith Thompson: "No, he can't leave. I want to hear his opinion. Kevin, tell us about marketers in general, keep Gliebers Dresses out of this for a moment. Are marketers stupid?"

Kevin: "Keeping Gliebers Dresses out of the equation, I think that Marketers are mis-guided. Look at what happened when Gap did a promotion with Groupon, that promotion was mentioned by the national media, by the trade journals, by just about everybody associated with direct marketing. If you are a Marketer at Gap, you were on top of the world for a few days, weren't you? Within a company, the marketer only gets beat up, beat up by the merchant for not promoting merchandise, beat up by the operations expert for not being a modern, smart marketer, beat up by the CFO for wasting valuable resources, beat up by the CEO for not growing top-line sales fast enough. In fact, it is always about top-line sales, isn't it? When is the last time a Marketer had to provide the CEO with a profit and loss statement of all marketing activities, illustrating both short-term and long-term profit?"

Roger Morgan: "We just went through a profit and loss statement here, so clearly, we're not stupid marketers, are we?"

Kevin: "All of the information technology systems in our industry are designed to illustrate what happened to top-line sales after a catalog was mailed, an e-mail was delivered, a search-term was clicked on, or after a social media follower visited the website. Our web analytics and business intelligence software packages seldom interface with financial information, so we seldom see the short-term profitability of our work, and we almost never get to see the long-term profit impact of our work."

Roger Morgan: "We have a lot of that type of data integration on our book of work, and we're likely to get to it in 2013, so we're moving in the right direction."

Kevin: "So you take a Marketer who isn't respected in her own company ... I mean honestly, how many times did you as a team ask Pepper today what her thoughts were on this Kewpon promotion, and remember, she's the Chief Marketing Officer, right? The Marketer isn't respected in her own company, because every employee thinks that he is a marketer at heart. So the Marketer gets little internal validation. The Marketer is forced to look outside for validation, and there are plenty of places to get external validation ... Twitter followers, Facebook fans, Kewpon promotions, keynote presentations at the Shopper.com conference, you name it. The Marketer doesn't get love internally, the Marketer doesn't have internal reporting to show how much profit the Marketer actually generates on a short-term and long-term basis, so the Marketer goes for the only things that provide love ... external recognition and internal top-line sales growth. Honestly, it's a failure of the entire ecosystem, when it comes right down to it. If the Merchant respected the Marketer, if the Executive team respected the Marketer, if the Operations executive didn't try to be a marketer, if the web analytics / business intelligence team / information technology team actually provided the marketer with the right information, and if the Marketer was motivated by profit and not by top-line sales, then the whole thing could work. In lieu of that, we've trained a generation of Marketers to be stupid, to look for recognition via the wrong metrics, metrics that don't correlate with profit. And we trained a generation of web analysts, business intelligence analysts, and IT professionals to provide the wrong information."

Lois Gladstone: "I agree with Roger, we shouldn't have asked Kevin to pop-off about his opinions of Marketing. If we assume that the Kewpon customer has average long-term value, then this promotion is fine."

Kevin: "But that's exactly the problem. Lois, that's 'truthiness'. You want to believe that the Kewpon customer has average long-term value, because the assumption of average long-term value leads one to a math calculation that enables one to run the promotion. And then you'll never measure the promotion on the back-end, because there will be some other priority that is more important than analyzing something that happened in the past. This is where Marketers are stupid. Marketers always get to manipulate figures in a way that enables the Marketer to do something stupid. We have more facts and figures than we've ever had in history, and yet, we don't have anything resembling actionable insights."

Lois Gladstone: "Roger, maybe it is time to ask Kevin to leave?"

Kevin: "We all know that the Kewpon customer is more likely to generate $5 long-term value, congruent with the discount/promotion customer, than the $15 long-term value of the average customer, or the $25 long-term value of the full-price customer. It is the lie that we tell ourselves that enables us to do stupid things, and then it's the failure to measure on the back-end and learn from our mistakes that makes us stupid. Roger, I can respect a person for wanting to test a promotion. There's nothing wrong with saying that you want to partner with Kewpon, I applaud that. I have zero problem with testing a promotion and losing money in the process."

Roger Morgan: "Keep talking."

Kevin: "But I am 100% against the way that Marketers use truthiness to justify actions. The Marketer will point out that the Kewpon promotion may lead to word of mouth, and that you cannot discount word of mouth, because word of mouth is the modern way that Marketers leverage customer interactions, right?"

Roger Morgan: "Yeah, we didn't factor word of mouth into our profit and loss statement. That alone might make this thing profitable! Think of the social media ramifications. Oh, this whole thing could go viral. You really don't want to stop something that could go viral, right? That's the strength of social media."

Kevin: "Or it might mean nothing. Or there might be negative word of mouth that devastates Gliebers Dresses. That's the problem. We'll use the social media and word of mouth argument if it benefits our objectives, and we'll discount the word of mouth argument if it works against us. We use all of these facts, figures, and metrics to defend truthiness, we don't use them to improve business performance."

Roger Morgan: "Well, folks, that's the end of the hour. I think between what Woodside Research says, what Shopper.com says, and Kevin's average long-term value estimates, this Kewpon promotion is the right thing to do, and Kevin just said that he lauds folks who test new strategies, so we're going to do it! Thanks for helping us flesh through this situation, Kevin, and thank you for spending an hour with us for free ... you know what trying economic times we are faced with, so every little bit helps us achieve our profit objectives."

