Posts from — October 2007

With a jaundiced eye, one could look at such crowdsourcing vehicles like Amazon’s Mechanical Turk and see its future as nothing more than a digitized sweatshop. The potential dark side of crowdsourcing is one that Jeff Howe at his Crowdsourcing blog touches on in his latest post. The concepts of digital peer production and crowdsourcing are still in their early, and somewhat idealistic, years. I’m pretty sure programmers associated with the early development of Linux did not feel like they were being exploited, but an outside observer could only wonder why would someone spend hours of work on something for which they would have no promise of receiving a monetary reward. The simple fact is that most crowdsourcing today is fairly benign and beneficial to its participants because the primary award is participation itself, along with some form of recognition. Unfortunately, that’s not to say that the future doesn’t hold something less idealistic when things come down to dollars and cents. While the Internet gave us all the benefits of email and the Web, it also gave us the negative developments of spam and identity theft from those looking to exploit the new systems.

Someone once said, “You are what you eat”. Is there are corollary in consumer goods that you are what you buy? I found this interesting study on product design by Ruth Mugge of Delft University of Technology while trolling through some posts on Advertising Lab. A summary of past research she references suggests “that consumers become attached to certain products, because they convey a personal and special meaning over and above the product’s utilitarian meaning”. This attachment is driven by four distinct factors:

1. Pleasure (provided by the product)
2. Memories (related to the product)
3. Group affiliation (does ownership of the product connect me to a group?)
4. Self-expression (can I distinguish myself from others with the product?)

What this becomes is an attachment gauge for brands and products. I’ve ordered the list so that it starts with the most basic, functional part of product attachment, which is how much pleasure it brings us. At this level, we buy products that taste good or that take care of a need (think potato chips or toothpaste). The next level is reserved for brands that are associated with positive moments in our past or have something of a personal heritage to them (i.e., the potato chips I got when visiting my grandparents or the toothpaste I’ve used since I was a kid). The last two categories belong to brands or products that move from the personal to the social. The products provide some form of social validation, by either attaching oneself to a group or separating oneself from another. These products could range anywhere from an NFL team jacket to an iMac to a Lexus sedan.

Obviously, brands or products that can move up the list can lever powerful emotional connections in their advertising that more functional brands or products cannot. Would you rather get a Lexus for Christmas or your father’s Oldsmobile?

In the “Wisdom of Crowds”, Jame Surowiecki writes about decision or prediction markets as tools for aggregating opinions through a form of speculation or betting. They can be used to predict many future events, including quarterly sales of a new product, presidential candidates, or who will win the best actress Oscar.

As I’ve tried to come up to speed on prediction markets and their application to marketing research, I’ve come across several resources that have been very useful. Two blogs in particular stand out as being very topical and focused on prediction market applications, rather than just theory:

1. Mercury’s Blog: this blog by Jed Christiansen has a couple of excellent videos that can quickly bring anyone up to speed on prediction market applications.
2. Midas Oracle: more like a community focused on prediction markets, rather than just a single blog. Lots of interesting posts.

Consumers don’t spend a lot of time thinking about brands. Certainly not as much as marketers in consumer packaged goods would like them to. Consequently, a consumers’ search behavior focuses more on the behaviors, attitudes and need states a brand is associated with, versus anything about the actual brand. Matt Wilburn from Yahoo describes it this way in a research summary posted on Mediapost’s Online Media Daily:

“It’s not just consumers searching for brand names,” added Wilburn. “That does happen, but this research illustrates that consumers are meeting a broader set of needs through search. They’re searching for products that can answer lifestyle questions like ‘I have hair loss,’ or ‘I’m fighting wrinkles’–and life stage questions like ‘I’m going away to college’ or ‘I just bought a home.’”

Sometimes the most important consumer insights about a brand can be heard by listening to the side conversations about the needs and wants that set the context for the brand, rather than waiting to hear what consumers say about the brand itself.

Two big numbers jump out from this press release from comScore on the worldwide scope of search: 750 million people conducted 61 billion searches in the month of August. These data points from their qSearch 2.0 service show just how big the Database of Intentions has grown and subsequently the potential of search to gain insights into human behavior and belief systems.

Where has random digit dialing for survey data gone? Gone with the “Do Not Call” lists. How about mall intercept or even door-to-door. Gone with a shake of the head or a slam of the door. No wonder market researchers have moved a great deal of research to online panels. According to findings from the Online Research Barometer survey (probably done online) as reported at Marketing Vox, 82% of market researchers surveyed said they would increase their use of online research panels in the upcoming year. The only thing surprising about those findings was the fact that there was that 18% group who would not be increasing their online panel usage. Were they planning on the return of mail panel surveys?

