May 31, 2015

Holdout Cell Sizes - You Are Likely To Disagree With Me - Again!

Can I let you in on a little secret?

I once worked with a company that executed a lot of mail/holdout tests in direct mail / catalogs. A funny thing happened at this company. First, a ton of Executives didn't like testing, so they told the statistical guru that he couldn't choose more than 10,000 customers for a holdout group. Second, there was a lot of variance in the mail/holdout groups ... customers could spend $25 or $0 or $1,250.

The statistical guru was caught between opposing sides of a cliff wall. The statistical guru should have stopped executing tests altogether given the bind he was in.

He moved forward anyway.

Because of the small holdout segment, the statistical guru could not prove that the mailed segment outperformed the holdout segment. The guru concluded that there was no need to continue to mail the catalog or direct mail piece, since he couldn't statistically prove that it generated sales.

Had he been allowed to have 50,000 customers in his holdout segment, he would have proved that the mailing generated several million dollars of sales and a half-million in profit.

But he concluded that the mailing should be shut down.

This is what can happen if you let the data-driven crowd drive the car.

At Nordstrom, and this is way back in 2003, I instituted a rule ... we would make sure that our holdout groups would not be smaller than 50,000 customers ... and would be 10% of the mailed quantity where applicable. In other words, if we were mailing 2,000,000 customers, the holdout group would be 200,000. 

Yes, 200,000. 

Oh, I know, I can hear the vendor thought leadership (#vtl) crowd, screaming all the way from Chicago or New England, thousands of miles away. "He's an idiot!". "Don't listen to him." "You don't need a holdout group, our machine learning algorithms can figure out everything for you, please pay us $100,000."

Take a moment, take a breath, and continue reading.

Because I had a large holdout group, I could learn more about the business than anybody competing with me. You like being smarter than everybody else in the industry, don't you?

I could measure item level performance in many cases. I knew if a top-selling item, AN ITEM, belonged in the catalog or not. If the top-selling item generated 1,000 units in the mailed group and 1,000 units in the holdout group, it was a top-selling item that had no place in the catalog. Yes, this happens. All the time!

I could always measure category-level performance. I knew that 24 pages of Mens generated no incremental sales, that the Womens pages drove customers online/in-store to buy Mens. With this knowledge, we pulled Mens from subsequent mailings. Tell me how your machine learning algorithm could figure that out without appropriate tests in place?

I could measure sub-segments. I knew the incremental gain online, and in stores, for recent buyers, frequent buyers, lapsed buyers, customers who lived 77 miles from a store, customers who lived 7 miles from a store, customers who lived in Denver, customers who lived in Dover. Now, the sample sizes started to get small ... so when you have small sample sizes, you measure response (0, 1, 0, 0, 1) instead of measuring sales (0, $184, 0, 0, $396), greatly reducing variance. Got that?

Channel differences. You learn that catalogs have minimal/no impact on pure retail customers. You figure out other ways to build a relationship with retail shoppers. You learn that catalogs have marginal impact online. You learn that catalogs mean everything to rural customers. You learn what the organic percentage actually is. Do you know what your organic percentage is, by channel?

I had co-workers, Executives, who wanted my hide. One plucky Executive visited the Nordstrom Family, and told them that I was "breaking the rules". She asked the family to bring the hammer down on me.

We executed holdout groups for email marketing as well. Today, if I speak with 100 email marketers, 96 do not execute email marketing holdout groups. In other words, 96 out of 100 marketers truly have no idea what the incremental value of email marketing is. How could you ever know how much traffic email drove into a retail store without ample holdout groups? Be honest!

Catalogers do not like holdout tests. Catalogers like to mail stuff.

Email marketers almost universally detest holdout samples - heck, I've had email vendors tell my clients to stop working with me. I know I am on to something interesting and useful when vendors tell my clients that my clients should not work with me.

Executives do not like holdout tests ("how could you cost our business top-line sales, you idiot?").

Attribution experts should demand a healthy supply of holdout tests ... but for some reason, they don't always advocate the tests, trusting flawed models instead.

But you, you are smart, or you wouldn't be reading this. You know you need to be executing holdout tests. And you know you need to have large holdout cell sizes.

Without holdout tests, you're making countless ad investment mistakes.

Without large enough holdout panel cell sizes, you're also making ad investment mistakes.

Without large enough holdout panel cell sizes, you're not learning anything. Don't you want to learn? Don't you want to be smart? Don't you want to be an expert in your field? Don't you want to be eight steps ahead of professionals you compete against?

With appropriate holdout cell sizes, you have the knowledge necessary to push your business forward.

Ok, your turn. If you disagree, use the comments section to defend your case. Show math that defends your case.

May 28, 2015

A Potpourri of Information

Click on this article from mid-January ... a data-driven argument about why a 19-20 Cleveland basketball squad would not win an NBA title (they still may not) and why they may not make the playoffs (they did make the playoffs and won three series, easily). One of the shortcomings of the data-driven movement is the inability of so many (me included) to forecast properly. Just because somebody can (artfully) analyze the past does not mean they can tell you what will happen as a consequence.

I am not the only person who gets fed up with vendor nonsense (click here).

