1989: Identifying Interesting Opportunities

Popular songs in 1989 include:

• "Like A Prayer - Madonna".
• "Eternal Flame - Bangles".
• "Another Day In Paradise - Phil Collins".
• "The Look - Roxette".
• "Love Shack - B52s".

We did some interesting things at the Garst Seed Company.

If I remember this correctly, one of our researchers wanted hybrids that yielded well, and dried down quickly.  The hybrid could be harvested faster than normal, helping the farmer achieve high yields when the growing season was shortened by bad weather.

The analysis was quite simple - but elegant - fit a line through a relationship between yield and days to dry down.  Any hybrid that was "below the line" in the graph above and had a high yield (> 120 bushels per acre) was a hybrid that would be "fast tracked".  The researcher wrote a paper, and invented a simple but effective methodology, and I got to do the actual work for the researcher.  All in all, it was a lot of fun!

You can have fun, too.

In e-commerce, the methodology above works for analyzing conversion rates.  Plot conversion rates on the x-axis, plot profit per conversion on the y-axis.  You want to explore scenarios where you have high conversion rates and high profit per conversion.

So much of what we're doing in 2013 is not new - it's a derivation of stuff done long ago, applied to a modern situation.  Borrow what has been done, twist it to apply to modern situations, and identify and interesting opportunity!

1988: Deer Romping

Top Selling Song of 1988 = Faith, by George Michael.

Ya gotta have faith!

Sometimes, however, it was hard to have faith in the data.

In the autumn of 1988, at the Garst Seed Company, we were analyzing the annual "harvest". Each research plot was harvested, yields were recorded, and data was sent to a twenty-three year old analyst named Kevin Hillstrom.

Research plots were in grids.  I analyzed each cell in the grid - a cell represented a corn or sorghum hybrid, the number represented the yield of the hybrid.

Look in the top half of the table.  See the red numbers?  Those "yields" don't look right, do they?

Turns out that a deer "romped" through those plots, ruining the experiment.  Yields are 70% too low!

These experiments were expensive to conduct, and took six months from planting to harvest to complete.  You couldn't just say that the experiment was wasted, you had to do something with the data.

We used a method called "GLM", or "Generalized Linear Models" ... to correct for cases where deer romped through and destroyed our experiment.  By adjusting for the rows and columns in the field (and by planting a second or third "rep" ... an identical but re-randomized test), we could predict (values in blue above) what "should" have happened.

This allowed us to save the company a fortune ... we used good data to predict what should have happened.

These days, you go out on Twitter, and you'd swear that folks just invented A/B tests.  Well, the good folks at Kansas State and Iowa State University were executing and analyzing randomized plots, using GLM to adjust for outliers ... as far back as the 1940s ... which means these methods were being used long before that!

We have the same problems today in e-commerce ... how do you measure conversion rate when the email marketing team adds/removes campaigns from the schedule?  Or how do you measure conversion rate when you toss a bunch of sloppy brand advertising on the home page, foregoing revenue generated by top-selling items that were previously sold on the home page?

You use this procedure, this "GLM" procedure, to predict "what should have happened".  Heck, you don't even have to get that fancy ... just eyeball the results, it's better than using sloppy, bad data, isn't it?

25 Years

There should be some kind of award for being gainfully employed for the past twenty five years.

But there isn't.  Instead, you get to read another in a long line of blog posts.

After graduating from the University of Wisconsin with a BS in Statistics, I started my first day of work at the Garst Seed Company, in Slater, Iowa, on July 15, 1988.  Next week, it's the twenty-five year anniversary of the magical moment when I put down the Des Moines Register, entered the corporate campus of the Garst Seed Company, and emerged nine hours later with a 1,200 page SAS manual in hand.

In the next seven weeks, we'll explore marketing and analytical topics from 1988 to 2013.  All examples will be tied to current issues, so no need to fret about blog posts focusing on the best ways to avoid SB37 errors in JCL code on a mainframe computer.

