How to interview for a Testing / QA engineer role? How to test a book?

If you're applying for the role of a SDET (software design engineer in test), STE (software test engineer) or QA Engineer (quality assurance engineer), then the interviewer will invariably ask you to test "something". Recently, I've spent some time brainstorming a few testing interview questions with a few friends who are applying for a QA role. Based on my personal experience both as an interviewer and an interviewee, here's how I would answer a typical "testing" question:

How would you test a book?

Functionality:A book can be categorized functionally into two primary categories:

  • Book to Read
    • Test if the printed words on the cover and inside the book are legible and in accordance with the specifications on every page.
  • Book to write
    • Test if all pages have enough friction, so that it can be easily written/drawn onto them with a regular writing device.



  • Check to see if the papers are of the right quality
  • Check to see if the papers are of right thickness
  • Check to see if the paper’s transparency is not too much so that one can read the book without seeing the words in the other page.
  • Test if the book can be held in hand or set atop a tabletop, in such a manner that the user can open it and flip through pages.
  • Check to see if the pages tend to stick together at any point
  • Check to see if any of the pages are sticking out. This could be caused due to incorrect cutting machines
  • Test, if the binding of pages and the cover is strong enough to hold the pages together, but flexible enough for the user to freely fold the book as per convenience.
  • Make sure the pages do not have any obnoxious smell. The book should smell new, if it is coming out of a printing press
  • Test for the acidity of pages, and ensure that it complies with regulatory standards.
  • Check to see if the printing has caused any ink smudges
  • Check to see if the printing has caused any aliasing of characters
  • Check the weight of the book. For some grade school text books lower weight is a premium concern.
  • Check the physical dimensions of the book like length, width, thickness etc.


  • Test if the index section (a.k.a. chapter listing) is present and whether it’s matching to the specification, if given.
  • Test if the book has a glossary of important terms
  • Flip through the pages to see if the indexing of the chapter page numbers corresponds to the actual chapter beginnings. This is the main functionality of a chapter listing and hence this test is very important.
  • Ensure that the content is appropriate for the target users. For e.g. ensure that there is no adult content in a children's book.
  • Now start going through the tables and figures in the book making sure the bottom numbers are corresponding to the format and are in the correct order
  • Next ensure that the fonts used are the same as the specification. If the testing time/resources is limited take random sample pages (try to get one from every chapter if possible)
  • Make sure the figures and tables come in the right page where the matter is supposed to be
  • Make sure that the margins are at right angle to the bottom of the book. This is to see if the entire text section is slanted after the printing
  • Make sure the fonts are consistent (or according to the standard) throughout the book

Data validation

  • First of check and see if the title and outer cover is what it’s supposed to be. This can be accomplished by cross-checking with a given specification. Particular things to test are the author’s name, title name and making sure the graphics are what they are supposed to be.
  • In the back cover check and see if the ISBN number is correct.
  • Ensure that the barcode representation matches. You will need to know how to decode barcodes or use a barcode scanner which outputs the corresponding text string.
  • Test if the price of the book is printed in proper format and as per specification. If the price is printed on both the inner cover of the book and the back of the book, ensure that both prices match.
  • Move to the publisher information section to see if the dates correspond to what we expect. This is a good time to see if the edition of the book is correct. Here is where the copyright legal information is present and it is VERY important to make sure that it is indeed what the legal advisors mentioned in the specification
  • Next is the most time consuming part. Reading the contents and verifying that the subject matter corresponds to the subject and the target audience. This is best done by a person well versed in the subject.
  • Test for spelling and grammatical errors.
  • Check to see if the author is able to articulate the ideas correctly, as per the user specification
  • Test if the book contains a list of references, if the author has referenced some other works in the book.

User Assistance

  • Test if the book has proper resources listed for user assistance. These may include but not be limited to author's contact information, if the reader has any concerns.
  • The book should also have information about the publisher, if the user wants to order more copies of the same book or want to file a complaint.


  • If the book contains a CD-ROM, test to see if the CD-ROM is readable in the computer environment given and that the data inside matches what is expected.
  • If the book has an accompanying web site, ensure that the basic information about the book matches on both the web site and the book.
  • If the book is a sequel/prequel to another book, then its contents should seamlessly correspond with the contents of its corollary.