"Loss of the Middle" ... go read point number six. Point number is in the process of hurting catalogers. In the old days, there would be 50 great catalog twelve month, 500 middle catalog twelve-month, and 1,000 low-value catalog twelve-month buyers. Today, those numbers are more like 50 / 150 / 1,000. Not surprisingly, much of the profit goes out the window when that happens, requiring catalogers to discount more, sending even more profit out the window (because the discounting never recaptures the middle). Run the data on your twelve-month buyer file, comparing 2000 to 2015. You're not going to like what you see.

Vox Media acquires ReCode (click here) ... you're seeing this trend everywhere you look ... without the power of a large brand, it is very hard to make it on your own. Smaller businesses are being rolled-up by larger ones, as it has always been. In our own little catalog world, you've witnessed the rolling-up of the consultant industry over the past eighteen months. Either you have a brand-name and resources and cross-selling opportunity, or you do something so unusual, different, and interesting, that you attract a profitable niche.

A report from RJMetrics (you may recognize a contributor) suggests that it is very, very difficult in e-commerce to move a customer from a first to a second purchase. As new customer audiences begin to dry up in e-commerce, you're going to see more folks come to this realization ... and that's going to fundamentally transform e-commerce. Countless dollars will be wasted trying unsuccessfully to get e-commerce buyers to become "more loyal". You'll see digital brands embrace offline marketing (ask Airbnb, Google, Wayfair, countless others), out of necessity. Catalogers will mis-read this as a sign that catalogs are back (they will be used more often, they won't come back, though I always volunteer that I could be wrong).

Speaking of ecosystems ... we're going to see more and more people using more an more methods to explain how somebody/something fits into an ecosystem ... like in this case, Android (click here).

A 60% reduction in PPC expense (click here).

Chip Kelly - "that's not the way things are done here". Do not get discouraged when vendors or pundits or experts tell you that you are wrong. Chart your own path.

May 27, 2015

Online Visitation Evolves Over Time

When I am analyzing an online ecosystem, I create a series of logistic regression models that predict next-month visitation within online advertising channels.

What is interesting about the equations is that customers exhibit behavior that is far more consistent than what you're taught to believe. You're taught that customers "do everything". Once you are taught that customers "do everything", you're taught that you must use attribution algorithms to parse out the activity that leads to a purchase.

In a recent project, more than half of visitors had just one daily interaction with the website. When predicting what should happen the next month, the models indicated that customers are far more likely to stay within the advertising channel they were in the month prior ... in other words, when a customer visits via an affiliate last month, we can expect that customer to visit via direct load next month, with a scattering of visits across all other advertising channels.

The projects consistently point out three things.
  1. Direct Load visitation is most important, and obviously, most cost effective. Most visitors, even toll-based visitors, migrate back to direct load over time.
  2. Push-based visitation (email) is second most important - and obviously.
  3. Toll-based visitation (catalogs, affiliates, remarketing, Facebook, search/Google) matters, and ultimately pushes a customer to direct load over time.
The projects also point out a consistent trend, over time.
  • Early customer life cycle visits are skewed to toll-based channels.
  • Later customer life cycle visits are skewed to push-based (email) and direct load (customer loyalty) visits.
The projects also illustrate that discount/promo centric companies drive more toll-based traffic ... imposing a double-toll on the business (pay an affiliate, generate less profit because of the discount/promo).

The interesting thing about all this stuff is that the vast majority of what we read in trade journals or from vendors focuses on toll-based traffic that converts via discounts/promotions. The vast majority of quality profit we generate comes from direct load (customer loyalty) and email push-based visits.

Yes vendors, I hear you yelling at me from afar ... "you're wrong, and I'm going to tell you why". Go ahead and run the monthly ecosystem models I'm running, and see if you can see something similar in your information.

Contact me (kevinh@minethatdata.com) for your own, customized, online ecosystem analysis.


May 26, 2015

Ad Blockers

You've been reading about ad blockers, right (click here)? Here's a popular version for Chrome ... with downloads allegedly exceeding 200,000,000 (click here).

In my projects, there are three types of traffic.

  1. Earned Traffic (i.e. customer visits your site because the customer loves your brand).
  2. Pushed Traffic (i.e. catalogs, email campaigns).
  3. Toll-Based Traffic (i.e. retargeting, Facebook, Google, YouTube, all the stuff you love to focus on).
It takes a tremendous amount of talent, effort, and merchandising/marketing brilliance to achieve earned traffic. The outcome is a phenomenal amount of high-quality profit.

It takes a tremendous amount of talent, effort, and merchandising/marketing brilliance to generate sales via Pushed Traffic. The outcome is a large amount of sales and a fair amount of low-quality profit (in the case of cataloging), or a smaller amount of sales paired with high-quality profit (in the case of email marketing). Either way, something gives.

It does not take a tremendous amount of talent, effort, and merchandising/marketing brilliance to generate Toll-Based traffic. It takes money and some amount of technical chops. The outcome is frequently a fair amount of sales paired with low-quality profit.

Almost everything you read demands a focus on Toll-Based Traffic generation. Many call this "omnichannel". Vendors, trade journalists, and consultants love omnichannel Toll-Based Traffic, because they get paid when you listen to them.

Now, imagine a world where 65% of your customers are using ad-blocking software. Can we agree that, in a world where the majority of customers are using ad-blocking software, it becomes much, much harder to grow your business via Toll-Based Traffic?

