September 30, 2013

New Item Failure

Folks, this story repeats more often than old episodes of M*A*S*H.

Take a look at this table. This is the outcome of a comp segment analysis on an e-commerce business. The comp segment analysis evaluates customers who, for instance, purchased two times in the year ending September 30, 2012. We then measure how much these customers spent in the year ending September 30, 2013. I've found that this style of analysis is the most reliable way to determine if a business is generating improved merchandise productivity or if merchandising issues are hurting the business.

Guess what? This business is being hurt by merchandising issues.

Comp segment demand was -3.9% last year, and was -2.8% in the past twelve months. This business is suffering from two consecutive years of merchandise failure.

Now look down the Existing Comp column. This column represents comp performance for existing items. For each of the past five years, existing items have remained flat, or have marginally improved in value.

Now look down the New Comp column. Oh boy. Ohhhhhh Boy.

The entire reason that the business is struggling is because new item comps were -26.5% last year, and are -25.2% in the past twelve months. In fact, new item performance is down more than 44% over a two year period of time. 

That's a catastrophe, folks.

Look at the 2007/2008 timeframe. Both new and existing items were down, suggesting that the economy may have been to blame for poor performance.

But in 2012/2013? It's new items.

And guess what new items in 2012/2013 become in 2014/2015? Yup. Existing items. We can confidently project that, without an infusion of highly productive new items, existing item performance is about to go in the tank as well.

That's what we do with a Merchandise Forensics project. We identify what the core issue is with a business. And in most cases, the core issue is merchandise productivity, not a failure to employ a proper omnichannel marketing strategy.

When you are at Shop.org today, ask anybody if they are measuring new item and existing item productivity. When you learn that few folks do this, you will realize that you are on the cusp of getting a head start on the competition.

Email me (kevinh@minethatdata.com) for your own customized project.

Click here for file layouts and pricing information.

Or buy the booklet on Amazon (click here).

September 29, 2013

Dear B2B Catalog CEOs: Average Order Value

Dear B2B Catalog CEOs:

Hi there - a special message just for you!

B2B cataloging is a bit different than B2C cataloging. Not as different as the pundits would have you believe, but different enough to create unique surprises.

Average Order Value is one of those surprises.

In B2C, you're likely to see something like this:
  • Response Rate = 2%.
  • Average Order Value = $130.
  • Dollar per Book = $2.60.
  • Book Cost = $0.50.
  • Profit Factor = 40%.
  • Profit per Book = $0.54.
In B2B, you're likely to see something like this:
  • Response Rate = 1%.
  • Average Order Value = $600.
  • Dollar per Book = $6.00.
  • Book Cost = $0.50.
  • Profit Factor = 40%.
  • Profit per Book = $1.90.
Average Order Value plays a major role in B2B catalog marketing. For B2B folks, AOV is dramatically bigger than it is for B2C. This causes each individual catalog to simply be more profitable, even if response rates are half or less of what is observed in B2C marketing.

As a result, the catalog must be part of the B2B marketing strategy, even if almost nobody responds to it.

It addition, large AOVs allow page counts to be inflated, compared to the B2C world where page counts are going to be pushed down in Darwinian manner.

When I started my consulting work back in 2007, the differences in circulation strategy between B2B marketers and B2C marketers were not all that much different. In 2013, for B2B marketers with large AOVs, the catalog marketing strategy is frequently going "Back to the Future", if you will - large AOVs promote a 1990 style catalog marketing strategy, opposite of almost every piece of advice you get anywhere else.

This only happens, of course, if your AOV is greater than about $400.

Ponder the consequences of AOV - B2B vendors love to capitalize on large AOV's, and for good reason.

September 26, 2013

Last Chance!

Here's your last chance to get Hillstrom's Health Index (HHI) at $2,900. 

Nobody accepted the free offer.

But demand for the $2,900 offer has been good.

Next week, I'll have enough data via the trial offer to begin offering this as a full-price product ... the price point will be $4,900, based on the test cases I've analyzed and the results obtained. That's a spectacular deal - one any business can afford. If you're wondering what it might be like to work with me, here's an entry-level opportunity to find out.

I've revised my product offering - click here for details - you'll read a brief description of the work that is most popular, you'll see pricing information, and you'll see file layouts

From this point forward, Merchandise Forensics is being sold at full price, all test cases have been analyzed, reported on, and completed - 2013 test pricing for Merchandise Forensics is no longer available.

Also, I'm getting feedback for four other types of projects:
  1. Clario Optimization Review: This has come up numerous times in 2013 - so far, 100% approval of Clario's work. If you have concerns, I'll be happy to review the impact they have on profitability.
  2. Advisory Role: It came up this week - a business wants on-demand assistance with various strategic issues. I'm available for advisory work - I don't just do data mining and geeky math. Contact me for details.
  3. Buying/Selling a Catalog Business: I get many private equity inquiries - folks looking to evaluate businesses. I also get inquiries from CEOs who lead struggling businesses. Sometimes, I get the inquiries from each side! If you're on either side of buying/selling, we can use the HHI framework above as a starting point for evaluation.
  4. Omnichannel Evolution: Specifically, folks want to understand what role mobile plays in the evolution of traditional e-commerce. There's data that supports growth opportunities, and there's data that suggests that mobile completely cannibalizes and obliterates traditional e-commerce. Hire me to find out what end of the spectrum your business is on!