September 22, 2010

Digital Profiles Part G: Profiles 13-16

It's time to describe the final four Digital Profiles!

Digital Profile #13 is called "Retail 1/2/3". 99% of these customers purchased in stores last year, and their merchandise preference was in merchandise divisions one, two, and three. These customers purchased 2.0 times last year, which isn't bad, but they are only expected to spend $144 in the next twelve months, considerably below the company average. These customers also seem to like merchandise division seven, they are more likely than average to live in retail trade areas, and they are more likely than average to be first-time buyers. As marketers, we should always pay attention to any Digital Profile that is comprised of first-time buyers.

Digital Profile #14 = "Retail Newbies Looking For Eight". These customers are more likely than average to be first-time buyers, with 96% shopping in stores, and 76% buying from merchandise division eight. These customers purchased 1.3 times last year, and are forecast to spend just $82 in the next twelve months, well below the company average. Obviously, the challenge is to somehow encourage these customers to experience more of the brand than just one merchandise division in just one channel across just one purchase ... and that's easier said than done! Sometimes, you simply have customers that are not going to be productive, end of story!

Digital Profile #15 = "Retail 1/2". Speaking of customers that are not going to be productive, "Retail 1/2" is a Digital Profile that isn't very productive!! In the next twelve months, these customers will only spend $83. 85% of this Digital Profile purchased in retail stores, with customers almost exclusively buying from merchandise divisions one and two. Thirty percent of these customers were first-time buyers last year. Interesting to note: 16% of these customers shopped via the telephone last year.

Digital Profile #16 = "Rotary Phone". This is a very interesting Digital Profile. 66% of the customers in this Digital Profile purchased last year via the telephone, placing only 1.1 orders last year. In the next twelve months, these customers are forecast to spend just $85. 72% of these customers purchased from merchandise division four ... so the marketer should be able to easily explain why customers shopping over the telephone are pre-disposed to purchases in just one merchandise division! A whopping 60% of customers in this Digital Profile do not live in a retail trade area, wow! Zip Code Forensics data strongly suggest that these customers live in traditional catalog-shopping zip codes, meaning that these are rural shoppers, and are probably in the top-half of the Baby Boomer generation. 32% of these customers were first time buyers, suggesting that the business used catalog marketing to acquire new customers via the telephone.

That's it!!! We've described each of the sixteen Digital Profiles. Up Next: A series of blog posts describing how we can use Digital Profiles to understand customer behavior in an actionable manner.

September 21, 2010

Digital Profiles Part F: Profiles 9-12

We're halfway through our evaluation of all sixteen Digital Profiles. Here's a description of profiles nine through twelve.

Digital Profile #9 = "Retail Fanatics". In our "multi-channel" obsessed marketing environment, we seldom praise the customer that spends a boatload of money with us in just one channel. That's a shame, folks. With $548 of expected spend in the next twelve months, Retail Fanatics are the fifth-most productive profile in our analysis. 99.9% of these customers shop in stores, and only 11.5% shopped multiple channels last year. In other words, these customers love stores, and they spend a lot of money in stores. Take a look at the merchandise divisions that the customers purchase from ... all but divisions seven and nine. Remember, I mentioned in earlier posts that customers buying from a robust cross-section of merchandise divisions are generally more productive than are customers buying across multiple marketing channels ... this theme continues to play out in this analysis! Also notice that these customers are likely to live in multiple retail store trade areas. When you have customers who live near multiple stores, it becomes more and more difficult to get the customer to buy from online channels. This doesn't mean that the customer doesn't use multiple online channels, s/he is simply less likely to buy from online channels.

Digital Profile #10 = "Retail 4/5/6". This Digital Profile will spend $293 in the next twelve months, just a bit above average. Their preference is in merchandise divisions four, five, and six, and they also tend to like merchandise division eight a bit. These customers are likely to reside in the trade areas of one or two stores. In all, these customers are very "average", with a specific merchandise preference.

Digital Profile #11 = "Retail 1/2/4". This Digital Profile will spend $243 in the next twelve months, and is generally similar to "Retail 4/5/6" except for the fact that these customers prefer a different combination of merchandise divisions. Nine out of ten customers in this profile shop in stores, and these customers are more likely than average to live in at least three retail trade areas.

Digital Profile #12 = "In And Out". Oh boy, we're about to get a preview of the portion of the customer base that doesn't shop much. These customers bought 1.2 times last year, and 90% of the customers purchased in retail stores, probably because these customers are more likely to live in zip codes that comprise three or more retail trade areas. 90% of these customers purchased from merchandise division four. In other words, these customers come into the store to buy from merchandise division four, and then they leave. In the next twelve months, these customers will spend just $106, less than half the overall average.

Twelve Digital Profiles have been described, we have just four to go. Up next ... Digital Profiles thirteen, fourteen, fifteen, and sixteen.

Stupid Marketers, Part 2

Did you watch the Groupon story on Nightline last night?

See, I'm one of the stupid marketers I talked about in yesterday's rant titled 'Are We Marketers Just Plain Stupid? ... I didn't know the story would be on television (see, television is old-school, only used by Luddites). My wife, however, told me to record Nightline and watch what they said about Groupon.

So I recorded the episode.

Turns out that 70% of Groupon users are under the age of 35, are female, and live in cities.

"It was free to sign up, It was a place were we would go anyway, and, you know, we saved money and it was fun". Those are the words of Erin Donahue, a 27 year old woman interviewed for the story.