In that world, the omnichannel theorists correctly assess that your customer is using many channels ... but your efforts to have a relationship with the customer outside of your site are increasingly thwarted. What do you do? Be honest with yourself.

Success in five years comes from efforts to build Earned Traffic today. It's probably time to get busy, don't you think?

May 25, 2015

Marketers Influence Merchandise Ecosystems

Some marketers raise a champagne glass every time somebody engages with brand, or every time 55% off plus free shipping results in an incremental order.

Maybe we should celebrate when we cause positive downstream change.

Recall our example? Category 03 negatively influences the entire ecosystem, whereas other categories positively influence the entire ecosystem.

So let's run a simulation. In my simulation (won't get into the geeky math here in this post, there's no easier way to lose an audience than to focus on math), I increased annual demand in Category 03 by 20%. Here's how many of last year's customers are forecast to shop next year, based on this change to the ecosystem.
  • Category 00 = 62 fewer.
  • Category 01 = no change.
  • Category 02 = no change.
  • Category 03 = 1,522 more.
  • Category 04 = 168 more.
  • Category 05 = 220 fewer.
  • Category 06 = no change.
  • Category 07 = 29 fewer.
  • Category 08 = 58 fewer.
  • Category 09 = 221 fewer.
  • Category 10 = 45 fewer.
  • Category 11 = 97 fewer.
  • Category 12 = 71 fewer.
Essentially, half of the customer gains in Category 03 are wiped out by customers lost in Categories 00, 05, 07, 08, 10, 11, and 12.

Now, look at what happens if we increase Categories 04/05 by 20% (these categories, in total, are about the same size as Category 03).
  • Category 00 = 76 more.
  • Category 01 = 62 more.
  • Category 02 = 7 more.
  • Category 03 = 350 more.
  • Category 04 = 1,125 more.
  • Category 05 = 560 more.
  • Category 06 = no change.
  • Category 07 = 38 more.
  • Category 08 = 33 more.
  • Category 09 = 338 more.
  • Category 10 = 57 more.
  • Category 11 = 205 more.
  • Category 12 = 198 more.
Here, all other categories benefit as much as 04/05 did, downstream. Everybody wins.

Ok, you are a marketer, and you now know how your ecosystem works. What do you do about it?
  • Be willing to pay more to acquire customers outside of Category 03.
  • If you have a catalog, feature Categories 04/05 at the front of the catalog.
  • Feature Categories 04/05 more often on your home page and landing pages.
  • Feature Categories 04/05 more often in email campaigns.
  • If you must offer discounts via promotions, do so more often in Categories 04/05.
  • Pay Google more for Category 04/05 keywords.
  • Feature Categories 04/05 more often on Pinterest.
  • Try something!
Do you know understand why it is so important to gain knowledge of your merchandise ecosystem?

May 24, 2015

#BestPractices For Winning The Indianapolis 500

In the spirit of all of those marketing articles you click on and then wish you hadn't, here is a list of best practices for winning the Indianapolis 500.


By Vic Montana, Industry Marketing Consultant

The Indianapolis 500 is one of the most prestigious races to win. Steeped in tradition, many men and woman devote their lives in an effort to drink milk in the Winners Circle.

Unfortunately, most race teams make mistakes, and those mistakes cost the race teams the opportunity to win. I have compiled a list of best practices that, if you follow them, will give you a better chance of obtaining a coveted victory.


Obtain Sponsorship. Racing is expensive. If you can convince a major brand to part with ten or fifteen million dollars of money, you will be able to invest in better equipment. Watch television commercials. The brands that spend the most on television are the brands that may be willing to give you money so that you can pursue your dream. This isn't a loan. You will not have to pay back the funds given to you. You may have to sit in a hospitality suit during a golf tournament and consume crostini with company Executives in exchange for a check with seven zeroes on it, but that is a small price to pay for a chance to win a race. Think of sponsorship as Kickstarter for racing.

Take A ClassLearn from the experts. Take a twelve-week "crash course", and build your skills on ovals, road courses, and city circuits. You'll be "up to speed" once you immerse yourself in the racing culture for a few months.

Qualify Fast. Cars that start the race at the front of the field have a much better chance of winning the race than cars that are slow. Be fast!

Where Possible, Pass Cars. In order to win, you have to run at the front of the field. Cars that are not in the top five during the race have a lower chance of winning than cars that consistently run at the front of the pack. If you find yourself running in 17th place, try harder. Be excellent.

Use Pit Stops To Your Advantage. Eventually, you will need to fill your gas tank. Much like pulling your Corolla into the Shell station for fifteen gallons of 87 octane unleaded fuel, your car requires frequent refueling of 85 octane ethanol-infused gasoline to get to the finish line. While refueling, take the opportunity to change tires. Adjust a wing! Make sure you have the best pit crew in the business. But don't pay the pit crew a lot of money, because every dollar spent on the pit crew is a dollar you cannot spend improving your equipment.