September 25, 2013

Business Is Bad - Blame Marketing?

This table, from the Merchandise Forensics booklet, illustrates A/B/C items for an actual business.

Remember, I categorize A/B/C/D/F items in the following manner ... "A" = Top 5% in Demand volume and Top 5% in Unit volume ... "B" = Top 5% in Demand volume, Bottom 95% in Unit volume ... "C" = Bottom 95% in Demand volume, Top 5% in Unit Volume ... "D" = Top 45% in Demand Volume, Not In A/B/C Category ... "F" = All Other Items.

Ok, back to the business being analyzed in the booklet.

In total, A/B/C items dropped from 120 in 2011 to 105 in 2012 to 94 in 2013. In other words, the number of highly successful items (which comprise 6-9% of items and 50% of annual demand) is on the decline.

Look at the middle portion of the table. This is where we depict the number of existing items. This business reasonably maintains high demand items (A/B), but has cut way back on the number of low demand / high unit volume items. In other words, this business is de-emphasizing low price point items.

Look at the bottom portion of the table. Here we observe a complete meltdown. The merchandising team sold 29 A/B/C new items two years ago - and only sells 18 A/B/C new items today. Notice that new "B" items remain relatively flat - suggesting that the merchandising team continues to churn out new, expensive items. Notice that both "A" and "C" new items are in free fall - these have lower price points.

This is a case where the merchandising team is fully accountable for the meltdown of the business. They are not offering enough new items, and they are de-emphasizing low price points.

I've been in a lot of meetings in my career. Of course, merchants are under a tremendous amount of pressure, and for good reason. But at some point in a meeting, especially when business is below plan, the merchandising team turns on the marketing team. It's "their" fault. If it is a catalog business, then the wrong customers are being mailed. If it is an e-commerce business, then email and search programs are not targeting the "right" customer.

The marketer must have this data at their disposal. Must. Have. This. Data. The marketer has all the metrics in the world to prove that catalogs, or email campaigns, or paid search, or social, or mobile works. None of it, and I mean none of it, matters when the merchandising team is making glaring mistakes.

This example, which I see happen all the time, has nothing to do with the marketing team. The merchandising team is hurting this business.



Purchase on Amazon (click here).

Contact Kevin (kevinh@minethatdata.com) for your own, customized project.

30% Off Plus Free Shipping!

When you see a whopper like this, there are three things I want you, the intrepid business leader, to do.

First, you're going to do your best to perform some sort of holdout test. Since you have retail stores, maybe you do a geographic holdout test - nobody in Arkansas gets to participate at this level of discounting. I'll get to why the holdout test is important in a moment.

Second, you're going to simulate what it takes to generate profit, before running the promotion. If you have 65% margins and you can expect your holdout group to provide shipping/handling revenue 40% of the time (at $12 per order), your profit and loss statement might look something like this (your mileage will vary).




You need a 115% increase in demand during the promotion, in order to cause the promotion to generate profit. Here, I am assuming no incremental ad cost whatsoever, that's why profit looks so dandy in a normal business situation.

Look at the "Increment" column - this is the true value of the promotion (more on that in a moment). Under normal business circumstances, we're actually making money on shipping and handling, so we generate 71% profit (you can do that when your gross margins are 65%). We generate about $71,000 profit on $100,000 of sales. The promotion generates $1,670 profit on $115,000 of sales, a bit greater than a 1% profit level.

Here's the third (and most important) part of the analysis. When the promotion ends - KEEP MEASURING THE DIFFERENCE BETWEEN TEST AND HOLDOUT - preferably for a month (or longer).

Why? Here's why!

  • Week of the Test, Promo Group = $215,000.
  • Month After the Test, Promo Group = $400,000.
  • Total Volume, Promo Group = $615,000.
The promotional group generates $615,000 in the five weeks, total. Now, let's see what happens to the holdout group, the group in Arkansas not offered the promotion:
  • Week of the Test, No Promo = $100,000.
  • Month After the Test, No Promo = $500,000.
  • Total Volume, No Promo = $600,000.
This happens all the time. Promotions pull sales "in early". The customer would have spent money in week three, but the promotion pulls the sales (and 30% off and free shipping) into week one.

THIS IS SO IMPORTANT!

I'll bet fewer than 15% of marketing analysts measure results this way. When you do this, your profit and loss statement changes, dramatically.


The profit and loss statement illustrates reality, doesn't it? In reality, the promotion causes the business to bleed profit. Sales are barely impacted, we give away precious gross margin dollars to bump up a promotional period, and then we wonder why business is flat in the following weeks?

And guess what happens when business is flat or down over the next four weeks? Somebody will want to add a promotion!!

So please, please, run this three-step analytical process. This isn't rocket science. Just do it! Whether you are the CEO, the CFO, or a smart marketing analyst, just do the math. Then, when you have the facts, you're free to decide whatever you want. But please do the math. Heck, the math might work out in your favor - it might be appropriate to discount and promote your business like this. But how else would you know it to be true?