Barnes & Noble

Multichannel / Omnichannel experts tout the "bricks 'n clicks" advantage.

But then we read about Barnes & Noble (click here for a link to an article about losses in the Nook Division - losses offsetting in-store profit).

Let's think about this for a moment.

• Experts say Omnichannel / Multichannel > Single Channel.
• B&N will reduce stores by a third in upcoming years.
• B&N will pull back on digital tablets.
• B&N has every advantage over Amazon, according to omnichannel experts.
• Amazon is not pulling back, are they?

At some point, we have to concede that the whole multichannel / omnichannel thing is designed to generate page views, to sell vendor solutions, to sell research reports, and to generate Management Consulting engagements.

Do what is right for your business, not what is right for somebody to generate more page views.

Dear Catalog CEOs: A Business Bubbling Under The Core Business

Dear Catalog CEOs:

Here's an interesting quote ... I heard it recently from one of your peers.  I had not heard a true catalog executive offer this previously.  It marks a shift in thinking.

• "We were told we had to align all of our channels.  Yet last week, we're looking at merchandise reporting, and we can see the proof in the pudding.  We have online items that are fundamentally different than our core catalog items.  They sell reasonably well, and with minimal ad cost, they're very profitable.  But more important, this tells me that we're serving different customers.  We can no longer think about the catalog as the center of the ecosystem.  We have different customers.  Older customers and younger customers.  We need to meet all of their needs.  The catalog can't accomplish our goals anymore.  This will disrupt our entire organization, how we do things."

I'm going to stop right there.  Take a moment, and let the paragraph sink in.  I'll be back in a moment.

...

...

...

I'm back.

This was one of your peers, not me, saying this.

It may just be that you have a new business, bubbling under the surface of your core business.  If you look at the data the right way, you'll see this.  The future is staring you right in the face.

Now that you are back from a weekend of enjoying fireworks, spend a little time thinking about the quote, thinking about what it means for your organization.

July 4

I'll be back on July 8 with scintillating facts that have the potential to revolutionize your business.

Until then, take a break, and celebrate the outcome of the Revolutionary War ... explode a few devices (safely, of course).  Spend a day at the beach.  Grill brat patties over an open flame.  Enjoy a cold drink.

Or experience a summer sunset.  Your choice!

Two Items

You have two items:

• Item #1 was promoted in your catalog, generating $10,000 in demand and $1,000 profit.  Not bad!
• Item #2 was not featured in your catalog.  It generated $3,500 in demand and $1,700 profit.

Which item do you prefer?

Most of you prefer item #1, don't you?

"It sold more!".

"We captured market share!".

"Businesses grow or they die."

There are reasons for favoring item #1.  At a $25 price point, it means you sold 400 units, most likely to about 350 customers.  For item #2, you most likely sold 140 units to maybe 125 customers.  Item #1 gives you what I call "file power".  I'm a big advocate of file power.  I'll take an incremental customer over an incremental dollar of net sales any day of the week.  In this case, you get both - incremental customers and incremental sales.  Incremental customers are good, because they pay us back in the next 1-3 years.  Ask Amazon how they feel about incremental customers.

There are reasons for favoring item #2.  Two big reasons.  First, you didn't have to spend ad dollars to generate the sales.  Sales that are generated by brand loyalty are more valuable than sales generated by advertising.  You get to save the ad dollars, and possibly do something else with them that will generate sales.  Second, you generated more profit.  Now, I get it, nobody looks at profit anymore.  It's only the most important metric in your whole business, it's the metric that allows us to earn a salary.  Without profit (or cash), you're sunk.  This item generates more profit/cash than the first item.  Therefore, in many ways, it is more valuable.  The incremental profit increase allows us to invest in advertising, buildings, new businesses, new items, salary increases, bonuses, you name it.  Business leaders that prefer item #2 and reinvest profit/cash in new activities tend to find a path to the future faster than those who are cash strapped due to a 30% ad-to-sales ratio.

Each item possesses strengths.

Which strength do you favor?

The strength you favor says a lot about the type of business you desire to create.