  • If the book has multiple-language versions, ensure that the printing press has not inserted the same contents of a page in a different language, maybe due to an error in the automated equipment
  • Test for cultural differences in content of the book, so that it might not offend anyone in its target market
  • If an international copy of the book has price written in more than one currency, then ensure that the price matches the specification
  • If the book is published in special languages like Arabic/Persian, then ensure that the text is printed to be read from right to left and the index on the left side is wide-enough to facilitate seamless reading of the book.
  • For a foreign language book check with a person knowledgeable in the script, to see if the characters are rendered correctly

Threat modeling and mitigation

  • Use chemical tests to verify that the paper & ink does not contain toxic elements
  • Check to see if the paper has the right amount of recycled material, either by any chemical tests or by verifying manufacturing records
  • Check to see if the paper will make “paper cuts” to a user who moves his/her fingers along the paper edge. This can be accomplished by using wax and running the edge along it measuring the depth of the mark caused, instead of injuring the tester’s fingers!


  • Take the book and throw it on the floor a few times to see if the bindings start giving up.
  • If that is passed, try flipping through the pages as fast as you can, as many times as you can. This is the physical activity that will be performed most on a book and a good book should stand this test. If the quality of the book is not good, papers will start ripping off or at least get torn at the edges. This can be automated by using a machine similar to a bank teller’s currency counting machine to flip through the pages
  • Test that the pages are not be torn when a certain amount of force is applied. Check this by trying to simulate that force
  • If the book is a paperback (as opposed to a hardcover) try rolling the book and see how much permanent damage is done. It can be quantified using the curvature of the book after folding as a unit for measurement.
  • If the above tests are passed, take the book through some testing of “the elements”. This includes high temperature to see if plastic coatings start melting before the temperature in the specs is reached, humidity to see if the paper quality gets eroded
  • It is worth noticing some special cases at this point. If the book is to be used as a manual in a chemical factory, try to recreate the environment there (e.g.: chemicals in the air) and see if the book withstands it. Another example might be submerging the book in water for some time to see any physical changes, if it is supposed to be used as a map for white water rafters! A good test for computer books is to see how it can withstand coffee spills, since many software engineers prefer to have coffee to keep them going through late night sessions!

Special Cases:

Book to write/draw:


  • If the book has lines, then they are equidistant from each other and all pages have a margin of same size.

Sketch book

  • Ensure that the pages have enough friction and are thick enough that, if the user erases and redraws, the paper is not torn apart.

Appointment book

  • Ensure that the appointment book meets the functional requirements of having necessary tools like a calendar, and writing space for every date and time-slot.
  • Ensure that the dates on calendar are correct

Scrap book

  • Test that the pages are thick enough, so that they don't get damaged, if the user applies glue to paste pictures, etc on the pages.
  • Test that the aesthetic appearance of the book appeals to the targeted user base.

Book to read:


  • First of all confirm that the book can be unzipped using the recommended utility if it comes as a zipped file.
  • Next make sure the book can be opened using the software it is supposed to be used with (e.g.: Microsoft Word, Adobe Acrobat Reader)
  • Now use up the memory in the system and try to open the book at this low memory conditions.
  • Scroll through the pages very fast and note how long it takes before graphic intensive pages become fully rendered.
  • Make sure all the pages are contained in the book i.e. the file is not partial.
  • Repeat these tests for various environments (e.g.: Windows, Unix, Apple etc)
  • Repeat similar tests for Kindle (completely separate set of test cases)

Children's book

  • Test if the content is appropriate and appealing to a typical targeted child. Ensure that the book contains enough pictures and colors, as opposed to the regular black and white pages.
  • Test for performance, as younger children are more likely to chew the book or try tearing away the pages more than adults.

Classified government/defense book

  • Ensure that the cover of the book clearly states "Confidential" or "Classified" or any other appropriate word on cover, to ensure that the user understands its importance.

User's manual

  • Test if the book covers the functional and technical specifications of the product, for which it is going to be used as a manual

Library copy

  • Test if the book has labels, with the call-reference-id, so that it can be easily spotted on the shelves.
  • Test if the book displays the name of the library
  • Test if the book displays information regarding, when it is due back and the pertinent late-fees
  • Hard cover
  • Test that the cover-flap on the hard-cover has the correct information about the book, including but not limited to name, author, price, etc.