And finally - yes, I get it, there's a viral component to these type of promotions - that's why you have to perform your test on a geographic basis, so that you minimize spillover of viral effects. All those allegedly "new" customers who would not have purchased need to be part of the profit and loss statement. Again, that's why this is done on a geographic basis, not on a mail/holdout basis.

September 24, 2013

Merchandise And Demographics Interact

I constantly run into situations where the marketing team wants to "attract a younger customer".

Let me tell you something. The merchandise you offer the customer, and the creative style used to portray the merchandise directly correlates with the customer you will attract.

Take a look at the blouse/scarf combination, courtesy of Paula Young. Do you think our friend, Jasmine, is going to be interested in this combo? I know, I know, you'll say that the discounts at an order level and the item level price adjustment are enticing, and they certainly are, but come on!

This is a Judy-centric item. Judy is going to buy this. And when presented in a catalog, it attracts a customer like Judy. Catalog marketing also attracts a customer like Judy. So the whole thing is a big ecosystem designed to attract a 60 year old customer. The merchandising strategy and creative strategy yield a predicable outcome, one that the marketer cannot challenge, without taking a profit and loss statement bath.

Here's an item, courtesy of our friends over at L.L. Bean (hi y'all!):



We're in Jennifer / Jasmine territory here, aren't we? It's unlikely that Judy will wear this. Notice the model - she's not Judy either, is she? What demographic profile will this item attract?

And then we have this example from Lands' End.



This is Jasmine, and heck, we're starting to push into Jadyn territory, aren't we?

This happens at every single company I put through the Merchandise Forensics framework. Every single company. Products align with Judy, Jennifer, or Jasmine. Have too many "Judy-centric" products, and your whole business aligns with Judy, locking you into catalog marketing. Have too many Jasmine-centric products, and you'll quickly observe that traditional marketing is less likely to work.

If you don't append age data to your customer file, then simply use a mail/phone ... online matched to a catalog ... email ... search ... pure online ... mobile/social framework. Each marketing channel simulates age ranges. Items that sell well in a pure online / mobile / social channel distribution align with Jasmine - hint, the winning items in these categories frame the future of your business. Meanwhile, items that appeal to the mail/phone shopper align with Judy. These items are your past, unless, of course, your target customer is Judy ... then these items are your bread/butter.

In every Merchandise Forensics project, I can identify the items that align with demographic profiles ... especially my Judy/Jennifer/Jasmine personas. This tells us how to merchandise landing pages, email campaigns, and the first twenty pages of a catalog marketing program.

Pretty important, don't you think?

Click here to purchase Hillstrom's Merchandise Forensics on Amazon.

Email me (kevinh@minethatdata.com) to get your own, customized project performed for you.

September 23, 2013

Hillstrom's Health Index (HHI): An Offer For You!!

Here's a problem that comes up, repeatedly. Each business has a series of metrics that are routinely evaluated. Some of the metrics look good, some of the metrics look bad.

Is the customer file healthy, or is it unhealthy?

I'd like to introduce a new product, called "Hillstrom's Health Index", or "HHI".

Think of this metric like you would think of a grade point average in college, or high school. If a metric is good, it earns a grade of "A", and receives four points. If a metric is terrible, it earns a grade of "F", and receives zero points.

Each metric is weighted, according to the importance that the metric has to the health of the customer file. New customers, of course, are terribly important - so you have to acquire more new customers than those you lose, or you earn a poor grade. And merchandise-related customer metrics are terribly important, they lead to good/poor repurchase rates. As a result, merchandise-related metrics are weighted higher.

The result is "Hillstrom's Heath Index", or "HHI" for short.

In the example above, the customer file is in dire shape. This business earns an HHI of 1.07, nearly a flunking grade. The business is losing customers at an epic rate. Repurchase rates and orders per buyer are crumbling. As we look down the report card, we notice that demand from new items is the worst performing metric. Demand from existing items, however, is on the increase - clearly, customers are not getting the new product diet they want, and are shifting demand as a result. Also notice that demand from best-performing items is on the decline. This business is generating volume from marginally performing existing items, with everything else crumbling.

One other good metric - only 54% of demand is generated via discounts/promotions, compared to 57% a year earlier. Notice that the metric is graded well.


An Offer For You - Valid For The Next Seven Days Only:

I've analyzed numerous businesses through this framework - with HHIs ranging between 1.07 (a catastrophe) and 3.40 (from a rapidly growing, existing catalog brand). I'd like to test the framework with additional data, before releasing the algorithm as a product.

And having said that, allow me to mention three common complaints I receive from you, the loyal reader.

  1. "I just wish you offered free products from time to time. Free is good!"
  2. "My business is too small to afford your services, I wish you had a product that was affordable and could help me make actionable choices."
  3. "We read your blog all the time, but our CMO wants to try working with you on a trial basis before committing to a large project."
Ok, your concerns have been heard. Here's a chance for you to "step up", as the pundits say, and follow through on your requests.  I am offering two choices while testing this product, prior to releasing it at full price.