Engage With Your Fans. During cautions, you are only driving fifty miles per hour. This gives you a great opportunity to interact with your fan base. Fans are going to have a lot of questions. "Are you going to win the race?" "What is it like to drive for a beloved brand like Verizon?" "Are you hungry?"  So feel free to tweet a handful of messages during cautions. Say something like "We're loose in turns three and four and have too much drag down the straightaway" or "I can feel your energy, keep cheering for me!" Connect your cell phone to the rear wing and take a Periscope or Meerkat of the race - your fans will really appreciate an unfiltered perspective of the event. You may not win the Indy 500, but you can win the Engagement Race with your fans!

Don't Quit. Just because you came out of the pits in eleventh place with twenty laps to go, don't get down on yourself. Quitters never win, and winners never quit.

Use Your Rear View Mirror. Many drivers will attempt to pass you during the race. You do not want to fall further behind. Use your rear view mirror to thwart off the competition.

Have An Omnichannel Strategy. To win today, you need to be successful in all channels. Complement your on-track success with a sound digital strategy. Make sure your fan site is mobile friendly, so that you can win the digital race as well.

Pump Your Fists When You Win! As you cross the start-finish line, thrust your left fist in the air, while steering with your right hand at 190 miles per hour. Thank your spotter, pit crew, and owner. Tweet something to your loyal fans - always put the customer at the center of this magical experience. You are a customer-centric brand driver.

2nd and 3rd Place Pay Well, Too. Only one driver out of thirty-three can win the Indy 500. However, second and third place pay well, and offer the driver the opportunity to stand on "the podium", giving your sponsor considerable marketing exposure. If you are lucky enough to be sponsored by Pepsi or Coke, be sure to have a can of cola in your hand, and take a long drink from the can of cola before answering any questions from the "gotcha media" who roam pit road in search of a larger narrative.

The Next Race Is Only A Week Or Two Away. In the unlikely event that you do not achieve the greatest honor of your career, do not despair. The racing calendar offers nearly twenty opportunities to earn a hard fought victory. #KeepTrying

If you follow my best practices, you, too, are likely to achieve your dreams of winning the Indianapolis 500. #MakeItHappen!

May 21, 2015

Gliebers Dresses: The Prawn Tower

As you know, Gliebers Dresses is a fictional story about a Catalog Executive Team. If this is not your cup of tea, why not do a Google Search to learn about how Adobe is hiring actors to read white papers in podcast form (or click here, your choice).


Welcome to the Gliebers Dresses Executive Conference Room.

Glenn Glieber (Owner, CEO): Ah, Memorial Day weekend. The memories. I hope you all have plans. Remember, your staff are free to leave at 3:30pm today.

Meredith Thompson (Chief Merchandising Officer): I already sent my team home.

Glenn Glieber: It's nine in the morning.

Lois Gladstone (Chief Financial Officer): It's a merchandising best practice, Glenn.

Glenn Glieber: Oh.

Pepper Morgan Pressley (Chief Marketing Officer): Did you see what Gwen Wilson wrote on her blog this morning?

Meredith Thompson: I get her posts in my in-box every morning. Fascinating stuff.

Roger Morgan (Chief Operations Officer): She's an idiot blogger.

Pepper Morgan Pressley: She's what?

Roger Morgan: An idiot blogger.

Pepper Morgan Pressley: She was an Executive for a decade at our biggest competitor. I think she knows something about our industry.

Roger Morgan: Please.

Lois Gladstone: Please tell me this doesn't have anything to do with Woodside Research.

Roger Morgan: It has everything to do with Woodside Research.

Meredith Thompson: I'm leaving.

Glenn Glieber: You're what?

Meredith Thompson: Have a nice weekend, everybody. (Meredith collects her papers, stands up, and walks out of the conference room).

Glenn Glieber: She can't do that.

Lois Gladstone: She excused her staff early, so she's just adhering to her own holiday weekend departure best practice.

Glenn Glieber: I should get HR involved.

Lois Gladstone: HR reports to Roger. Ask Roger to stop Meredith.

Glenn Glieber: Roger?

Roger Morgan: So I was at the Woodside Research Aspire Conference last week.

Lois Gladstone: The what?

Roger Morgan: The Woodside Research Aspire Conference. I'm sure you followed along on Twitter. Hashtag Aspire.

Lois Gladstone: Can I leave, Glenn?

Glenn Glieber: Shhhhh. I want to hear about this hashtag thing.

Roger Morgan: Woodside Research collected every thought leader in the industry, and put them on the stage to share their vision of the future.

Lois Gladstone: How does Woodside Research decide who gets to be on the stage? Merit? Profit generated for clients? Rampant client growth?

Pepper Morgan Pressley: You pay Woodside Research $25,000. Then you get to be on the stage.

Roger Morgan: The fee keeps the idiot bloggers off the stage.

Pepper Morgan Pressley: Gwen Wilson is not an idiot blogger.

Roger Morgan: The thought leaders at the conference disagree. 

Lois Gladstone: Thought leaders?

Roger Morgan: You know, the smartest people in the industry. They share their thoughts with the audience. They try to stretch the minds of brilliant attendees. People like me, for instance. It's an invigorating way to spend a day in the Bay Area, I'll tell you that much.

Lois Gladstone: If I wanted somebody to lead my thoughts, I'd join a cult.

Pepper Morgan Pressley: Thought leaders don't actually lead. They tell people what to do, then they stand far away while their ideas fail in the real world.