It’s not often that Apple is seen taking a stumble, which is why the recent brushfire over the iPhone price drop was quite unique. Steve Jobs belated gesture of $100 back to earlier iPhone adopters quieted things down a bit until the whole 1.1.1 update “iBrick” thing set things off again. At the same time last summer, Jobs was in a tiff with NBC over iTunes pricing of their TV shows, which led to the dropping of NBC new shows from the service.

From his article last month in the New York Times, David Carr compares these two events, and puts a sharp point on inconsistencies how Apple prices and consequently places value on the products and services it markets. The point that both a song by Kanye West and a Barry Manilow/Lily Tomlin duet can’t be worth the same 99 cents is apt. The recent price drop for DRM free recordings is a move in the right direction, since DRM provides no value to consumers.

Dynamic pricing is something that Web 2.0 should be bringing to consumers. What is strange is that one of the most dynamic technology innovators of this latest tech boom is still clinging to a Web 1.0 pricing model.

In this iMedia post on Behavioral Targeting, Dilip DaSilva rightly calls out the problem of behavioral targeting focusing too much on product specific behaviors or attitudes. Many web advertisers assume that the key behaviors they need to target are related to the purchasing or consuming of their product. However, sometimes it is and sometimes it is not.

Take the hypothetical case of the core Duff Beer consumer, Homer Simpson. The most “relevant environments” for purchasing or consuming his product are liquor stores, bars, and in the majority of times, in front of the TV. However, Homer’s “relevant environment” for online behavior may in fact be a couple of degrees removed. If you knew that the core Duff Beer consumer generally checks sports sites on Sunday evening to see how his fantasy football players did, you’d probably want your ad somewhere in that “relevant environment”, even though it is only tangentially related to buying or drinking beer.

In this Ad Age interview with Sue Davidson at Atmosphere BBDO, we get several good best practices on setting up online media for success and then measuring results on the back end. These include starting with a business objective, rather than a media tactic: “I want increase brand awareness 90%” versus “I want to do an online promotion”. Another example is moving beyond such necessary but ultimately soulless measures as time-on-site, and move to tracking changing consumer attitudes and beliefs through surveys.

All good stuff, but as someone who has worked on the client side for a long time, I wonder how healthy it is to have both your creative team and the analytic team that reports on how well the creative team did, both on the same side of the table. I’m sure at some level objectivity rules, however I think I know which way the report will lean if the data isn’t so black and white. This is not to fault BBDO: if their clients don’t bring their own digital insights and analytic power to the table, then they are providing an invaluable service. However, the marketing teams that have their own marketing research group minding their digital ROI will always have a their own advantage, this time on their side of the table.

So everything we know about fat and dieting is wrong. At least that’s what this article in the NY Times asserts, and its historical overview of how we got there is fascinating. Basically, the demonization of fat in our diets by the scientific community turns out to have been based upon dubious data and has never been corroborated by subsequent research. But because scientists were simply agreeing with other scientists without doing their own fact checking, something what social scientists call a “cascade” formed. And then the American public bought billions of dollars of low-fat foods and they still gained weight and had heart attacks.

As a big fan of James Surowiecki’s “The Wisdom of Crowds”, I’m interested in how this article shows how supposedly wise crowds can also be dumb. In “The Wisdom of Crowds”, one of the key features of a wise crowd is independence of thought and opinion. Cascades form when something called “social proof” occurs, which Surowiecki defines as “the tendency to assume that if lots of people are doing something or believe something, there must be a good reason why”. Especially if those people are respected scientific peers.

A bigger question is whether our modern distributed data networks (i.e., the Web) will make cascades such as this more prevalent, or will diminish their longevity. My thought is that both are likely to occur. The fact that Web 2.0 diminishes the sole authority of “experts” (see Wikipedia) and mitigates the penalties for disagreement (read for peaceful conformity in any forum or chat room), the potential for false cascades to run unchecked for a long time will drop. However, the same medium which mitigates the length of false cascades also allows them to start fairly quickly to begin with.

Cascades are something that all who are interested in consumer trends, insights, and beliefs need to be aware of and be ready to react to. Especially if you want to know how long consumers will be interested in that low-fat cookie you have on the shelves.