Previous month:
December 2016
Next month:
February 2017

January 2017

Amazon Books, Amazon Prime

Jf17techreview(Photo source: The January/February issue of MIT Technology Review has an article by Nicholas Carr about Amazon's foray into brick-and-mortar bookstores, called Amazon Books. The plan for such bookstores was first announced in late 2015, for instance in this New York Times article (the NYT also has a video about it here) as well in that one. It isn't apparently viewed as a return to traditional bookselling but rather as a "physical extension of", where physical bookstores play the role of showrooms for the books. The journalist adds that "The retailer, which is spending billions to refine and speed up home delivery, suggested that some people would come to the store to look at the books and then go home and order them online."

According to Carr in his MIT TR article, though, "[ CEO Jeff] Bezos underestimated the allure of bricks and paper." We learn that "in the mainstream trade-book market, e-book revenues dropped 11 percent in 2015 alone" while "sales in bookstores grew 2.5 percent in 2015." He also wonders whether the launch of Amazon Books is "the vanguard of a much broader push into brick-and-mortar retailing by the company," that will include convenience stores to boost its grocery business (his source is this WSJ article). Such a move would be well-aligned with current retailing trends toward omnichannel strategies, "which blend physical stores, Web stores and mobile apps in a way that makes the most of the convenience of smartphones and overcomes their limitations." Could be a future multichannel behemoth? 

In the retail location in Seattle, as described by Carr, "all the books are displayed with their covers facing outward" and "beneath each volume is a small placard that displays the book's Amazon star rating - only books that have earned at least four stars from Web buyers are stocked - and that also includes a brief excerpt from a customer review." 

First, there are obvious issues in selecting books based on the Amazon star rating. I'll get the issue of fake reviews out of the way by saying that only the reviews of Amazon customers with verified purchase (meaning they purchased the product on Amazon) should be included in the computation of the star rating displayed at Amazon Books. (I also think that should offer two ratings online, the one based on verified purchases only and the one including all purchases, because the fact that you didn't buy the product on doesn't mean you didn't really buy it.)

Further, the reason bookstores employ book buyers is that those people have a deep understanding of their store's customers, what sells and what doesn't in their market, and books those customers might not have heard from on national media but would love reading. And of course the book buyers at chains like Barnes & Noble or the defunct Borders may not all have upheld this ideal vision of the local professional with his ear on the ground, finely attuned to the interests of his store's frequent customers, but you get the idea. Should the books in the Amazon Books retail space be selected only based on the ratings of the Amazon customers in that area? Should the books on display be of the "popular but not very popular" kind, no matter the star rating, to raise awareness of certain books? Isn't the point to discover new books? currently displays in one long row books that other people who purchased this book have also bought, sometimes with comical results when someone buys together books for very different purposes, such as a business book for work and a book of, say, vegetarian recipes. The fact is, we often don't need to scroll through 20 pages of other people's purchases to find books we like. Often those books can be clustered according to themes or authors, and of course should distinguish between books very often bought with this one and books bought once out of one thousand times with this one. What is curious is that bought Goodreads back in 2013 (read more about this here and there) and Goodreads's recommendations system -- based on the ratings users give to books they've read -- is massively more sophisticated than

Why doesn't Amazon try to provide better-quality suggestions? It is almost as if it didn't know what to do with the troves of data it has so went ahead with the simplest way to use the data, to do something. But the Goodreads system, right in their backyard, actually seems very good. The Amazon and Goodreads systems don't seem to be "talking" to each other since Goodreads is recommending books I've actually already bought on Amazon but haven't put into Goodreads yet. If that's not data going to waste, I don't know what is. (On the other hand, it's an easy argument in favor of the Goodreads recommendation system, since it recommends books I was indeed interested in enough to buy.)

Finally, Amazon Books's differentiated pricing is worth pointing out: Amazon Prime members pay Amazon's discounted online price for a book, while non-members pay the list price. Unless non-Prime members struggle with delayed gratification (and since they willingly choose not to have the free 2-day shipping that comes with the paid membership and instead wait 3-4 days for their online purchases to arrive, they probably don't), it therefore seems like a strategic move by to push more customers into the Amazon Prime membership. Over recent weeks, I've noticed that some items I was once able to buy, for instance Green Mountain Coffee Half-Caff Keurig K-Cups, 72 count, are now reserved for Prime members only. If one day foregoes online discounts for non-Prime members, it might not be completely unthinkable that book lovers would order online elsewhere. I think Amazon's advantage at the moment is that their online payment and delivery system seems more robust, but if publishers developed their own payment systems, then direct purchasing from publishers could emerge as a serious alternative to's growing hegemony. (I wish there were a buzz of excitement surrounding B&N's strategy, talk of an innovative approach etc, but really B&N seems to just be plodding along, except in redesigned stores that focus on games and K-12 education, in a push to reposition itself as a "lifestyle brand". You can read about that here.)