Roger Morgan: Do you know how exciting it is to rub shoulders with the smartest people in the industry?

Pepper Morgan Pressley: We spend every day with you, Roger. It's like a little slice of heaven.

Roger Morgan (looking curiously at Pepper): For instance, I was standing next to the Prawn Tower, chatting with Nate Carson.

Pepper Morgan Pressley: You were what?

Roger Morgan: Chatting with Nate Carson.

Pepper Morgan Pressley: The other thing.

Roger Morgan: They had this eight foot tall pyramid of prawns. Oh my goodness. You just stuck a toothpick into a prawn, put it on your plate, dipped the prawn in cocktail sauce, savored the flavor, then repeat. I'll bet it took the two thousand attendees a good four hours to take down that Prawn Tower.

Glenn Glieber: Was it anything like the vat of popcorn shrimp at the Golden Corral? Because let me tell you, it doesn't matter how many vans of Sr. Citizens they bring in there, they're not going to get to the bottom of that vat!

Roger Morgan: Then there was the chocolate waterfall. They made it look like Niagara Falls!

Lois Gladstone: Did they send any of the thought leaders over the falls in a marshmallow raft?

Roger Morgan: I'm trying to make a point. When you go to an event like this, when you rub shoulders with people like Nate Carson, when you pay an additional $750 to sit at a breakfast roundtable sponsored by Omicron and you chat with thought leaders about engagement and personalization and relevancy in the modern era, you come to realize just how special Woodside Research is.

Lois Gladstone: We spent $2,750 of Glenn's money, plus travel expenses, so that you could rub shoulders with thought leaders and eat prawns?

Glenn Glieber: He spent my money?

Roger Morgan: You bean counters, you're all the same. In fact, one of the thought leaders gave a three minute lightning lecture ...

Pepper Morgan Pressley: A lightning lecture?

Roger Morgan: You know, one of those quick-fire sessions where thought leaders rapidly offer their guidance in a progressive and fast paced setting.

Pepper Morgan Pressley: Ugh.

Roger Morgan: Anyway, he gave a lightning lecture titled "Bake the Beans". He gave five quick tips for disarming your Chief Financial Officer, who he labeled, "the biggest stumbling block to omnichannel success in the modern enterprise." Did you know that, according to a Woodside Research report, 77% of marketers are fed up with the way the CFO limits the ability of the marketing department to do their job properly?

Lois Gladstone: I'll bet 100% of CFOs are fed up with employees wasting company resources to attend marketing nerd prom.

Roger Morgan: She feeds right into the narrative, doesn't she?

Pepper Morgan Pressley: Are we about done here?

Roger Morgan: I haven't even had a chance to share facts about the cataloger happy hour.

Pepper Morgan Pressley: Catalogers have a reason to be happy?

Roger Morgan: Bill Blatt from Hex Graphics shared a best practice case study about a large retailer mailing one catalog, just one, back in April. He says that one catalog is a symbol of the potential re-birth of an industry. We're really well positioned for the future, when you think about it.

Lois Gladstone: You've been telling us for the past six years to mail fewer catalogs. Now you're telling us we're well positioned?

Roger Morgan: If a printer tells you that paper is making a comeback, you have to believe the printer has your best interests at heart, right? Why would a vendor lie to you? It's like when social media experts tell you to focus on engagement, or when email vendors demand that you focus on personalization. They're leading your thoughts for a reason!

Lois Gladstone: They sure are. I'm leaving.

Pepper Morgan Pressley: Me too.

Roger Morgan: The two of you don't want to participate in the future, do you?

Glenn Glieber: Hold on, you two, let's hear Roger's vision for the future.

Roger Morgan: Really?

Glenn Glieber: Go.Tell us what our vision is.

Roger starts thumbing through papers on his desk, looking frantically at various reports from Woodside Research.

Glenn Glieber: I don't want to hear what somebody else thinks about the future. I want to hear your vision of the future.

Lois Gladstone: He's not capable of saying anything unless a vendor or a research brand first authored the sentence.

Roger continues to look at various documents. He is sweating. His cheeks are red. He has red blotches on his neck.

Pepper Morgan Pressley: Alright, it's been fun. I'm going to leave and join Meredith. Have a nice weekend everybody.

Lois Gladstone: I'm outta here, too.

Glenn Glieber: We haven't heard Roger's vision yet.

Roger Morgan: I think I left the Woodside Research report on marketing in 2025, sponsored by Omicron, on my desk. I'm sorry. I'll bring it next time.

May 20, 2015

When One Category Works Against Most Of Your Categories

Sometimes, you have a merchandise category that works against the rest of the business. It happens more often than you'd think.

When I worked at Nordstrom, we had Cosmetics, Mens Footwear, and Womens Footwear. These categories helped all other categories - meaning that when a customer purchased Cosmetics, the customer instantly became more likely to buy from Accessories, for instance. But the opposite wasn't necessarily true. Growing Accessories helped Accessories. Growing Cosmetics helped Cosmetics and helped Accessories.

In my Nordstrom example, which category should be featured in marketing, on landing pages, in email campaigns? Which keywords would you pay more for on Google? Cosmetics, right? Cosmetics helps Cosmetics and Accessories. Accessories only helped Accessories.