  • Option #1 = For $2,900, I will run your data through the HHI Algorithm. You'll get to see how your business stacks up against the metrics listed above, as well as a series of comp segment metrics (metrics not published in this post). Your business will earn a grade (A/B/C/D/F - based on the grade point average, or as I am calling it, the "HHI"). You will get recommendations on how to fix any problems, and you'll get praise for the things you're doing well!
  • Option #2 = For FREE, I will run up to TWO companies through the HHI Algorithm. You get a free analysis, however, you agree to have your name, your company name, and results analysis published on this blog as a case study. You will be interviewed, and your findings and comments will be shared with at least 7,000 blog subscribers, Twitter followers, and site visitors.
If you've been sitting on the fence, looking for an opportunity to get in on the action, then this is your opportunity, why not take advantage of it?

Click here to obtain the file layout necessary to complete this analysis - only File #1 is needed to complete an HHI project. Then email me (kevinh@minethatdata.com) to get your project loaded up in the queue.

How Does It All Fit Together? The Merchandise Forensics Map

One of the geekier elements of a Merchandise Forensics project is this image - a Merchandise Forensics Map.

Geeky, but useful!

I use a Factor Analysis to produce the map, using 12 month buyers and weighted historical demand percentages for each variable.

Ok, enough of the geekiness.

On the left hand side of the image, we see the average attributes for first-time buyers. Notice that the price points are low. On the bottom left, we see Christmas focused first-time buyers ... on the upper left, we see Spring/Summer first time buyers (acquired primarily via organic search and third parties).

On the right hand side of the image, we see loyal buyers. What other attributes are tied to loyal buyers?

  • Promotions.
  • High Price Points.
This is a fascinating dynamic. Loyal buyers tend to buy high price points ($50+), but buy them using discounts/promotions, eliminating the gross margin benefits derived from higher price point items.

The Merchandise Forensics Map helps us observe attributes that belong together - we get to see the attributes associated with Cyber Monday, or purchases in May, or via Paid Search. This helps us understand why there are issues with the business - in the case above, this business struggles with customer loyalty - notice that new buyers buy low price point items, while existing buyers purchase high price point items at a discount. This is a huge disconnect, the business is being managed in two separate realms. That's not good.

September 22, 2013

Dear Catalog CEOs: Circulation Evolution Over TIme

Dear Catalog CEOs:

Yes, I'm watching you!

I'm watching how you manipulate circulation over time.

Back in 1995, life was good, and catalogs were at the top of the food chain. Quite honestly, we were lazy back then ... profit was just sitting there, waiting to be harvested, and we didn't pick it all up.  It was common to see a profit factor of 30% - 30% of demand flowed-through to profit. I recall arriving at Eddie Bauer in 1995 - I saw metrics in the 25% to 29% range - and that company generated 5% pretax profit. That's some serious laziness, when your gross margins were more than 50%.

In 2004, we applied matchbacks. Catalogs weren't quite as productive as in 1995, but we managed declining productivity by cutting costs (yielding a profit factor of 35%). As a result, slightly higher book costs and better profit factors and lower productivity yielded about the same amount of profit.

Since 2004, customers fled to the internet. Today, it is very common for pure catalogers to see between 50% and 80% of demand happen anyway, without mailing a catalog ... and for catalogers with customers age 65ish, the ratio is between 25% and 50%. This leads to dire consequences for the catalog (not corporate) profit and loss statement. And in 2013, those under the age of 30 have now fled the internet, moving to a hybrid of mobile+social ... moving two full steps away from the catalog business model.

In response, we cut even more expense - today, it's common to see 45% to 50% of demand flow-through to profit. That's amazing! And thoroughly depressing.

We accomplished this by removing humans from our business. 

  • No humans at the call center (out of necessity).
  • No humans in the information technology department (outsourced to vendors).
  • No humans in customer acquisition (outsourced to co-ops).
  • No humans in search (outsourced to vendors).
  • No humans in email marketing (outsourced to vendors).
  • No humans analyzing results (outsourced to matchback/attribution experts).
  • No humans in the warehouse (outsourced to robots).
  • No humans managing the website (outsourced to vendors).
  • No humans managing cross-selling, pay-per-click, retargeting (outsourced to vendors).
Yes, we still have humans coordinating these programs - but not managing these programs. There is a big difference between being accountable, and holding vendors accountable.

What happens when we remove humans from merchandising and creative?


But we get to keep mailing catalogs.

Take a look at the red numbers in the table. In 1995, 80% of the profit came from 60% of the circulation. In 2004, 80% of the profit came from 60% of the circulation. In 2013, 80% of the profit comes from 30% of the circulation.

In other words, the economics of catalog marketing are changing. There's only so much expense that can be trimmed ... and catalog costs will continue to rise ... and demand will continue to leak out of the catalog ecosystem into e-commerce (and mobile), leaving fewer and fewer catalog-centric customers to mail to.

So what does the scenario look like in 2022? Maybe like this:
Oh boy!

In 2022, given current trends, productivity will drop further, as e-commerce is old-school, mobile is established, and hologram marketing is capturing market share. This leaves us with little profit - no amount of financial manipulation fixes the problem.

And did you notice that profit is close to unchanged at any circulation level? Some will cut back to almost no circulation at all - some will push their chips all-in and mail deep - and everybody will make peanuts, regardless. It's already happening!