"I would say 60% of the people who come in here through the Groupon didn't even know I was here ... it put me on the map." said Richelle Ciluffo, Owner of "The Spa on Oak".

It was mentioned that Groupon is on track to rake-in a half a billion dollars in sales this year and is already turning a profit.

Somehow, a portion of our readership missed the most important point of the article.
  • "The men and women at Groupon are the smartest marketers in the room".
I am not criticizing Groupon. Please demonstrate to me the last time you founded a business and in 22 months you were able to grow it to a half-billion dollars in sales with 500 employees in the teeth of something just shy of a depression?! They are very, very smart marketers. They found a way to give their customers something cheap, and they found a way to capture our profit.

I am not criticizing Groupon.

I am criticizing us!


We are stupid.

Let's use me as an example. I wrote the article yesterday, knowing it had a chance of "going viral". And it went viral, all right! Last I checked, more than 1,300 marketers read the article.

And for all of that viral magic, I earned an additional 25 followers on Twitter.


See, Twitter, those folks are the "smart marketers" ... they theoretically benefit over and over and over again from the content we produce. On the other hand, I got 25 additional followers. Of the 700 incremental blog readers who visited my site and read the article, fewer than a dozen clicked on any of the links on the blog, and I sold zero consulting projects.

I am a "stupid marketer". My intellectual capital benefited Twitter readers (i.e. Groupon customers). My intellectual capital benefited Twitter (i.e. Groupon). My intellectual capital didn't benefit me (i.e. the retailer).

In fact, in the two years since "joining the conversation" on Twitter, I've generated a whopping $2,500 of consulting revenue, and that revenue won't be booked until November ... so at this time, I've technically generated $0.

Increasingly, we marketers partner with intermediaries. And when we partner with intermediaries, we darn well better accomplish our goals, because intermediaries suck the profit out of our business model.

Our goals are simple.
  1. We will acquire a new customer, a new customer with a positive lifetime value that greatly offsets the loss we suffer when we acquire the customer.
  2. We will increase the lifetime value of an existing customer.
Take the first point. Say you sell a $20 widget that costs you $8. You offer a Groupon, half-off. Groupon gets $5, you get $5. You acquire a new customer.
  • You just lost $3 on the item.
  • You probably lost $2 on operating expenses and salaries as well.
  • In total, you lost $5 to acquire the customer.
Now, the pundits will tell you that when that customer shops with you next time, you'll generate ten dollars of profit ($20 item less $8 cost of goods less $2 of salaries and operating expenses), so the transaction is a net positive.

That's theory.

Go get real customer data from a real retailer or real online brand. I've done this, more than fifty times. The reality, unfortunately, is very different than marketing theory.

In practice, for businesses that acquire customers via discounts and promotions, around 20% of the customers will come back and buy again. And when the 20% of customers do come back, half will purchase only if a promotion is offered to the customer again.
  • This means that 10% of the customers buy again, and they maybe buy twice in the next year, usually at full price. Here, you make maybe $20 profit.
  • This means that 10% of the customers buy again, buying maybe twice in the next year, using a promotion, often generating no incremental profit because of their lust for discounts and promotions, a lust we're happy to feed. Let's assume these customers generate $2 profit in the next year.
Total future profit is 0.80 * $0 + 0.10 * $20 + 0.10 * $2 = $2.20. And this assumes that you do not spend money advertising to this customer again in the next year.

In other words, you lose $5 acquiring the customer, and then the customer pays you back $2.20 in the next year.

In other words, you are a stupid marketer.

Then you have the second example, where you had an existing customer, and you took $20 of future value and you destroyed it by offering the customer who would have purchased anyway at full price a wonderful promotion. This is the example that "Erin" mentioned on Nightline last evening ... she said she used promotions at places where she would have shopped anyway. As marketers, this is our true blind spot ... we almost never view purchases from the standpoint of "the customer would have purchased at full price anyway." The results are devastating.
  • The customer would have generated $10 of profit.
  • Instead, you gave the customer a wonderful promotion, and the customer now generates a $3 loss (salaries and overhead are constant in this situation, the customer would have purchased anyway).
  • Net change to customer lifetime value ... a loss of $13.
In this case, you must get this individual customer to purchase an additional 1.3 times above and beyond what the customer would have purchased, at full-price mind you, in order to recoup this "advertising investment". And this assumes that the customer will shop with you again, which, when analyzing 50+ projects over the past 42 months, I know doesn't happen ... heck, good companies are able to get 50% to 65% of good customers to shop again in the next year ... so you may have just lost $13 of profit and a third of the customers won't come back and the remaining customers will not change their buying behavior and now may only buy from you when you offer promotions when in the past the customer would have purchased anyway from you at full price.

Again, you're a stupid marketer. Why would you do that? Be honest with yourself. Why would you do that?

Does anybody do the math outlined in this blog post? Do you do the math outlined in this blog post?

Old-school catalogers, these folks always place a "hurdle" around their free shipping promotions ... have you ever noticed that? They give you free shipping if you spend $100. Why do they do this? Well, they've "done the math". They know that the hurdle provides enough profit to offset the losses generated by the promotion.

That's being a smart marketer.

Those Groupon folks, they are the smartest marketers in the room.

Please do the math before going down the discount/promotion route. Please be a smart marketer.

September 20, 2010

Digital Profiles Part E: Profiles 5-8

Let's take a look at Digital Profiles five through eight.

