The Curse of Vanilla

Posted in General on April 27th, 2010 by petekim – 1 Comment
Have you ever noticed that it’s the things you didn’t think of that always cause the most trouble?
I’ve been working on Dynamic Creative Optimization (DCO) programs for almost three years now.  And looking back, the issues that have cropped up over and over again are those that I had no clue were coming.
The one that I’d like to discuss today, I call the “Curse of Vanilla.”
A core tenet of DCO is the notion that creative variation leads to higher performance.  This occurs because of a blindingly obvious fact: People are different.  Have you ever thought about how strange it is for a major retailer to target females aged 18-54 with the same exact ad?  Does anyone believe the 18 year-old high school senior in Beverly Hills, CA will respond in the same manner as the 54 year-old grandmother in New York City?  Of course not.
To use a food metaphor, everyone has their own favorite flavor of ice cream.  I love dulce de leche (or, at least, I did until I went vegan – but that’s another post).  The people sitting around me as I write this enjoy: chocalate, mint chocalate chip, cookies and cream, butter pecan, strawberry, pistacchio, coffee, etc.   (One guy kind of looked at me funny and said, “I don’t like ice cream.”  I’m not positive he was in fact referring to ice cream.)
Unfortunately for advertisers, the broadcast paradigm that has dominated advertising for over a century carries with it a limitation: everyone gets the same ad.  In our food metaphor, everyone has to eat the same ice cream.  Now if you’re an advertiser, and you can only provide your customers with a single flavor, the choice is more or less obvious.  Go with vanilla.  It’s not anyone’s favorite, but no one really objects to it.  In fact, the very word vanilla has become synonymous with bland and inoffensive.  If you are forced to pick a single flavor for everyone, vanilla is probably the least bad choice.
Enter online advertising and DCO, where (finally!) advertisers can provide customized creative!  Chocolate? No problem!  Dulce de leche? Fine!  Jalapeno mint chocolate chip?  Sure thing – want a double scoop?  Dynamic creative allows advertisers to have a virtually unlimited number of flavors at their command.   Wonderful – we can finally give everyone what they truly want.
Of course, there is no such thing as free flexibility.  I won’t get into a boring discussion of Bayesian Statistics, machine learning, Taguchi methods, genetic algorithms, and all the other topics that make geek propellers spin.  Instead, I’ll just ask you to trust me: too many flavors of ice cream can make it hard to figure out the perfect flavor for each individual.
So as we began to launch dynamic creative into the marketplace, I made a mental note to myself: “Be sure that advertisers understand that they should keep their creative variations within reason. We have to control them – don’t let them get too excited.”
And sure enough, as we went to market, I re-discovered that age old truth.  I had absolutely no idea what I was talking about.
“Look at what you can do! Isn’t this great?  You can have billions of ad variations!  You can find the perfect message for everyone!  You can have Jalapeno Mint Chocolate Chip!  Woohoo!”
And as the orders started to roll in, as dynamic creative started to run, I took a close look and noticed that something was amiss.  Given the power to create a virtually unlimited number of ad variations, the advertisers hadn’t created too many – they had created too few!  I didn’t see 10,000 crazy flavors of ice cream – I saw 8.  And every single one of them were minor variations of… vanilla.  Argh.
In retrospect, I shouldn’t have been surprised.  People are creatures of habit.  The danger was never that they would over-use the power of dynamic creative.  The real danger was that they would ignore it altogether, sticking to the ideas that had been drilled into their collective industry norms for over a century.  One identical ad for everyone…  the Curse of Vanilla.
Now that we’re aware of the issue, we’re taking steps to overcome it.  It’s not rocket science: we’re focusing on education and outreach, coupled with an emphasis on those executions where there are natural sources of creative variation – large numbers of products and services, large numbers of geographic locations, etc.
Lessons learned, or perhaps, re-learned:  You can’t predict the future; change is hard; or, as Robert Burns put it: “The best laid schemes o’ mice an’ men / Gang aft agley.”
Now, someone pass the sorbet!