Has anybody pointed this out to you?

Have your marketing team perform this analysis for you, over time. Apply proper incremental rates after evaluating mail/holdout tests. You're going to see something similar to this table.

How does this analysis change your thinking about the future?

I know, nobody talks about this. But if your average customer is age 55+, the 2022 scenario is coming. For many, it's already here.

September 19, 2013

Cyber Monday Tip #4

If you're going to forego profit to "remain competitive" on Cyber Monday (and by the way, if you're going to remain competitive on Cyber Monday, why don't you care about remaining competitive on July 8, or April 14, or January 27?), at least do it with some element of strategy.

There are numerous Cyber Monday strategies:
  1. 30% off plus free shipping. The classic. Just give everything away in a big Cyber Monday fire sale that will leave NPR's Markeplace giddy with joy!
  2. Free shipping only. In modern times, everybody is giving away free shipping, so you won't stand out, but you won't take the profit bath that your competition will suffer.
  3. 30% off key items. This may be a viable tactic - via email and your home page, you advertise a half-dozen items, limiting discount exposure.
  4. 50% or more off liquidation items. Also could be a viable tactic, solving a liquidation problem while limiting exposure to best sellers.
Yes, I know, there's an infinite number of tactics and combinations.

If you opt for #3 above, be sure to identify key items that encourage repurchase. There's no need to take a profit bath on the transaction, and then cause those customers to not come back and buy again in the future.

Need help analyzing key items? Contact me (kevinh@minethatdata.com) for Cyber Monday assistance.

September 18, 2013

Kevin Spacey, Omnichannel, and Merchandise Forensics

Start your day with this five minute video of Kevin Spacey talking about the media version of "omnichannel". Click here, please. Yes, I'll wait for you to finish.

At around 1:30, you hear a series of metrics ... here they are:
  • Last year, 113 TV pilots were made.
  • 35 of the 113 were chosen to go to air.
  • 13 of the 35 were renewed, but most of those are gone, now.
Mr. Spacey goes on to talk about the media version of Omnichannel, watching whatever you want on any device at any time.

Notice, however, that he starts by talking about Merchandise Forensics ... not Omnichannel ... the media version of Omnichannel comes much later, and does not exist without great content to watch. He starts by talking about the tournament that new content goes through (133 to 35 to 13).

Omnichannel works (+/- 5%). Merchandise works much better (+/- 50%).

I've worked on twenty Merchandise Forensics projects since the start of 2012 ... sixteen of the twenty indicated a significant merchandising problem, most suggesting that new product development is lagging far behind where it needs to be. When you fix merchandising problems, omnichannel tactics work a whole lot better!

Choices:


September 17, 2013

Merchandise Forensics: Loyalty Killers

You know what Loyalty Killers are, right?

Loyalty Killers are items that are both popular, and cause customers to not repurchase at above-average rates in the future.

Pick every item sold during 2012. Select only the best-selling items ... I use a three-pronged approach.
  • "A" Items = Top 5% in demand volume and Top 5% in unit volume.
  • "B" Items = Top 5% in demand volume, not in Top 5% in unit volume.
  • "C" Items = Not in Top 5% in demand volume, in Top 5% in unit volume.
This represents a small percentage of items sold, but represents the majority of demand you generate, each year.

These items are popular - featured on landing pages and email campaigns and on your home page. Or in the first twenty pages of your catalog. They're everywhere.

And some of them kill the long-term success of your business!

For the A/B/C items listed above, identify the customers that bought them in 2012. Then measure, in 2013, the rebuy rate ... how many of the customers repurchased in 2013? No, not repurchased that item - but repurchased in general.

You'll often notice that a third of these items generate repurchase rates in the 50%ish range, while a third generate repurchase rates in the 25%ish range. The latter third ... these items cause customers to not come back and buy again in the future.

If you knew that a third of your best selling items caused customers to purchase in the future at significantly reduced rates, would you feature them in your email campaigns, landing pages, catalogs, and on your home page?

Perform a Loyalty Killer analysis - it's pretty darn important, folks!

Choices:

September 16, 2013

Slooooooow Decay

Maybe this is happening in your business as well - you have items that generally sell well, and have been around forever.  Over time, these items experience a sloooooooow decay:
  • 2008 = $7.00 million.
  • 2009 = $5.60 million.
  • 2010 = $4.48 million.
  • 2011 = $3.58 million.
  • 2012 = $2.87 million.
  • 2013 = $2.29 million.
Think iPod, for instance.

My projects show a consistent trend. Best items, sometimes called "winners" or "heroes" or something equally compelling, erode at a slow rate. This is called the "decay rate". The average decay rate in my projects is between 10% and 30%. Those are big numbers! In the table above, the decay rate is just 20% - but over five years, 67% of the demand is gone.

Even worse, when a whopper of a winner decays, you'll frequently observe that best customers continue to buy the item - they're in the habit of buying it - and they tend to buy it in old-school channels (i.e. catalogs or email). This is destructive in two ways. First, the blend of customer and demand results in the item continually being featured, frequently with the best creative. This ages the catalog (or email marketing program) faster than normal. Second, it prevents newer items from getting the real estate necessary to become winners. I see this all the time. What would you do? Of course you're going to give the best real estate (catalogs, home pages, landing pages, email marketing) to the items that you know will work. In so many ways, we need to almost do the opposite - we need to give precious real estate to new items, so that new items can become winners.