Digital Profile #5 = "Shop Online, Buy In Store". Oh, isn't this one just plain dreamy? Marketers love the concept of a customer browsing online and then buying something in a store, and it turns out that there is a profile that represents this behavior. These customers sure are valuable, spending $317 in the next twelve months. Now, what do you see when you look down this column? First of all, these customers buy expensive items, $72 each. Maybe the customer needs to touch/feel expensive merchandise before buying it, who knows? Notice that all online channels except for social are above-average ... these customers want to transact online! Then notice that more than half of these customers buy merchandise in stores as well. Also notice that merchandise divisions one, two, three, eight, and nine are above average ... we haven't seen this combination yet, so there is something about this type of merchandise that warrants an online shopper who might want to buy in stores as well. 55% of these customers are the fabled "multi-channel" shopper that marketers love to talk about, and these customers do live in online-responsive zip codes.

Digital Profile #6 = "Direct Newbies Looking For Eight". The "eight" represents merchandise division #8, which is about all these customers want to purchase. Look at the price points as well, at $147 per item, it's the most expensive of all price points in this analysis. These customers tend to buy online and via search at an above-average rate, suggesting that these customers are searching online for this merchandise. These customers are only worth $135 next year, in the lower half of all segments for future value. Many customers were first-time buyers, suggesting search might play a role in recruiting newbies, or that newbies like merchandise division #7?

Digital Profile #7 = "Online One-Twos". These customers purchase via the website, do purchase via search, and like buying from merchandise division one and two. These customers do tend to live in retail trade areas, and tend to live in areas where catalogs are traditionally responsive. These customers are similar to "Direct Newbies Looking For Eight", except that they prefer merchandise divisions one and two instead. They are also worth $135 next year. Again, many customers were first-time buyers.

Digital Profile #8 = "Ship It To Me". They are worth only $103 in the next twelve months. They buy expensive items via online/search, they like merchandise division #7, and they are second-most likely to not live in a retail trade area, living instead in areas traditionally served by catalog marketing. An above-average number of these customers are first-time buyers.

Ok, we're half-way home. Up next ... profiles nine through twelve.

Are We Marketers Just Plain Stupid?

There's a reasonable chance that you or somebody you know is attending Shop.org's annual Summit next week. And you'll ask me ... "Why aren't you attending?" ... or ... "Why don't they ever invite you to speak?"

The latter question comes up all of the time ... "Why don't the big US-based conferences ever invite you to speak?"

I've been consulting for three-and-a-half years now, and there's one thing I've learned that is a "best practice" if you want to be invited to speak at US-based conferences.
  • "Follow the storyline."
The storyline is the script that folks want you to follow. And the more we follow the storyline, the less we pay attention to what really matters.

Remember, the storyline is designed to maximize the return on investment for those who host conferences.

The storyline fails us, over and over and over again, because the storyline isn't designed to maximize our return on investment.

Honestly ... go poll 100 companies that are similar to yours. How many of the 100 companies generate 10% of their sales via Social Media? Seriously, go poll 100 companies that are similar to yours ... you'll find fewer than five of one hundred that generate at least 10% of their sales via Social Media. So if Social Media is where all the cool kids hang out and is so vitally important, then why isn't Social Media a vehicle that drives sales on a consistent basis for all but a few retailers?

Now go check out the conference agenda for your favorite industry conference, and you'll find a buffet of Social Media topics ... "Eight ways to reap the rewards of employee blogging" or ... "The future of location-based micro-services" or ... "Nine tips for success on Facebook and Twitter from a panel of industry experts."

Fortunately, Social Media is harmless ... it costs nothing, so there's no risk to it ... you should experiment in low-cost endeavors, right?

The worst fraud perpetrated on us this year is Groupon. This is not a low-cost endeavor.

Groupon represents "the smartest men and women marketers in the room". They truly are. We, on the other hand, are just plain stupid.

Where else but in marketing would one cut the price of an item by 50%, then share half of what is left with a company that does a great job of driving traffic? Groupon gets $5 to spread across the cost of driving traffic, while you are left with a curious math problem ... how do I generate profit when I have $5 and I have to pay for the $10 cost of goods sold?

It's sort of like what happens when you make $4,800 a month in salary and you have to pay $9,600 a month in mortgage payments and other expenses. Would you run your personal life that way?

Go ask the merchant who is responsible for the sales of widgets what happens to her if she generates negative gross margin dollars ... SHE LOSES HER JOB!

But in marketing, WE CELEBRATE THIS OUTCOME AS A SUCCESS! Seriously, in what profession other than Wall St. or Marketing is losing money considered a success? Maybe the average Chief Marketing Officer keeps his job for only twenty-three months because, as a marketing community, we follow the storyline, constantly giving up profit dollars in order to adhere to the storyline.

The storyline offers us a lot of low-probability routes to profitability ... enticing us with tantalizing treats that work for some, but seldom for the masses.

Social Media is always part of the storyline, but is more like a roulette wheel, generating significant sales and profit for maybe 3% of participants. As one popular Social Media consultant recently said ... "It's where the cool kids hang out." You want to generate profit, you don't want to be a cool kid ... right?

Mobile may be really important in a few years, but it sure isn't going to be the key factor that determines Holiday sales effectiveness ... YOUR MERCHANDISE will be the key factor that determines your Holiday sales effectiveness, it has always been about YOUR MERCHANDISE! Are you focusing on an iPad app, or are you focusing on how to sell merchandise on an iPad app?