Have you ever noticed that it’s the things you didn’t think of that always cause the most trouble?
I’ve been working on Dynamic Creative Optimization (DCO) programs for almost three years now.  And looking back, the issues that have cropped up over and over again are those that I had no clue were coming.
The one that I’d like to discuss today, I call the “Curse of Vanilla.”
A core tenet of DCO is the notion that creative variation leads to higher performance.  This occurs because of a blindingly obvious fact: People are different.  Have you ever thought about how strange it is for a major retailer to target females aged 18-54 with the same exact ad?  Does anyone believe the 18 year-old high school senior in Beverly Hills, CA will respond in the same manner as the 54 year-old grandmother in New York City?  Of course not.
To use a food metaphor, everyone has their own favorite flavor of ice cream.  I love dulce de leche (or, at least, I did until I went vegan – but that’s another post).  The people sitting around me as I write this enjoy: chocalate, mint chocalate chip, cookies and cream, butter pecan, strawberry, pistacchio, coffee, etc.   (One guy kind of looked at me funny and said, “I don’t like ice cream.”  I’m not positive he was in fact referring to ice cream.)
Unfortunately for advertisers, the broadcast paradigm that has dominated advertising for over a century carries with it a limitation: everyone gets the same ad.  In our food metaphor, everyone has to eat the same ice cream.  Now if you’re an advertiser, and you can only provide your customers with a single flavor, the choice is more or less obvious.  Go with vanilla.  It’s not anyone’s favorite, but no one really objects to it.  In fact, the very word vanilla has become synonymous with bland and inoffensive.  If you are forced to pick a single flavor for everyone, vanilla is probably the least bad choice.
Enter online advertising and DCO, where (finally!) advertisers can provide customized creative!  Chocolate? No problem!  Dulce de leche? Fine!  Jalapeno mint chocolate chip?  Sure thing – want a double scoop?  Dynamic creative allows advertisers to have a virtually unlimited number of flavors at their command.   Wonderful – we can finally give everyone what they truly want.
Of course, there is no such thing as free flexibility.  I won’t get into a boring discussion of Bayesian Statistics, machine learning, Taguchi methods, genetic algorithms, and all the other topics that make geek propellers spin.  Instead, I’ll just ask you to trust me: too many flavors of ice cream can make it hard to figure out the perfect flavor for each individual.
So as we began to launch dynamic creative into the marketplace, I made a mental note to myself: “Be sure that advertisers understand that they should keep their creative variations within reason. We have to control them – don’t let them get too excited.”
And sure enough, as we went to market, I re-discovered that age old truth.  I had absolutely no idea what I was talking about.
“Look at what you can do! Isn’t this great?  You can have billions of ad variations!  You can find the perfect message for everyone!  You can have Jalapeno Mint Chocolate Chip!  Woohoo!”
And as the orders started to roll in, as dynamic creative started to run, I took a close look and noticed that something was amiss.  Given the power to create a virtually unlimited number of ad variations, the advertisers hadn’t created too many – they had created too few!  I didn’t see 10,000 crazy flavors of ice cream – I saw 8.  And every single one of them were minor variations of… vanilla.  Argh.
In retrospect, I shouldn’t have been surprised.  People are creatures of habit.  The danger was never that they would over-use the power of dynamic creative.  The real danger was that they would ignore it altogether, sticking to the ideas that had been drilled into their collective industry norms for over a century.  One identical ad for everyone…  the Curse of Vanilla.
Now that we’re aware of the issue, we’re taking steps to overcome it.  It’s not rocket science: we’re focusing on education and outreach, coupled with an emphasis on those executions where there are natural sources of creative variation – large numbers of products and services, large numbers of geographic locations, etc.
Lessons learned, or perhaps, re-learned:  You can’t predict the future; change is hard; or, as Robert Burns put it: “The best laid schemes o’ mice an’ men / Gang aft agley.”
Now,meone pass the sorbet!

Have you ever noticed that it’s the things you didn’t think of that always cause the most trouble?

I’ve been working on Dynamic Creative Optimization (DCO) programs for almost three years.  And looking back, the issues that have cropped up over and over again are those that I had no clue were coming.

I’ve named the one I’d like to discuss today the “Curse of Vanilla.”