Your business is a big, bubbling ecosystem. You've been trained to ignore the ecosystem. You've been trained by vendors to focus on campaigns, to focus on how campaigns convert.

Look at Category 03 in the table above. Are the coefficients positive, or negative? Positive coefficients mean that dollars spent in that category today will help other categories in the future. Negative coefficients meant hat dollars spent in that category today hurt response in other categories next year.

Category 03 hurts nearly every other category in the next year. When a customer buys from Category 03, the customer will spend less next year in the following categories ... 00, 05, 07, 08, 09, 10, 11, and 12.

In other words, the merchant responsible for growing Category 03 is the same merchant who hurts the future performance of the merchants responsible for categories 00, 05, 07, 08, 09, 10, 11, and 12. And the marketer who actively promotes Category 03 is the marketer who hurts his/her efforts to properly market categories 00, 05, 07, 08, 09, 10, 11, and 12.

Read across the row for Category 11. This category doesn't harm the future performance of a single category. The merchant responsible for Category 11 helps the entire business.

These relationships happen in your business too. Absolutely! The best marketers know these relationships. They know that Category 03 is hurting the future potential of the rest of the business. They know that when they have a choice, they give Category 11 marketing priority over Category 03. They protect the future of their business by not promoting one category in a way that harms other categories.

Show of hands ... how many of you thoroughly understand how your merchandising categories interact with each other?

May 19, 2015

David Letterman

The guy was throwing watermelons off of a building. That's the kind of thing that was interesting to a seventeen year old aspiring to become a statistician. Mr. Letterman had re-written the rules of late night television.

Around the time Larry Bud Melman delivered periodic overnight comedic thrills, companies like Lands' End and L.L. Bean were re-writing the rules of cataloging. Prior to 1980, "big books" dominated the scene ... JCP, Wards, Spiegel. Five hundred pages, delivered a few times a year. Their model was upset by 124 pages mailed twelve times a year ... customers had not seen that before, content especially for them, targeted merchandise, a limited assortment (curated). Customers responded, and specialty catalogs ascended the mountain.

By the mid-1990s, Mr. Letterman was at the height of his game, competing against Mr. Leno at 11:35pm (10:35pm Central and Mountain). However, we couldn't see the future. We couldn't see how unlimited cable channels and this thing called "the internet" would fragment the audience, allowing dozens of late night hosts to have their own audience - we couldn't possibly see how somebody would create content at 12:35am so that it would go viral on YouTube a day or two later.

By the mid-1990s, catalog companies were at the height of their game. However, catalogers could not see the future, either. How could we? We couldn't see how unlimited commerce websites on this thing called "the internet" would fragment the audience (some probably did), allowing thousands of business leaders to have their own customer audience - we couldn't possibly see how a company would generate no profit for nearly twenty years and yet would be on a trajectory toward a trillion dollars in net sales by 2025 by finding a way to cope with an ever-fragmented audience (think about that one for a moment - the audience fragments, so one company finds a way to appeal to everybody - kind of like selling via a 500 page catalog without the onerous variable costs).

If you watch Mr. Letterman today, you can barely recognize what existed in the 1980s. The "World's Most Dangerous Band" was replaced more than twenty years ago by "The CBS Orchestra". No more watermelons. Still a top ten list. The spirit lives, the old-school silliness is gone. The audience aged. The 30 year old watching the show in 1985 is a 60 year old watching the show today. Younger folks don't bother. It became hard to fight against a nearly infinite number of channels.

If you read through a catalog today, you can barely recognize what existed in the 1980s. The entrepreneurial spirit has been replaced by best practices. Still 124 pages. The spirit lives, the old-school innovation was gone. The audience aged. The 30 year old purchasing from Lands' End in 1985 is a 60 year old purchasing today. Younger folks don't bother. It became hard to fight against a nearly infinite number of (omni)channels.

Mr. Letterman either quit or was forced out. He will be replaced, but the brand name goes on. Just ask Jimmy Fallon ... he hosts a show that has been on more than sixty years. The merchandise changes ... the brand lives ... the audience gets younger each time. We recycle.

You're being taught that you have to offer your merchandise in more channels, that customers love your merchandise but you need to have it everywhere, at all times. 

Do you want David Letterman on all devices at all times?

Maybe we need new merchandise?

Just this week, I analyzed a project for a company that makes full use of the co-ops. The merchandise purchased by co-op customers was fundamentally different than the merchandise purchased by customers acquired via online marketing. Run an ecosystem simulation for your business - or ask your favorite vendor to run an ecosystem simulation for you - you may well see the same thing. The co-op customer, in television terms, loved David Letterman. The online marketing customer, in television terms, loves Stephen Colbert. By maintaining strong loyalty to the co-ops, this cataloger sold merchandise in the theme of David Letterman. The cataloger is trapped - the cataloger must subsequently offer merchandise loved by the long-term core customer.

But a broad, diversified customer acquisition portfolio gives life to your merchandise assortment. You'll have enough new customers who like a different set of merchandise, causing your merchandising team to respond differently, to evolve the assortment in a manner that helps your business, long-term.

Television (a medium that may be dying or may be growing stronger, who knows) evolves the merchandise assortment on a non-stop basis, often failing, sometimes succeeding spectacularly.