E-mail was promised to be the "best return-on-investment" of any marketing vehicle. And yet, I keep analyzing e-mail campaigns that deliver $0.08 or $0.12 net sales per e-mail delivered, yielding $0.03 profit. That's awful, pitiful, just plain lousy ... and yet we're told that e-mail and social media are like peanut butter and jelly ... yup, they sure are ... they're like a teaspoon of peanut butter and jelly when what we really need is a full sandwich to satisfy our hunger.

The men and women at Groupon are the smartest marketers in the room.

We, however, are just plain stupid. We willingly give up profit in exchange for traffic that is loyal to Groupon, traffic not loyal to us. And if you don't agree with me, then go find a dataset of actual customer purchase transactions from a real retailer or online brand, and prove to all of us that you'll make up what you lose on the Groupon transaction in long-term profit (or even short-term profit). Go get the data, and present your analysis for all to see ... I'll even publish your results here on this blog if you can prove to me that you have actual customer data from a real retailer or online merchant that validates your result.

The storyline is leading us astray.

I don't get asked to speak at a lot of mainstream conferences, and for good reason. As you know, we focus on actual customer behavior on this blog, using actual customer data to validate actual profitability. This data, unfortunately, isn't always congruent with the storyline told at mainstream US-based conferences. One wonders if our blind allegiance to the storyline is causing us to become stupid marketers.

September 19, 2010

Dear Catalog CEOs: Fractured Audience

Dear Catalog CEOs:

Have you ever studied the actual customers who purchase from your business?

No, not a focus group of nine customers lured by $60 and free pastries. I'm talking about, at minimum, a demographic / psychographic overlay of data, or an even more comprehensive study where you actually, personally speak with a thousand customers.

It is in this type of study that you actually learn what the past five years of catalog marketing have delivered.

Increasingly, I see that we've delivered an audience that is fracturing away from the mainstream.

The mainstream audience is firmly anchored in e-commerce. And e-commerce is rapidly fracturing into mobile (the pundits will tell you that mobile IS e-commerce ... just like catalog vendors told us in 2003 that e-commerce was the same as catalog marketing). Catalog marketing now happens among a fractured audience ... among the folks who chose not to embrace the internet at the same rate as the mainstream public.

Fractured audiences are interesting to analyze.

For most of my catalog analytics projects, I am seeing the fruits of a fractured audience. This may sound backward, but bear with me.

Eight years ago, a catalog customer base might be comprised of, say, five-hundred customers.
  • 100 Customers = A.
  • 100 Customers = B.
  • 100 Customers = C.
  • 100 Customers = D.
  • 100 Customers = F.
The best customers had a grade of "A", the most marginal customers had a grade of "F".

Today, after adjusting for the impact of the Great Recession, the customer file of a typical catalog brand looks something like this (and I see this over and over and over and over again).
  • 100 Customers = A.
  • 50 Customers = B.
  • 50 Customers = C.
  • 150 Customers = D.
  • 150 Customers = F.
In other words, there are just as many good customers as there were years before ... those are the classic catalog shoppers of years gone by (Hint, they are increasingly 55+ year old shoppers).

The middle of the file is shrinking, as those customers abandon classic catalog marketing on a macro level.

The bottom of the file is growing ... a plethora of one-time shoppers seeking to get a need met at a low price. They are not brand loyal shoppers, and often, they are online shoppers.

Strategically, this is a problem, because it is awfully hard to run a business on a small audience of loyal shoppers.

Mechanically, this is a blessing, because you can save a boatload of expense by identifying the audience that has now fractured from the catalog core audience.

Instead, we keep trying to figure out how we are going to get the Ds and Fs to "love catalogs". Those days are gone. We need to move on.

Call me or e-mail me to schedule your Multichannel Forensics project for this November/December, and get your results in plenty of time to impact the profitability of your business when postage increases next year.

September 15, 2010

Digital Profiles Part D: The First Four Digital Profiles

Let's take a look at the first four digital profiles in our sample dataset.

The first digital profile is called "Gold Mine", and for good reason! There are 8,600+ customers in this digital profile, and these customers can be expected to spend $841 per customer in the next twelve months!

Look at the attributes that comprise this segment, the attributes that are shaded. These customers purchase many items per order, they buy from the retail channel, and they buy from all online channels. Notice that these customers buy from all merchandise divisions as well ... we so often talk about customers being "multi-channel", but so much more often, it is critical to have a diverse customer base willing to buy from all merchandise divisions. That being said, 73% of this audience buys from at least two of the following physical channels ... phone, online, and stores.

In other words, this segment is a "Gold Mine", they do everything you could possibly want!!

The second Digital Profile is called "Multi-Channel Mavens". These customers are not as valuable as are "Gold Mine" customers, spending $503 in the next twelve months. These customers do shop from all online channels, and do purchase from retail stores. However, these customers fail to shop merchandise divisions one, two, three, and nine ... this dropoff in merchandise preference yields a customer that is simply less valuable.

The third Digital Profile is called "The Digerati". These buyers will spend $291 in the next twelve months, so clearly, these buyers are not as valuable as prior Digital Profiles. These customers purchase more items per order than any other Digital Profile. These customers are unlikely to shop in stores, and have yet to adopt the Mobile channel. Still, these customers focus purchases in all remaining online channels. Notice that these customers prefer merchandise divisions one, two, four, and five ... a subset of the entire merchandise assortment ... this is a problem that multi-channel brands have, a problem where the online audience skews to a merchandise assortment that is not representative of the total business.



