- Interested in advertising on ‘thefacebook’…
- Hired!
- The Boy Who Followed Somebody Else’s Dream (The Dream Comes Around)
- “Try not to embarrass yourself”
- “Your voicemail is so annoying…”
- @GusLage i agree.
- @gregrhenderson do they really break? i've dropped it plenty.
- @cesaradominguez I keep mine in a suit pocket or my jeans, and have never had a scratch (knockonwood)
- @psu_chris good point Chris! (or the Star-Tac of 1998!) :)
- @kylejomo i havent had that problem (yet) wonder if I'm just lucky.
- @NovaCat91 i've dropped mine a ton, never had that problem. (yet).
- @akor17 funny. :)
- Why do people buy iPhone cases? Seems like a waste and just more bulk to drag around. Didn't Jobs demand iPhone be made from sturdy stuff?
- Congratulations @UrgentSpeed & Lenddo team on the $8 mm round. For-profit is the right model in micro-finance. Huge idea & right guy.
- Remarkable. Fifty photos from fifty years ago... http://t.co/18r7HnUz
Recent Posts
Latest Tweets »
The Digital Soul, XV: The Rule of Two Magnitudes
April 24, 2011 | (6 Comments)
Physical goods businesses reliably transform raw materials into finished products. A chair manufacturer knows that a chunk of wood and a dozen nails will always turn into a chair and a pile of wood shavings, time after time after time, when processed by her factory. The outcome is known and deterministic, the maker makes no guesses in the manufacturing processes about what the end result will be.
Some information goods share this characteristic. Buying a movie ticket reliably leads to your sitting in a darkened theater while each minute of movie is displayed, in the order intended, on the screen in front of you. Putting your money into an insured savings account yields a set number of pennies in return, day after day after day. Purchasing Malcolm Gladwell’s Outliers in hardcover provides precisely 309 pages of book, from which the application of your literacy to each page in succession will yield a comprehension of his argument.
Although immaterial and intangible, each of these information goods reliably produce the outcome their maker intended — tale told, pennies gained, or thesis understood. And just as you may have a positive or negative experience with an Aeron, how you may feel about The King’s Speech, Outliers, or your ever-incrementing bank account, will vary. The producer of the experience will endeavor to ensure your chair-sitting or movie-watching enjoyment, but can not control your behavior or reaction directly, only indirectly via your experience of the product put out by his factory or bank or studio or publishing house.
Behavior goods are very different.
Behavior goods are those that intend to cause a behavior on the part of the user.
In the 20th century, offline, world we understood behavior goods to be advertising, sales or marketing channels. Newspapers intended to cause their users to visit the local car dealer, or to ring up the local real estate agent showing the nice home on Walnut Street, or to shop till they dropped at the Spring Savings Event. Similarly, the matchmaker intended for an appropriate opposite number to meet with you in person, the Tupperware host wanted you to feel comfortably social while buying at the party, and WeightWatchers hoped your meeting attendance would drive weight loss.
While each of these 20th century businesses comprehensively addressed their users’ behavior from expression of intent to cusp of actual purchase, it was nonetheless the case that the Tupperware party, the newspaper circular, or “It’s Just Lunch” could present the opportunity to buy, but not calculate the outcome of each individual’s behavior in response.
What this means is that behavior goods businesses — the companies involved in creating and channeling demand — “manufacture” user behavior probabilistically.
In contrast to the physical goods manufacturer, who knows that one unit of raw materials will yield one unit of output, 10 units input will yield 10 units output, 100 will yield 100, the behavior goods “manufacturer” knows the ratio of input to output, at best, on average. He knows that publishing newspapers, setting up dates, or throwing product parties, will lead to the desired user behavior, but only in aggregate. He can not count on any particular user to buy, can not predict the likelihood of any specific transaction coming to pass.
The Internet has engendered a revolution in these behavior goods businesses. In the online world, the steps between passing thought and purchase have been articulated into the web hierarchy of: information – intentions – inferences – interactions. We are presently witnessing the rise of enormous businesses to address each of these discrete steps, with Google being only the most prominent of many to come.
What we observe is that for a given number of instances of user behavior at one layer, there is an expected, but not determined, outcome in the subsequent layer.
To walk through, step-by-step, what this means is that a user consuming information will, after a time, have intentions about what to purchase. A user expressing a number of intentions, opinions, or preferences, will after a time, enable the generation of inferences about what else he might like. And the expression of intentions and generation of inferences will, after a time, imply interactions in which the user may be interested.
Concretely, we observe the following user behavior:
- Scanning “a number” of search results on Google produces a click on an ad
- Reading “a number” of job listings generates one application
- Skimming “a number” of profiles on a dating site leads to one email from a suitor
- Replying to “a number” of suitor emails leads, on average, to one meeting in person
- Reviewing “a number” of restaurant listings, movie possibilities, shows available on Hulu, or flight options leads, in each case, to one action taken: a dinner, a show, a trip.
For each of the above domains, I suspect that the number represented by the placeholder “a number” is two magnitudes, as we are already seeing in the available data:
- For Google, every 100 displays of an ad leads to 1 click
- For TheLadders, every 1,000,000 page views lead to 10,000 job applications
- For online publishers, every 1,000 pageviews yields 10 clicks
This Rule of Two Magnitudes is important for understanding the nature of the web factory: what we must and what we mustn’t do in order to successfully generate behaviors from and for our users.
Behavioral economics will have more to say as we seek to better understand the transition of behavior goods businesses online. As further data from disparate sources are published, I suspect the possible ranges for the Rule are either tight (10^1.75 to 10^2.25) or loose (10^1.5 to 10^2.5), that is, between 50 and 175 if tight and between 30 and 300 if loose.
These numbers will vary by domain, perhaps widely, but what they share in common is the implied requirement for sustained user attention. We know that if humans are to grasp dozens or hundreds of data points at a time, they require sustained focus in a contextually relevant environment.
Thus the most salient implication of the Rule of Two Magnitudes will be that firms on the internet can not simultaneously conduct business in two or more layers successfully.




Pingback: Why does Yahoo! fail? | Recruiting & Job Search, New York City Start-up blog by Marc Cenedella
Pingback: How to handicap the Groupon vs. Facebook battle | Recruiting & Job Search, New York City Start-up blog by Marc Cenedella