We catalogers need to evolve as well. We need to evolve our programming (merchandise). Merchandise, not channels, attracts a profitable audience.

Oh, I know. You'll see Mr. Letterman do a piece on CBS Sunday Morning in six months, and the old-school folks will roar with enthusiasm ... "He's back, this proves his style of humor isn't dead!!" ... much in the same way the trade journalists and consultants roar with approval when JCP decides to mail one (1) catalog.

I will miss Mr. Letterman. I already miss Mr. Ferguson. And I will miss Mr. Stewart, too, in a few months. The merchandise assortment changes - but their shows go on. I change, I go on. We can change our merchandise offering as well.

Merchandise Category Interaction

You've enjoyed an annual physical, right?

You learn that your cholesterol is too high (too many nachos), so you are prescribed a drug. Before the drug is prescribed, however, your doctor determines if the drug will interact with other drugs you are taking.

You don't want drugs interacting negatively, do you?

The same issue happens in your business. But we're not trained to measure interactions. The vendor community prescribes solutions. What if their solutions cause negative interactions within your business?

I assure you, the solutions some of those folks prescribe for you yield negative and unfavorable interactions.

The table above is a geeky table of logistic regression coefficients for a business with thirteen merchandise categories (categories 00, 01, 02, 03, 04, 05, 06, 07, 08, 09, 10, 11, 12). The dependent variable is a 1/0 indicator, measuring whether a customer purchased from that category in the past year. The independent variables are demand ($/1000) amounts for each category for the period 13-24 months ago (this is the point in the discussion where some in the vendor community will tell you I'm nuts - remember, they are prescribing drugs, we are diagnosing problems, two very different disciplines).

Read down the "Category 05" column. Here, I am measuring the merchandise categories that positively/negatively interact with future purchasing in Category 05 (yes, these are actual equations from an actual business ... I just omitted what the actual category meanings/definitions are). What do you see?
  • Category 00 interacts positively.
  • Category 03 interacts negatively.
  • Category 04 interacts positively.
  • Category 05 interacts positively (duh).
  • Category 06 interacts positively.
  • Category 07 interacts positively.
  • Category 08 interacts positively.
  • Category 08 interacts positively.
  • Category 10 interacts positively.
  • Category 11 interacts positively.
  • Category 12 interacts positively.
This is one, big, bubbly ecosystem, don't you think? Only two of the thirteen categories (01, 02) don't interact with this category. Ten of the interactions are positive. One interaction (03) is negative.

This category goes as the entire business goes. When a customer purchases from Category 08 today, this category gets benefit tomorrow.

Let's preview something for tomorrow. Read across the row for Category 03. What do you see? Given what I've explained so far, what does this mean to you?

May 18, 2015

You Are A Doctor, Not A Pharmacist

The modern theme of marketing misses out on all of the magic and glory bubbling within your merchandising ecosystem. You are asked to market in all channels (#omnichannel), and you are asked to pay tolls (Google, Facebook, Abacus) in an effort to get customers to purchase.

In other words, the industry wants to help you prescribe a drug ... the industry wants you to count pills and put the pills in a bottle and then communicate side effects (omnichannel sales gains).

Your job, quite honestly, should be the opposite. You should be a doctor. You should diagnose problems, and then you, yes you, should prescribe the solution.

Do not be a pill counter. (in the medical world, by all means, I fully encourage you to be a pharmacist - I'm talking about marketing, not medicine).

We don't have enough doctors in our industry.

Let me give you an example. Let's go back to 1998. Yes, I'm at Eddie Bauer, Director of Circulation, fingers in every pie. The Home business was awful. Those folks always lost money. And those folks somehow kept their jobs in the process ... posting -10% earnings before taxes year after year. But now that I'm the circulation director, I'm accountable for that -10% as well. I don't want to be responsible for posting -10% earnings before taxes. 

Bad business is not going to be my fault.

*** Here is where the vendor community will tell you that I'm using math incorrectly - patiently listen to their complaints, then get busy implementing solutions ***

I created a series of logistic regression models. The independent variables were demand spent by twelve-month buyers, by merchandising category, in 1996. The dependent variables were 1/0 indicators ... did the customer purchase from a series of merchandise categories in 1997? Each logistic regression equation featured prediction of a 1/0 buy/no-buy variable regressed against demand spent the year prior in each merchandise category.

Have you ever run these equations? The coefficients reveal the secrets embedded within your business!! The coefficients tell you why some categories never grow, while other categories grow by default.

In the case of the Home division - the coefficients told a very interesting story.
  1. Existing Eddie Bauer customers did not want to buy Home product.
  2. Home customers, after buying Home product, did not want to buy Home product. Home customers wanted to buy Apparel next.
When you see (1) and (2), what would you do to grow the home business, knowing these two facts?

#Omnichannel? 


Heck no.

Here's what I did:
  • I rented every single Williams Sonoma and Pottery Barn name I possibly could. Often. All of them.
In fact, half of my circulation plan (it took WS/PB two years to figure out what was happening) consisted of Williams Sonoma and Pottery Barn names. 

Oh, I caught heck within Eddie Bauer for doing this. "Why aren't you mailing our own customers, you idiot?" Well, I'm not mailing them because our own customers won't buy the product, that's why!!