The fourth Digital Profile is titled “A Long Drive”. Customers in this profile will spend $211 in the next twelve months, a bit below the overall average in this analysis. These customers have a very specific merchandise preference, in divisions four and five. These customers generally live outside of retail trade areas, and are skewed a bit to very traditional zip codes. Online, these customers are skewed to web purchases, e-mail purchases, and search purchases, and are not pre-disposed to purchase via mobile or social channels. We might infer that these customers live in rural areas, not near retail stores, given what the data indicates. Given the dramatic skew to merchandise divisions four and five, one might infer that there are many opportunities to personalize the website experience and to work on targeting strategies for this unique audience.

Four Digital Profiles are in the books, twelve to go. Hint: We've covered many of the high-value profiles already!

September 14, 2010

Digital Profiles Part C: An Introduction To Each Profile

Well, we've made it to the fun part of the project!!

In Part A, we collected the data necessary to create Digital Profiles.

In Part B, we executed a whole bunch of geeky math, math that was necessary to create each Digital Profile.

In Part C, I introduce a table that outlines the composition of each Digital Profile.

Here's how we read the table. Each row represents an attribute. Each column represents one of the sixteen Digital Profiles. Each cell represents the actual value of that metric for the customers in the Digital Profile. Any cell that is shaded is a cell that has an above-average value for that metric. The shadings are used to name each Digital Profile.

And in case you didn't notice, it isn't easy to come up with names for Digital Profiles! You're free to come up with whatever names you feel are appropriate ... what matters is how the Digital Profiles are used, not what the Digital Profiles are named.

Take a look at the Online Buyers row. Notice that the first eight Digital Profiles all have above-average values for Online Buyers, whereas the final eight Digital Profiles are virtually absent of Online Buyer. This happened because the first factor was heavily weighted toward online buyers. So, in our Digital Profile analysis, we'll want to see how the first eight Digital Profiles perform in comparison with the final eight Digital Profiles.

Take a look at the Online, E-Mail, Search, Social, and Mobile rows. You'll see that across the first eight Digital Profiles, there is a continuum of digital "savvy" ... Social and Mobile in the first two profiles, then Social drops off, then E-Mail drops off. This will be interesting to monitor over time as Social and Mobile increasingly replace Online and E-Mail as complementary channels.

Up Next: Four posts, with the description of four Digital Profiles each.

September 13, 2010

Digital Profiles Part B: Diagnostics And Math

Welcome to the next part in our series about Digital Profiles!

Recall in Part A that we created a dataset, a spreadsheet if you will. The spreadsheet has one row per customer, and has a series of columns that describe how the customer behaves.

Next, I create a table that shows the average value and standard deviation for each variable. This will be used later in the methodology to score each customer prior to segment assignment. The table is illustrated here on the left hand side of this blog post.

Now, things get a little complicated, and for good reason!

I am about to use a methodology called "Factor Analysis" or "Principal Components Analysis". Essentially, I want to ask the computer to combine any/all similarities in the database. Maybe all store customers buy from merchandise division #3, while all e-mail customers buy from merchandise division #8. If that is the case, then the methodology finds those similarities, and creates "factors" that are combinations of all of the similarities in the data.

Ok, here's where the statistician-elite jump all over me for violating 22,493 different assumptions. They'll tell me that I cannot us
e 1/0 variables in the analysis, they'll come up with a ton of reasons why my work is garbage.

Horsefeathers!

Listen, I'm not trying to protect a cancer patient from an experimental drug that may cause devastating side effects ... I'm simply trying to understand customer behavior better, and then convert that understanding into a series of pr
ofiles/segments that can be utilized in an actionable manner.

The factor analysis yields a rotated component matrix. I look for any value below - 0.200 or above 0.200. When I see this, I know that the variable with that coefficient plays a role in that factor.



In most cases, the first four factors capture enough relevant information for a Digital Profile analysis.

In this case, the following attributes were important in the first factor.
  • # of Orders
  • Online Buyer
  • E-Mail Buyer
  • Search Buyer
  • Social Media Buyer
  • Mobile Buyer
  • Retail Buyer (Negative)
  • Merchandise Division 5
  • Merchandise Division 7
  • Multi-Channel Buyer
The second factor listed these attributes as being important.
  • Number of Orders
  • Price per Item (Negative, meaning low-cost items are favored)
  • Items per Order
  • Retail Buyer
  • Merchandise Division 4
  • Merchandise Division 5
  • Merchandise Division 6
  • Merchandise Division 8 (negative)
  • First Time Buyer (negative, meaning existing customers are favored)
  • Multi-Channel Buyer
The third factor listed these attributes as being important.
  • Number of Orders
  • Price per Item (high-cost items are favored)
  • Retail Buyer
  • Merchandise Division 3
  • Merchandise Division 5
  • Merchandise Division 6
  • Merchandise Division 8
  • Merchandise Division 9
  • Multi-Channel Buyer
The fourth factor listed these attributes as being important.
  • Number of Orders
  • Price per Item (negative, meaning low-cost items are favored)
  • Items per Order
  • Merchandise Division 1
  • Merchandise Division 2
  • Merchandise Division 3
  • Merchandise Division 4 (negative)
  • Merchandise Division 5 (negative)
So, each factor describes something interesting about customers.

Now, here's where I go and once again ruin any remaining reputation I have with MS-level statisticians.

Each factor results in a score ... a standardized variable (mean = 0), with positive values reflecting the presence of the attributes listed above, a negative value representing a lack of the attributes listed above.