A core tenet of DCO is the notion that creative variation leads to higher performance.  This occurs because of a blindingly obvious fact: people are different.  Have you ever thought about how strange it is for a major retailer to target females aged 18-54 with the same ad?  Does anyone believe the 18 year-old high school senior in Beverly Hills, CA will respond in the same manner as the 54 year-old grandmother in New York City?  Of course not.

To use a food metaphor, everyone has their own favorite flavor of ice cream.  I love dulce de leche (or, at least, I did until I went vegan).  As I write this post, the people sitting around me enjoy: chocolate, mint chocolate chip, cookies and cream, butter pecan, strawberry, pistachio, coffee, etc.   (One guy kind of looked at me funny and said, “I don’t like ice cream.”  I’m not positive he was in fact referring to ice cream.)

Unfortunately for advertisers, the broadcast paradigm that has dominated advertising for over a century carries with it a limitation: everyone gets the same ad.  In our food metaphor, everyone has to eat the same ice cream.  Now if you’re an advertiser, and you can only provide your customers with a single flavor, the choice is more or less obvious.  Go with vanilla.  It’s rarely anyone’s favorite, but no one really objects to it.  In fact, the very word vanilla has become synonymous with bland and inoffensive.  If you are forced to pick a single flavor for everyone, vanilla is probably the least bad choice.

Enter online advertising and DCO, where (finally!) advertisers can provide customized creative!  Chocolate? No problem!  Dulce de leche? Fine!  Jalapeno mint chocolate chip?  Sure thing – want a double scoop?  Dynamic creative allows advertisers to have a virtually unlimited number of flavors at their command.   Wonderful – we can finally give everyone what they want.

Of course, there is no such thing as free flexibility.  I won’t get into a boring discussion of Bayesian Statistics, machine learning, Taguchi methods, genetic algorithms, and all the other topics that make geek propellers spin.  Instead, I’ll just ask you to trust me: too many flavors of ice cream can make it hard to figure out the perfect flavor for each individual.

So as we began to launch dynamic creative into the marketplace, I made a mental note to myself: “Be sure that advertisers understand that they should keep their creative variations within reason. We have to control them – don’t let them get too excited.”

And sure enough, as we went to market, I re-discovered that age old truth:  I had absolutely no idea what I was talking about.

“Look at what you can do! Isn’t this great?  You can have billions of ad variations!  You can find the perfect message for everyone!  You can have Jalapeno Mint Chocolate Chip!  Woohoo!”

And as the orders started to roll in, I took a close look and noticed that something was amiss.  Given the power to create a huge array of ad variations, the advertisers hadn’t created too many – they had created too few!  I didn’t see 10,000 crazy flavors of ice cream – I saw 8.  And every single one of them were minor variations of… vanilla.  Argh.

In retrospect, I shouldn’t have been surprised.  People are creatures of habit.  The danger was never that they would over-use the power of dynamic creative.  The real danger was that they would ignore it altogether, sticking to the ideas that had been drilled into their collective industry norms for over a century.  One identical ad for everyone…  the Curse of Vanilla.

Now that we’re aware of the issue, we’re taking steps to overcome it.  It’s not rocket science: we’re focusing on education and outreach, coupled with an emphasis on those executions where there are natural sources of creative variation – large numbers of products and services, large numbers of geographic locations, etc.

Lessons learned, or perhaps, re-learned:  You can’t predict the future; change is hard; or, as Robert Burns put it: “The best laid schemes o’ mice an’ men / Gang aft agley.”

Now, someone pass the sorbet!

ad:tech – let’s meet up!

Posted in General on April 20th, 2010 by petekim – Be the first to comment

I’m back from Mexico – if anyone is in town for ad:tech and wants to catch up tonight or tomorrow, give me a shout!

Consolidation in the optimization space picking up steam…

Posted in General on April 15th, 2010 by petekim – 2 Comments

Since uncooperative conditions (rain, thunderstorms) in Playa Del Carmen, Mexico have killed my golf game, I may as well throw up another blog post….

Story released yesterday:  Mediamath Acquires Adroit, Combines DSP with Dynamic Ads

Analysis: At first blush, this deal makes sense.  I haven’t seen performance data for MM, but assume their optimization algorithms are up to snuff, and will dovetail nicely with the dynamic creative and self-service UI chops of Adroit.