Yup, happens all the time.

Here's another interesting problem. You can slow down the rate of decay by not introducing new products. If you don't give the customer new choices, the customer is forced to spend money on the old choices. Of course, this cannot work in the mid-term or the long-term, but I see it happening all the time.

If I asked 100 of you what your existing product decay rate is, how many of you would know the answer?

It's important to know the answer. Your marketing programs depend on you having product decay knowledge.

Hillstrom's Merchandise Forensics:

September 15, 2013

Dear Catalog CEOs: The 80/20 Rule

Dear Catalog CEOs:

Did you read this article (click here)?

A portion of this article made sense to me, and it directly relates to your catalog business.

Most of us are dealing with catalog marketing as if it was 1995. Deceived by the way matchbacks overstate results, we mail the same circulation depth we always have. If we used to mail 400,000 housefile names, then we still mail 400,000 housefile names.

Back in 1995, our profit and loss statement looked like this:
  • Demand = $3.00 per customer.
  • Profit Factor = 30% of demand.
  • Cost of the Catalog = $0.40.
  • Profit = $3.00 * 0.30 - $0.40 = $0.50 per customer.
Today, when we properly evaluate incremental demand via mail/holdout tests, our profit and loss statement looks like this:
  • Demand = $3.00 per customer.
  • Incremental Demand = $1.50 due to catalog, $1.50 happens anyway online.
  • Profit Factor = 40% of demand.
  • Cost of the Catalog = $0.60.
  • Profit = $1.50 * 0.40 - $0.60 = $0.00 per customer.
Now, 9 in 10 catalog CEOs would look at this situation, and say "fine, we'll keep mailing customers".

Of course, in five years, the relationship is going to look like this:
  • Demand = $3.00 per customer.
  • Incremental Demand = $1.00 due to catalog, $2.00 happens anyway online.
  • Profit Factor = 45% of demand.
  • Cost of the Catalog = $0.65.
  • Profit = $1.00 * 0.45 - $0.65 = ($0.20) per customer - a loss.
But there's no sense talking about what is coming ... you are busy projecting 2014 as we speak.

Back in 1995, the 80/20 rule was largely intact ... but it meant that you could mail very, very deep, and keep generating profit.

Today, with the internet (and, increasingly, mobile) generating sales that happen independent of a catalog, without any catalog mailings at all, the 80/20 concept allows you to greatly free up your advertising budget. 20% of the 12-month buyer file is responsible for 80% of the demand generated by a catalog - the world has changed.

In fact, in most of my Catalog PhD projects, I can cut 25% to 30% of the housefile circulation with only a 2% to 5% impact on demand, greatly increasing profit.

Go rank your housefile circulation, top to bottom - then sum cumulative profit. I'll bet you learn that almost all of your profit comes from the top of the housefile. 



Purchase on Amazon, click here.

Hire Kevin for your own Catalog PhD project - email me (kevinh@minethatdata.com).

September 14, 2013

Merchandise Forensics Being Discounted By Amazon - $8.63

If you like discounts and promotions, and I think you do, then you'll love the fact that Amazon is discounting the new Merchandise Forensics text ... for Amazon Prime members ... from $11.95 to $8.63 ... print version only.

Here's a link to the print version (click here).

And a link to the Kindle Version (click here), $7.95.

Click here for file layouts if you're thinking of hiring me for a project - Merchandise Forensics projects are about 60% of my work these days. This file has been downloaded 39 times in the past two days alone, so there's interest, folks!

If you work in a large office, don't you think every one of your employees deserves a copy? Don't you? Why not order a dozen or two dozen copies, one for everybody!


September 12, 2013

Dell and Merchandise Forensics

By now, you've read that Dell is going private. Click here for some of the details.

This post is not meant to pick on Dell. It's meant to get us to think.

Why do I want you to study Dell? Well, they did an awful lot of the things that the experts tell us to do - good decisions.

  1. They are social media leaders, heck, they're monitoring every word of text I'm writing as we speak - it's almost like the NSA! If you're best at social media, it should cause breakthrough sales gains, according to the experts.
  2. They are omnichannel leaders. They moved into retail, not their own branded stores, but you can go to Wal-Mart and get the goods. That's called "bricks 'n clicks" by the experts. We're told that retail plus e-commerce is unbeatable.
  3. They are promotional gurus - the carousel on the mobile site (remember, mobile or die) as of this writing (9/12) shows summer clearance, a $200 promo eGift card, and a coupon for smart shopping. Of course, that ignores merchandise, and what Dell sells is merchandise, but what the heck, it's the tickling of the buying bone that marketers love. We're told we have to do this, all day, every day, to "remain competitive".
Ok, how did omnichannel + social pay off?
Maybe sales went bonkers, given the bricks 'n clicks approach and the omnichannel tactics and promotional strategy and social media wizardry?
  • Net Sales, February 1, 2008 = $61.1 billion.
  • Net Sales, February 1, 2013 = $62.1 billion.
So, what happened? How could a combination of omnichannel excellence, social media power, and promotional tactics yield ... well ... this?