I run a quick analysis on each factor, identifying the median value of each factor. The median value should be reasonably close to zero, but seldom is with this kind of marketing data.
  • Factor 1 Median = - 0.542.
  • Factor 2 Median = 0.064.
  • Factor 3 Median = - 0.193.
  • Factor 4 Median = - 0.295.
Why did I do this?

Well, this is going to form the basis of the segmentation strategy.

I create a new variable, called "DP1". If factor 1 > - 0.542, set DP1 = 1, otherwise, set it equal to 0.

I create a new variable, called "DP2". If factor 2 > 0.064, set DP2 = 1, otherwise, set it equal to 0.

I create a new variable, called "DP3". If factor 3 > - 0.193, set DP3 = 1, otherwise, set it equal to 0.

I create a new variable, called "DP4". If factor 4 > - 0.295, set DP4 = 1, otherwise, set it equal to 0.

Ok, still with me? Good!

The final step is to evaluate every single combination of DP1, DP2, DP3, and DP4 values. Since each variable is a 1/0 indicator, there are 2*2*2*2 = 16 combinations. These, ladies and gentlemen, are the Digital Profiles!

If (DP1 = 1) and (DP2 = 1) and (DP3 = 1) and (DP4 = 1) Digital Profile = 1.
If (DP1 = 1) and (DP2 = 1) and (DP3 = 1) and (DP4 = 0) Digital Profile = 2.
If (DP1 = 1) and (DP2 = 1) and (DP3 = 0) and (DP4 = 1) Digital Profile = 3.
If (DP1 = 1) and (DP2 = 1) and (DP3 = 0) and (DP4 = 0) Digital Profile = 4.
If (DP1 = 1) and (DP2 = 0) and (DP3 = 1) and (DP4 = 1) Digital Profile = 5.
If (DP1 = 1) and (DP2 = 0) and (DP3 = 1) and (DP4 = 0) Digital Profile = 6.
If (DP1 = 1) and (DP2 = 0) and (DP3 = 0) and (DP4 = 1) Digital Profile = 7.
If (DP1 = 1) and (DP2 = 0) and (DP3 = 0) and (DP4 = 0) Digital Profile = 8.
If (DP1 = 0) and (DP2 = 1) and (DP3 = 1) and (DP4 = 1) Digital Profile = 9.
If (DP1 = 0) and (DP2 = 1) and (DP3 = 1) and (DP4 = 0) Digital Profile = 10.
If (DP1 = 0) and (DP2 = 1) and (DP3 = 0) and (DP4 = 1) Digital Profile = 11.
If (DP1 = 0) and (DP2 = 1) and (DP3 = 0) and (DP4 = 0) Digital Profile = 12.
If (DP1 = 0) and (DP2 = 0) and (DP3 = 1) and (DP4 = 1) Digital Profile = 13.
If (DP1 = 0) and (DP2 = 0) and (DP3 = 1) and (DP4 = 0) Digital Profile = 14.
If (DP1 = 0) and (DP2 = 0) and (DP3 = 0) and (DP4 = 1) Digital Profile = 15.
If (DP1 = 0) and (DP2 = 0) and (DP3 = 0) and (DP4 = 0) Digital Profile = 16.

We're there!!!!

Each customer has been assigned a Digital Profile, one of sixteen combinations of variables created from the Factor Analysis.

When I score files in the future, I use the standardized variables, created from the mean and standard deviation of each variable in the first table in this post, and use the Component Score Coefficient Matrix (I know, really, REALLY geeky) to calculate each factor.

Up Next: We'll begin to look at the composition of each Digital Profile!

September 12, 2010

Dear Catalog CEOs: Online Marketing Doesn't Work

Dear Catalog CEOs:

"Online Marketing Doesn't Work."

That's a quote from one of you, to me, at a recent conference.

And then an online marketing executive recently told me this gem:

"Print, catalogs, offline marketing, none of that stuff works, we tried it all."

Oh.

You know what? Here's where I've come out on this topic.

Everything works.

There seem to be four things that cause marketing to work or not work.
  1. A receptive customer audience.
  2. A blend of best practices that are combined with "rule breaking" practices that make a brand "unique".
  3. Talent ... human talent and technological talent.
  4. You have to "want it".
Continually, I find that folks don't "want it". So many catalog marketers don't really want online marketing to work, as it invalidates a career spent honing offline marketing skills. And so many online marketers don't want any form of offline marketing to work, as it invalidates all of the excitement of a population shifting long-standing habits online.

I thought things were bad ten years ago. Things are worse today.

I continually run into business leaders who, in spite of fifteen years of e-commerce growth, do not have an instinct for thinking about e-commerce. Meetings are held without the online marketing team, offline strategies are devised in ways that force the online folks to only execute offline tactics in the online environment. No wonder online marketing doesn't work!

It's no better on the other side of the spectrum.

Online marketers do not have the analytics tools to understand offline customer behavior, and often have no interest in learning about offline customer behavior. Offline folks are continually blasted by the online elite for being Luddites, for not understanding the intricacies of long-tail search marketing, for not having a clue about the viral benefits of a strategically timed "tweet".

What does work is merchandise that customers are passionate about. Just ask Apple, for instance. Channels are irrelevant when customers crave your merchandise.

Do customers crave your merchandise? That's a question worth spending a few hours talking about with your Executive team.

Go find a marketer, online or offline, who is passionate about merchandise, and you're likely to find a marketer who says that any channel works!