In summary, what I got from this article is that the Amazon Books retail locations come across as glorified airport bookstores and that Amazon is finding new ways to push for Prime membership. 

PS: According to MIT Technology Review, has opened two more bookstores in San Diego and Portland, OR, and has plans for Chicago and Boston. One can only hope Barnes & Noble mounts a massive campaign of top-notch analytics, high-quality book-buying and stellar community events to defend its turf (because it is naive to expect's online discounts to continue, especially to non-Prime members, on the same scale as before if Barnes & Noble goes out of business and customers have few other choices to get their books). I was surprised to read that would venture into the Boston area, which has thriving indie bookstores in Brookline (Brookline Booksmith) and Cambridge (Harvard Bookstore), but it turns out Amazon Books's retail location is in Dedham, about 20 miles and half an hour away. 

PPS: If you're getting really bored and are reading this on a laptop or tablet, flip it upside down and see a different cover image appear for the January/February issue of MIT Technology Review

PPPS: Two unrelated articles I liked in the same issue of MIT Technology Review are The Frontline Reporter: Dickey Chapelle '39 blazed a trail for combat photojournalists and Obama's Stand-up Economist (For Austan Goolsbee, PhD'95, doing his bit for the country meant helping tackle the recession - and performing a little comedy on the side).

NYT on America's Great Working-Class Colleges

NYTSundayReviewDavid Leonhardt, who gave the Tate Lecture at SMU back in November, wrote an interesting op-ed in this past Sunday's New York Times, using a recent study that tracked students from nearly every college in the country and measured their earnings after they left campus. He argues that such colleges continue to do a strong job pushing students into the middle class and beyond, in spite of a challenging environment marked by state budget cuts, more unprepared students and high dropout rate.

He provides the following statistics, among others:

  • Among people born in 1980 who attended the City University of New York, 30% of the students were in the bottom 20% of the income distribution as children, but only 14% were at that bottom 20% as adults. 
  • "At City College, in Manhattan, 76 percent of students who enrolled in the late 1990s and came from families in the bottom fifth of the income distribution have ended up in the top three-fifths of the distribution."
  • "The equivalent number at the University of Texas, El Paso, is 71 percent. At California State University in Bakersfield, it’s 81 percent. At Stony Brook University, on Long Island, it’s 78 percent, and at Baruch College in Manhattan, it’s 79 percent."

For fun, the NYT has an interactive page where you can learn about the upward mobility at any college. For instance (I'm not sure what to make of the fact that exactly 67% of the students from Harvard, Lehigh and SMU come the top 20 percent, so the results are to be taken with a grain of salt),

  • "The median family income of a student from Harvard is $168,800, and 67% come from the top 20 percent. About 1.8% of students at Harvard came from a poor family but became a rich adult.
  • "The median family income of a student from Lehigh is $167,600, and 67% come from the top 20 percent. About 1.9% of students at Lehigh came from a poor family but became a rich adult."
  • "The median family income of a student from Southern Methodist is $198,900, and 67% come from the top 20 percent. About 1.8% of students at Southern Methodist came from a poor family but became a rich adult."
  • "The median family income of a student from University of Texas at Dallas is $89,800, and 40% come from the top 20 percent. About 2.4% of students at University of Texas at Dallas came from a poor family but became a rich adult."

Many students on what Leonhardt calls working-class college campuses don't resemble the typical Ivy League student. There, "students often work while they’re going to college. Some are military veterans, others learned English as a second language and others are in their mid-20s or 30s." OpenIDEO recently had a challenge about higher education, entitled "How might we reimagine the cost of college in the U.S. and how it's paid for". I didn't participate, but I think it's key to realize the difference between parents paying for their children's education, as is the case at most Ivy League colleges, and individuals paying for their own undergraduate education, as is the case in "working-class college campuses" with older students already well into adulthood. The unspoken assumption when parents pay for their children's education is that most of them have been saving for it since they've had two decades in the professional world to prepare for their kid going away to college. That's just not the case when a student is in his 20s or 30s, doesn't have a college degree yet, and his parents can't afford helping him pay for it (if they could, they probably would have sent him to college at age 18.)