Merchandise?

Dell has tons of smart people, more than most companies. If they performed a Merchandise Forensics analysis, they'd quickly observe that existing products were tiring, while recent new products failed to generate the volume prior new products generated, leaving Dell with no choice but to try to maximize omnichannel + social + promoitons to compensate for merchandise challenges. 

This is a big theme, folks, one I see repeated everywhere - merchants struggle, marketers get yelled at, so marketers go bonkers on channels and ad dollars and promotional strategies. It's a time honored tradition. It's a sign of a ailing business model.

On February 1, 2008, the iPhone was reasonably new, the Android operating system was in a stage of relative infancy compared to today, and the iPad was thirty months away from introduction.

In other words, the merchandise world that Dell participated in changed.

You, too, can see these things happening in your business. They begin to happen subtly, the metrics appear but net sales largely hold their own. It's pretty darn important to run a Merchandise Forensics analysis, to stay on top of key issues.

Hillstrom's Merchandise Forensics:

September 11, 2013

It's Here! Hillstrom's Merchandise Forensics

We spent much of the summer talking about Merchandise Forensics, and for good reason. Twenty recent projects strongly suggest that merchandising issues, not marketing challenges, are the primary reason why a business struggles to grow.

In "Hillstrom's Merchandise Forensics", I explore analytical techniques that clearly illustrate how merchandising challenges result in reduced customer productivity, and decreased marketing productivity.

Here's the thing, dear readers. I'm tired of watching the marketing department get pummeled every time the business does not grow. I mean, it's late 2013. Think about how many times a twelve-month buyer learns about your brand:
  • Six to forty catalogs from catalog brands.
  • +/- 100 email marketing messages.
  • A mobile experience.
  • A website that can be visited 24/7/365.
  • Search, paid and organic.
  • Social media.
In other words, a customer is going to hear about your business multiple times a week. Multiple times each week!! What more do you want the marketing department to do? Should they sit in the living room of every customer and demand that the customer purchase merchandise immediately?

Of course, the marketing department can do a better job. But so can the merchandising department.

This booklet helps the marketing team figure out why a business is struggling to grow, via merchandise productivity. I use actual data to explain how new items, existing items, price points, discounts/promotions, seasonal issues, Cyber Monday (yes, it causes problems), merchandising categories, new customers, and existing customers all interact to limit overall business productivity. The reports are reasonably simple and straightforward to create. There's no reason somebody in your analytics/marketing department cannot produce these reports for you, on a monthly basis. And if they won't, then you contact me (kevinh@minethatdata.com) and I'll do it - now!!

As mentioned earlier, Merchandising Forensics projects now make up the majority of the work I do. I worked on at least twenty projects of this nature since 2012, in sixteen of the projects, I was able to easily identify a merchandising issue that held the business back.

If we've learned anything, it's that having an infinite number of marketing channels and a half-dozen customer contacts a week does not cause sales to increase. What causes sales increases are customer love for merchandise. And increasingly, my projects illustrate that we are doing a very, very bad job of developing new merchandise.

You can buy this booklet in one of two ways.
And you can contact me right now (kevinh@minethatdata.com) to have your own, customized Merchandise Forensics project prepared ... contact me now, because projects are going to line up really fast (click here for file layouts).

Again - I'm tired of watching the merchandising team point fingers at the CEO, CFO, CMO, and marketing team, as if the merchandise they sell was perfect and they were not accountable for challenges. Granted, merchants suffer a disproportionate amount of scrutiny - but it isn't the fault of the CFO, CMO, or marketing team that nobody wants to buy odd merchandise. So buy this booklet, then hire me, then have data at your disposal to help everybody understand why 3,439 channels haven't resulted in a sales increase.

Sales are not increasing because customers don't want to buy the merchandise. Measure it! Buy the book right now!

Cyber Monday Tip #3

Cyber Monday does not happen uniformly, across channels. Not at all. Cyber Monday aligns with a specific customer who has specific interests.

This outcome is pretty common:
  • % Nov-Dec Phone Demand On Cyber Monday = 2.5%.
  • % Nov-Dec Online Demand On Cyber Monday = 7.4%.
  • % Nov-Dec Email Demand On Cyber Monday = 8.1%.
  • % Nov-Dec Search Demand On Cyber Monday = 8.8%.
  • % Nov-Dec Affiliate Demand On Cyber Monday = 9.2%.
What does this tell us about Cyber Monday demand?

Phone demand is usually catalog-driven, and therefore, was generated weeks prior. Catalogs mailed in early-mid November are largely responsible for phone demand on Cyber Monday. Or it tells us that old-school, 55+ rural customers don't care about Cyber Monday. Either way, they're not participating.

Online demand is driven from a mix of many activities. Overall Cyber Monday awareness yields a higher percentage here. Also - keep in mind, there are 61 days in Nov/Dec ... when you generate 7.4% of demand on just 1.6% of the timeframe, well, that's staggering. If profitable, this is good. If the sales are all discount-centric, well, this is potentially very harmful.