Not surprisingly, the business went from losing a couple million dollars a year to losing a hundred thousand dollars within a year. The business wasn't fixed - but the business was no longer floundering into oblivion, either.

Do you understand, dear marketers, that you have power? You - yes you - can fix major problems on your own, if you simply act as a doctor and diagnose a problem instead of being a pharmacist who counts omnichannel pills.

Tomorrow, I'll share more information on the methodology I used at Eddie Bauer, way back in 1998. Vendors will tell you that I'm using math incorrectly. I'm here to tell you that I work in the real world, and the technique works very, very well in the real world.

May 17, 2015

The Little Red Campfire

It's not warm at night on the west side of the Cascade Mountains. You don't just sit outside and sweat. In fact, you probably need a sweatshirt or two if you're going to sit outside for a few hours.

Unless you have the Little Red Campfire.

This little thing cranks out a whopping 65,000 BTU's of propane-infused thunder. Within ninety seconds, it is as if you are sitting outside in Topeka in June ... but without the bugs.

Anytime I sit somebody down in front of this thing, I hear this:
  • "I have to have one!"
Do you understand the gravity of that sentence? You know you're a successful Digital Merchant when customers "have to have one".

Let's view the purchase process from the eyes of a customer who "has to have one".

Start with Google. 


We compare two companies ... Amazon ... and Camping World.
Camping World doesn't look too promising, based on price ... but the unproven omnichannel thesis suggests that a customer will ignore a higher price because of the myriad benefits of a digital relationship with a store-based brand. So let's follow some in the vendor/consultant community down the rat hole.

I start by filling out my shipping address ... I hit a wrong key while trying to get Google to auto-fill information for me, and I ended up in another tab. When I get out of that tab, I am greeted with the following message:


Because I was a complete idiot and hit the wrong sequence of keys, I am rewarded with a five dollar discount! Why do I deserve to pay less, when a customer fully capable of using a computer keyboard ends up paying more? Is this a new best practice?

Camping World won't let me check out unless I tell them what kind of RV I own. Is this a new best practice? Because in the old days, the pundits demanded a frictionless experiences.


Here comes the cash-based reason to add friction.


No thanks (but Kevin, you could have received a $20 merchandise certificate ... #crm #personalization #bounceback).

Not interested in roadside assistance? How about a Camping World Visa with Rewards? Mind you, I have yet to see what shipping is going to cost me.


It is interesting that they place this offer over the shipping and handling window ... because they probably realize that the customer is going to be really disappointed with the cost of shipping and handling.


Oh.

I get standard 10-14 day shipping for free (does it really take 10-14 days to ship an item to my home, when I can pick it up from a store 45 miles away - that's about four miles of transit per day, I sure hope I get a tracking number so that I can actively monitor foot-by-foot progress). However, this is considered a heavy item, so tack on another $10.00.

Total omnichannel value proposition at Camping World? $139.41 with sales tax. #Omnichannel!!

Total cost at Amazon? $120.79 with sales tax. And the item will arrive in two days, no surcharge, no nonsense.


Again, just so we understand what Digital Marketers with an omnichannel preference are asking of the customer:
  • It took me four clicks on Amazon to buy the item for $120.79. I'll have the item in two days.
  • I took a scintillating omnichannel journey at Camping World, one where I was given a $5 discount for being unable to properly utilize a computer keyboard, one where I was offered the chance to purchase Good Sam Roadside Assistance for less than eighty dollars, one where I was offered the opportunity to acquire yet another Visa card, before learning that the net cost of the item would be at least eighteen dollars more than I could procure the same item from Amazon for - and I would enjoy a 10-14 day delivery window unless I was willing hand over even more money.
There's a reason that everything you read tells you that the scintillating omnichannel journey at Camping World is a best practice ... it's simple ... vendors and consultants and trade journalists and researchers all get paid. Only the brand and the merchant supplier get paid in the Amazon example.

Allow me to define two terms for you.
  • Digital Merchant: A marketer who cares deeply about aligning must-have products with great customers/prospects. Once the connection is made, the digital merchant steps out of the relationship, and allows the merchandise to speak for itself.
  • Digital Marketer: A marketer who cares deeply about being a middleman. The Digital Marketer does not believe in the merchandise being sold, but instead, believes in a series of cross-sells, up-sells, cart abandonment offers, discounts, promotions, omnichannel strategy, CRM, cross-brand credit offers, cross-brand partnerships, strategy, attribution algorithms, optimization, personalization, engagement, conversations, sale events, and anything else that shine sun on the Digital Marketer.
I know, I know, I'm going to get angry emails from the industry ... not from those of you working at companies like Camping World, however. Let that one sink in for a moment.

Which purchase process do you, as a customer, prefer? Not as a marketer, but as a customer?
  1. Four clicks via Amazon, resulting in a lower price and product arrival in two days?
  2. The world created by the Digital Marketer at Camping World, a world where building a relationship with Visa is prioritized as more important than the communication of shipping costs to the customer, a process that yields a higher price and glacial shipping?
The process you prefer dictates whether you are a Digital Merchant or a Digital Marketer.

Now, be honest - which process does you company employ?

Are you a Digital Merchant, or a Digital Marketer? There's a big difference.

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...