Leonhardt also has impressive charts about elite colleges being packed with rich students. This helps explain why "working-class colleges have become vastly larger engines of social mobility." Yet, the importance of investing in public education to foster a thriving economy has become increasingly challenged over recent years, with sharp declines in state funding (the movie Starving the Beast is worth seeing if you haven't already done so, see my post here). Offering lower-income students a shot at excellence, social progress and increased opportunity is a critical goal for a nation that wants all its citizens, and not only the wealthy ones, to become the best they can be.  

Jim Schutze from Dallas Observer on School Accountability

Logo-dallasobserverToday, I came across a thought-provoking, well-written blog post by Jim Schutze of the Dallas Observer entitled "Mediocrity Lobby Angry Because Grades for Schools Expose Their Incompetence", published last week. In it, Schutze argues that students who are given mediocre education in school are more likely to end up in prison, and therefore we should all be concerned with improving public education.

In particular, he gives the following stats: "Two-thirds of prison inmates in this country lack high school diplomas. All black men between the ages of 20 and 24 have a greater chance of being locked up than of having a job. Meanwhile, research has found that a 10 percent increase in high school graduation rates can produce a 9 percent decrease in crime rates."

He argues that some measure of accountability is needed to make sure taxpayers' money is "tightly and strategically focused on achievement", describes the "trial version of a new letter-grade system for schools based on a whole matrix of measurements" that the Texas Education Agency is rolling out, touches upon powerful themes such as the unforgiveness of the world toward people who can't read or do math, as well as the machinery of envy where poor people are bombarded with a "relentless barrage of wealth and glamour", takes issue with the reaction of a specific school district to the Texas Education Agency's efforts to introduce this grade-based evaluation system, which said school district apparently branded as an attempt to destroy public education, and finishes with reflections on the attitude of Texan superintendents, who he says have an average base salary of $350,000 a year, when confronted with results that suggest a large number of their students are headed for failure.

Schutze's writing is very powerful so you really have to go and read his post rather than whatever summary I can try to make of it (if you don't have time to read the whole post, the last four paragraphs are particularly sharp). I'd love to know what you think. 

Engineering to the rescue of threatened art (from the New Yorker)

IMG_4820 Back in late November the New Yorker ran an article by Daniel Zalewski on The Factory of Fakes: How a workshop uses digital technology to craft perfect copies of threatened art, especially from Iraq and Egypt. (I only got around to reading it after the semester at SMU was over.) The digital tech company in question is Factum, led by Adam Lowe. The New Yorker article is full of examples of art now endangered or destroyed, for which the company has been able to recreate very high-quality facsimiles. Studying the pictures has even made possible some new discoveries, such as ridges in Tutankhamun's chamber. They could mark the entrance to the tomb of Queen Nefertiti, which has never been found. Factum's website also offers images from the facsimile of Tutankhamun's tomb, one of its flagship projects - he issue with the original tomb is that it was sealed before the painting could properly dry, because Tutankhamun had died suddenly, allowing bacteria to foster on the walls. 

My favorite part in the article was Factum's role in creating replicas of destroyed or stolen work, such as a winged-lion statue that had stood in Nimrud before the city was destroyed in 2015 or the "Nativity with St Francis and St Lawrence" painting by Caravaggio, which was stolen by the Mob from a chapel in Palermo in 1969. Factum "made its reputation in 2007, with a replica of Paolo Veronese's monumental painting "The Wedding at Cana," which Napoleon presented to a new museum, the Louvre, after ripping it off the wall of a refectory in Venice in 1797."

The high-resolution technology allows scholars and tourists to study details of the work that they would not otherwise notice (tourists aren't allowed to stay more than about fifteen minutes in the real King Tut tomb, for instance). Where the quality of hand-crafted copies depends on the copier's eye for detail, Factum's approach captures all the details of a work, exactly the way its author experienced it, down to the same brushstrokes, even showing the spots where the painter was running out of paint. This doesn't come easily or cheap: it took seven weeks to scan King Tut's tomb, and  "the cost of scanning and rematerializing [=recreating] the Seti I tomb [another tomb in Luxor] will likely approach twenty million dollars". But it makes accessible phenomenal art that we would be deprived of otherwise. I enjoyed reading about Lowe's business model, which apparently involves getting enough commissions from big shots such as  Anish Kapoor, Maya Lin and Marc Quinn, for whom he fabricates sculptures, to fund his true passion of art preservation. 

It is a powerful testimony to the role high tech and engineering can play in bringing art back to the public and ushering conservation into a new area.

Another great article about Factum was published in 2013 by the Financial Times (Conservation: Factum remaking history), with more pictures of the facsimile of Veronese's "Wedding at Cana" and of the church of San Petronio, where Factum 3D-scanned and replicated the statue of the city's patron saint.