Email demand is at an even higher rate. I see this happen when the business promotes Cyber Monday via email. Too often, email buyers are already discount buyers, and are exactly the ones who are compelled to buy on Cyber Monday. We flush gross margin dollars out the door, not helping anybody but the journalists who crow about Cyber Monday.

Search demand is at an even higher rate still. Attribution lovers know that email campaigns drive customers online to search for the lowest prices and best promotions. Cyber Monday is a timeframe where email + search = big discounts.

And finally, Affiliate demand possesses the highest Cyber Monday demand rates. The customer hunts relentlessly for the best deal - I mean, who wants to pay 40% off plus free shipping on Cyber Monday, right? That's too much when you can find 50% off plus free shipping somewhere else!

So, tip #3 is to measure the share of demand in Nov/Dec that happens in each channel on Cyber Monday. If you see rates comparable to what we observe above, you'll want to flag Cyber Monday buyers in your database by channel, so that you treat them differently during the rest of the year.

Or, you can contact me (kevinh@minethatdata.com), and I'll do the work for you!

September 10, 2013

Cyber Monday Tip #2

Continuing our series on Cyber Monday, the giddy online holiday designed to give away gross margin so that the media have something to talk about ... let's move on to Tip #2.

Tip #2 involves existing customers and new customers. In most of my projects, Cyber Monday does not yield a significant increase in new customers, compared to the rest of November/December.
  • % January - October Buyers New = 45%.
  • % November Buyers New = 57%.
  • % Cyber Monday Buyers New = 55%.
  • % December Buyers New = 53%.
The scenario painted above is typical. We see that customer rates are similar through November/December.

Let's assume that you promote more heavily on Cyber Monday than the rest of November/December.
  • November Buyers = 38% Use A Promotion.
  • Cyber Monday Buyers = 90% Use A Promotion.
  • December Buyers = 42% Use A Promotion.
In this situation, you begin to expose yourself to a challenge. You essentially acquired customers via promotions instead of a more reasonable mix of full-price purchases and promotions.

I like to run 12 month repurchase models - measuring how much customers spend on full-price and discounted products. Here's a typical outcome for full-price demand.
  • November Newbies - Future Full Price Demand = $40.00.
  • Cyber Monday Newbies - Future Full Price Demand = $27.00.
  • December Newbies - Future Full Price Demand = $38.00.
And here's a typical outcome for future demand tied to a discount/promo code.
  • November Newbies - Future Demand Tied To A Discount Code = $25.00.
  • Cyber Monday Newbies - Future Demand Tied To A Discount Code = $36.00.
  • December Newbies - Future Demand Tied To A Discount Code = $25.00.
Remember, in our example, we brought on newbies via a promotion during Cyber Monday. And while future demand is largely similar (counting demand at full price, and demand tied to a discount/promo code), demand shifts to discount/promo codes. This causes the demand to be less profitable, causing us to conclude that shifting demand into Cyber Monday is not optimal for this hypothetical business.

Of course, your mileage will vary.

But how do you know unless you run the analysis yourself?

Contact me (kevinh@minethatdata.com) for your own Cyber Monday analysis.

September 09, 2013

Cyber Monday Tip #1

I know, you're pushing your chips all-in on Cyber Monday. You "have" to, in order to remain competitive. Or so you tell me. And the press love it - you give away 35% plus free shipping, they make ad dollars by yelling about it.

So let's take some time to share a few tips with you.

First, I want you to analyze how business shifts over time. Compare 2008 to 2012, and look at three time periods:
  • Month of November Demand, Excluding Cyber Monday.
  • Cyber Monday Demand.
  • Month of December Demand, Excluding Cyber Monday.
Here's an example I witness, often:
  • 2008 November = $10,000,000.
  • 2008 Cyber Monday = $1,000,000.
  • 2008 December = $9,000,000.
  • 2008 Seasonal Totals = $20,000,000.
So far, so good. Now let's look at 2012:
  • 2012 November = $9,000,000.
  • 2012 Cyber Monday = $2,500,000.
  • 2012 December = $9,000,000.
  • 2012 Seasonal Totals = $20,500,000.
What do you observe?

If all things are equal, then Cyber Monday grew from $1,000,000 to $2,500,000 ... but the total business grew by $500,000.

This tells us that $500,000 / $1,500,000 = 33% of Cyber Monday business is truly incremental, and new to the business.

In other words, you gave away the farm (i.e. 30% off plus free shipping) to gain $1,500,000 of business, when $1,000,000 of it would have happened anyway.

Your profit and loss statement is likely to look like a catastrophe, assuming you give free shipping away in Nov/Dec - assuming the 30% off is due to Cyber Monday.
  • Incremental Demand = $500,000.
  • Discounts on Gain of $1,500,000 = $450,000.
  • Gross Margin = 50% = $250,000.
  • Net Contribution = $500,000 - $450,000 - $250,000 = ($200,000).
So - Tip #1 is an easy one ... quantify how much of Cyber Monday is incremental, is truly new business. Run a profit and loss statement (the one above is terribly simplistic, for illustrative purposes), and measure if you are actually helping your business.

Contact me (kevinh@minethatdata.com) for your own Cyber Monday analysis.