September 08, 2010

Digital Profiles Part A: Necessary Data

The first part of the Digital Profile process is data acquisition. Think about data in terms of a spreadsheet.
  • Each row in the spreadsheet represents a unique customer.
  • Each column in the spreadsheet represents customer behavior during a finite period of time.
Of course, I do the majority of my work in e-commerce, retail, and cataloging. For these businesses, an appropriate period of time is one year. You really don't want irrelevant data from four years ago clouding how customers behave today, this results in an uninteresting segmentation scheme that is ultimately reduced to "RFM" attributes.

Instead, you want a short period of time. Maybe your business is different, and if so, that's fine, just take the concepts I explore here and apply them to different timeframes.

In my example, I pulled data from the customer database I am analyzing ... using 2009.05.01 to 2009.04.30 as my one year time interval.

Next, I calculated a series of "KPIs", to use the parlance of the day, metrics that illustrate traditional customer behavior during this one-year period of time. Each metric is represented as a column in the spreadsheet.
  • Demand: Amount spent by the customer in the past twelve months.
  • Frequency: Number of orders in the past twelve months.
  • Items: Number of items purchased in the past twelve months.
These three metrics allow me to create another series of metrics, metrics that are actually quite important in the development of a Digital Profile:
  • Items per Order: Divide total items by frequency.
  • Price per Item Purchased: Divide demand by number of items.
At this point, we have six columns in our "spreadsheet" (household_id, demand, frequency, items, items per order, price per item purchased).

I've found, via trial and error, that the remaining variables are best categorized as "1" or "0" indicators.

The business I am analyzing has nine merchandise divisions (tabs that run across the top of the homepage on the website). I create nine columns in the spreadsheet. If a customer purchased from a merchandise division in the past year, I record a value of "1". If a customer did not purchase from a merchandise division in the past year, I record a value of "0".
  • Merchandise Division #1: 1 = Yes, 0 = No.
  • Merchandise Division #2: 1 = Yes, 0 = No.
  • Merchandise Division #3: 1 = Yes, 0 = No.
  • Merchandise Division #4: 1 = Yes, 0 = No.
  • Merchandise Division #5: 1 = Yes, 0 = No.
  • Merchandise Division #6: 1 = Yes, 0 = No.
  • Merchandise Division #7: 1 = Yes, 0 = No.
  • Merchandise Division #8: 1 = Yes, 0 = No.
  • Merchandise Division #9: 1 = Yes, 0 = No.
Ok, that takes care of merchandise. Now, I want to create a series of 1/0 variables for each physical/advertising channel utilized by the company. You can choose to use first-touch, last-touch, sophisticated allocation, whatever methodology you want ... I've found that the methodology doesn't make much difference.
  • Telephone Buyers: 1 = Yes, 0 = No.
  • Website Buyers: 1 = Yes, 0 = No (true catalogers would split this into two variables, orders that are purely online and orders matched-back to a catalog).
  • E-Mail Buyers: 1 = Yes, 0 = No.
  • Search Buyers: 1 = Yes, 0 = No.
  • Social Media Buyers: 1 = Yes, 0 = No.
  • Mobile Buyers: 1 = Yes, 0 = No.
  • Retail Store Buyers: 1 = Yes, 0 = No.
Each channel has been accounted for, each merchandise division has been accounted for, and each KPI (demand, frequency, items, items per order, price per item purchased) has been accounted for. If you don't like what I've come up with, create your own variables, the sky's the limit!

Finally, I want to create columns in the spreadsheet for zip-based variables. These are also 1/0 variables.
  • Customer Lives in Zip Not in a Store Trade Area: 1 = Yes, 0 = No.
  • Customer Lives in Zip Claimed by 1 Store as a Trade Area: 1 = Yes, 0 = No.
  • Customer Lives in Zip Claimed by 2 Stores as a Trade Area: 1= Yes, 0 = No.
  • Customer Lives in Zip Claimed by 3+ Stores as a Trade Area: 1= Yes, 0 = No.
  • Customer Lives In Zip Populated By Great Catalog Shoppers: 1= Yes, 0 = No.
  • Customer Lives In Zip Populated By Average Catalog Shoppers: 1 = Yes, 0 = No.
  • Customer Lives In Zip Populated By Poor Catalog Shoppers: 1 = Yes, 0 = No.
  • Customer Lives In Zip Populated By Great Online Shoppers: 1 = Yes, 0 = No.
  • Customer Lives In Zip Populated By Average Online Shoppers: 1 = Yes, 0 = No.
  • Customer Lives In Zip Populated By Poor Online Shoppers: 1 = Yes, 0 = No.
Finally, I add two variables to the spreadsheet.
  • First Time Buyer In Past Year: 1 = Yes, 0 = No.
  • Multi-Channel Buyer (Phone, All Online Channels, Store): 1 = At Least 2 Of Those Channels, 0 = 1 Or Fewer Of Those Channels.
That's my dataset! The dataset is a spreadsheet with one row per customer, each column representing the attributes mentioned above.

If you are a product manager at Unica, Webtrends, Coremetrics or Omniture, pay attention to this series, as Digital Profiles represent an outstanding way to unlock customer behavioral secrets embedded in your applications.

Up Next: A discussion of the methodology used to create each of the sixteen Digital Profiles.

Items That Appear In Multi-Item Orders

In a typical Life Stage Analysis within a Merchandise Dynamics project, it is common to see exaggerated trends when comparing first-time buy...