As with the Adobe/Omniture move into Level 2, the question will now be – can they execute?

Zooming up a level, the pace of consolidation seems to be accelerating.  In this case, Level 1 (media optimizer) acquires Level 2 (creative optimizer).  However, if you look at each of the three levels of media optimization, there are multiple recent examples of consolidation/integration:

1.    X+1 venturing into landing pages (level 1 to level 3)
2.    Adobe/Omniture announcing creative optimization (level 3 to level 2)
3.    MediaMath purchasing Adroit (level 1 to level 2)
4.    Adchemy (building levels 1, 2, and 3 from scratch)
5.    Aggregate Knowledge (building levels 1, 2, and 3 from scratch)
6.    Dapper (building levels 1 and 2 from scratch)
7.    Choicestream / MyBuys / other recommendation engines moving into advertising/creative (level 3 to level 2)
8.    … and more that are only rumors and/or under NDA at this point

There’s an interesting dynamic (no pun intended) evident in the quotes by CEO Zawadzki. On one hand, he stumps for the simplicity of a unified solution:

“Even agencies that recognize the value [...] are deterred by the complexity of navigating multiple technologies and making sense of the results. The combined Mediamath and Adroit systems could remove some of those obstacles, said Zawadzki.”

Yet in the same article he leaves the door open for working with other (now competing) creative technologies:

“‘We will and do continue to support other dynamic creative solutions,’ like Tumri and Teracent, stressed Zawadzki.”

The reason why such seeming double-talk is necessary is because it has not yet become clear whether integration via partnership or consolidation/unification via acquisition (or internal build) will become the dominant trend among the various participants in the display optimization landscape.  Consolidation yields simplicity and likely, in my humble opinion, better performance.  Integration allows flexibility of partner choice – enabling the mixing and matching of participants from each of the individual optimization tiers.

My bet, and apparently the bet of companies like MediaMath and others mentioned above, is that simplicity and performance will eventually carry the day over flexibility.  It will be interesting to see how this plays out.

Ok, enough ad tech talk – back to vacation!

(PS – on a more personal note, congratulations to Dan Faga and Vivianna Padilla on their impending nuptials!)

Even worse than anticipated: USPS projects deficit of $100-200 Billion over next 10 years

Posted in General on March 3rd, 2010 by petekim – Be the first to comment

Wow.  Courtesy of the official blog of the USPS Inspector General:

“The Path Forward” of the Postal Service

The money quote:

“The Postmaster General warned that if the Postal Service continues to operate as it is, it will run a cumulative debt of $238 billion over the next 10 years. Even if the Postal Service institutes every conceivable control within management control – product and service actions, productivity improvements, workforce flexibility improvements and purchasing savings – it can only shrink the debt to $115 billion.”

Adobe enters dynamic ad market

Posted in General on March 3rd, 2010 by petekim – 3 Comments

The longstanding rumors about Adobe/Omniture entering the dynamic ad space came true today.

Omniture Announces Display Advertising Solutions for Increased Return on Ad Spend

Interesting on a number of levels:

1. As discussed, we see a continuing trend of optimization companies migrating into adjacent segments.  In this case, a L3 player has migrated into L2.  (Did ADBE just become a potential buyer of DSP technologies? That would complete their stack.)  Strategically, I like it – execution will now be the key.

2. Adobe/Omniture is much larger than most other participants in this space, the principal exception being Teracent/Google.   Though they’re new in the space, expect them to leverage existing relationships to sell these additional services through.

3. The potential impact this may have on Flash as the de facto standard for dynamically generated ads.

Should the US Government Bail Out the Post Office?

Posted in General on March 2nd, 2010 by petekim – 1 Comment

Thanks to GH for forwarding the following story this morning:

USPS to propose 5-day mail schedule, major cuts

The most important quote:

“USPS posted a $3.8 billion loss in its 2009 fiscal year, the latest in a multiyear string of whopping losses. Mail volume was down 12.7% for the year, a trend the agency expects to continue over the next decade as more consumers opt for online bill payments and message delivery.  The Post Office was $10 billion in debt as of Sept. 30 — not far off from its $15 billion debt limit, which the agency expects to hit in its 2011 fiscal year.”

The USPS will require a bailout, count on it.   Is that a good idea?

Reasons that support a bailout:

1. The USPS is the second largest employer in the United States (behind Wal-mart) – a bailout would preserve much-needed jobs in an economy that is still on less-than-solid footing.

2.  ? (I’m sure there are other reasons that I’m missing..)

Reasons against a bailout:

1. Bailout money will be used to hold down postage prices for junk mailers – some of the largest companies in the world.  (Over 50% of mail volume is junk mail.)

2. Bailout money will increase and prolong the environmental impacts of junk mail.  (I’m not a dyed-in-the-wool green, but even I know that printing credit card offers on felled trees, and flying them all over the country is bad for the environment.)

We should also note the difference between a bailout and a permanent subsidy.  It’s probably a good idea to use government funds to cushion and ease the exit of an information distribution medium that, in the words of the Postmaster General, represents a “macro shift in society.”  It’s a terrible idea to use governments funds to artificially prop up a failing industry ad infinitum.


US Postal Service and Lottery Tickets

Posted in General on March 1st, 2010 by petekim – Be the first to comment

The USPS is becomingly increasingly, ahem, creative in their effort to close their multi-billion dollar budget gap. According to this post on the official blog of the US Postal Service Inspector General, the Postal Regulatory Commission is considering a USPS lottery as a potential source of much-needed funds.

Could Longer Lines Be Coming to Your Local Post Office…Lottery Lines?

I don’t know about you, but the phrase “grasping at straws” seems somehow appropriate…

Smart Ads Video

Posted in General on February 7th, 2010 by petekim – 1 Comment

The incomparable Mitch Spolan, Yahoo!’s Vice President of North American Sales, put together this video for my product, Yahoo! Smart Ads. Mitch did the entire thing himself, and I have no idea where he finds the time. It’s not like hitting sales goals that have little words like ‘billion’ in them can be easy on the schedule. Anyway, hope you like the video. Thanks, Mitch!

Philly Fight Night 2010 – February 27

Posted in General on February 7th, 2010 by petekim – Be the first to comment

What do you get when you combine a student-run charity boxing match, the number one boxing venue in the world, 1600 cheering Wharton MBAs, and a little bit of intra-school rivalry thrown in? Philly Fight Night!

Since its inaugural event in 2005, Philly Fight Night has become one of the biggest events of the year for Wharton students. Rather than describe the event in even more detail, I’ll just link to this year’s promo video instead. You’ll get the picture.

Credit where it’s due: This event was the brainchild of David Birnbaum, Schuyler Coppedge, and R.T. Arnold, with the inimitable Greg Battle serving as the evening’s MC . I’m proud to say that all of these fine gentlemen were my classmates in the Wharton MBA Class of 2005. Five years later, and still going strong. Nice work, guys!

Teachers Unions

Posted in General on January 31st, 2010 by petekim – Be the first to comment

I’ve stayed mostly away from political discussion in this blog, but I suppose it was inevitable that we got there eventually….

Like many of my peers, I consider myself to be a social liberal and and a fiscal conservative. I choose to vote most often with the Democratic party because, generally speaking, questions around social values are the political issues that I personally care more about. (That said, fiscal concerns are making a strong comeback with me considering the current sad state of our economy.)

However, as with any coalition, there are elements of the Democratic party with whom I just cannot reconcile my own beliefs. Top of that list for me? Teachers Unions.

I read this New Yorker piece while on vacation in the Far East. It made me so angry that I wanted to fly back and start beating people about the head.

And then this afternoon, I read this opinion piece by Joe Klein.

I am struggling to keep an open mind here. Maybe these unions are being unfairly maligned…?

But I’ll tell you what – if half this stuff is true, then I’m ashamed to associate myself with these people. As incomprehensible as they are to me, at least the right-wing religious conservatives can claim that they are taking a stand for something they believe to be right. Whether or not I agree with them, I can at least respect their fidelity to principle – something that is increasingly hard for me to say about the apparent naked and destructive self-interest of the teachers unions.

So – members of the teachers unions out there – please, set me straight. What am I missing? Why is there more than meets the eye here?