  • For non-hardcover books, stack it vertically next to other books in a shelf to see if the shape degrades. This is a very common way for people to store books which makes this test relevant

Music Notes book

  • Test that the pages do not make enough noise while flipping, because the music notes books may be used during live-recording of music

Manuscripts and antique books for museum

  • Test if the pages are chemically treated to prevent any environmental damage to the book

Pocket book

  • Test if the physical dimensions of the book are small enough to fit in a regular pocket.
  • Test if the font-size is at least legible
  • Very large folio-sized books
  • Test if their size is justifiable
  • Test if the size serves its purpose


  • Test if the atlas is bound in such a manner, that if two adjacent pages form a single map, they can be completely opened flat to view complete details of the mapReligious books
  • Test that the religious books are printed and bound in accordance with the beliefs of that religion. For e.g. A book on Hinduism should not be bound in cow's leather

Braille book for blind

  • Test if the contents of the book are easy to find and comprehend by the visually impaired users
  • Test if the size of the fonts, thickness of the pages and other physical dimensions of the book meet the compliance standards for Braille

Catalog book

  • Test if the products and prices match
  • Ensure that the user can find ordering information on the catalog

Cloth book

  • Usually cloth books are written in form of scrolls, with wooden handles on both sides. Test that scrolls are arranged in the right order

Audio book

  • Test that the format of the audio file is in one of the popular standards and can be listened to in one of the popular audio device

If there's more interest on this topic, I'll follow up with a detailed post on testing interviews (how to apply for and how to recruit for awesome QA roles).

Let's break things!

IMDb: onward to 100 M unique visitors through social content, personalization and Asia

I decided to skip watching Wolverine on Friday night and thought about the business of “movies” from my vantage point (as a tech entrepreneur).

I’ve used  the Internet Movie Database (imdb) for more than ten years as my primary source for all things “movies”. Along the way, I’ve also frequented Yahoo movies, Rotten Tomatoes, my college newspaper’s movie reviews and Flixster (on faceboook) for American movies and IndiaFM for Hindi movies. As IMDb continues to grow towards Col Needham’s vision of putting a Play button on every page (the Big Hairy Audacious goal of streaming all movies and TV episodes for free), it has unique opportunities to dramatically increase traffic from 57 M unique visitors per month to 100 M. In this post, I want to share my thoughts on:

  1. The mental model of a movie viewer’s intentions around movies – as a finite state model
  2. How would I approach to grow the user-base by providing an immersive and delightful “movie” experience to users
  3. Social graph, personalization and Asian markets – why they matter!

Where do people go today to find information about movies/celebrities?

While IMDb is one of the most frequented web sites in the world and the most frequented movie destination site in the world, competitors like Yahoo Movies and Flixster (200+ % growth!) have been growing at a faster rate. Here’s a comparative snapshot of US traffic from Compete.(Alexa, Quantcast, Comscore show similar relative results)


Mental model of a movie viewer’s intentions and actions:

Based on my personal experiences and conversations with friends, people visit an entertainment/movie website to read movie reviews, find out more information about a movie/celebrity, find showtimes for movies, watch trailers, buy a movie ticket or learn about upcoming/new movies. Being a computer scientist, I couldn’t help but create a finite state model.


The ideal web experience would enable the user to transition from one state to another in an intuitive and delightful way, without leaving the website. To further understand the four main scenarios, let’s take an example of a fictitious persona : Shelly Jones, who has recently graduated from college and loves movies.

  • Discover: Shelly may discover a new movie or a new celebrity through multiple sources (online and offline) – by talking with friends, watching a trailer in a movie theater, seeing a poster, watching an ad on TV, reading an article in a blog/newspaper, reading the list of upcoming movies on a website, etc.
  • Learn: After discovering about a movie or a celebrity or a TV show, Shelly wants to learn about the movie by reading synopsis, reading a review, learning more about the cast, watching trailers online, listening to/reading interviews of the cast, etc. such that it helps her to decide whether or not she should watch that movie.
  • Watch: If Shelly decides to watch the movie, she may also want to find out the
    showtimes near her location and potentially buy her tickets online. If
    feasible, she may want to watch the movie online right away on her
    computer or rent it (on Netflix or Blocksuster)
  • Discuss: After watching a movie, Shelly may want to write a review, rate the movie and discuss more about the movie with others who may have watched it. She may want to learn more about the cinematographer for the movie and watch other movies by the cast members.

While most movie sites address some or all of the above listed scenarios, the successful ones focus heavily on making the transitions (represented by numbered arrows in the diagram) obvious, intuitive and delightful, thereby reducing any friction for the user to immerse herself in the world of movies. Let’s explore the transitions:

  1. Discover > Learn: Once Shelly discovers a movie, TV show or a celebrity, she should be able to easily search for more information with at most two clicks. The search process should be enjoyable and exploratory, the content should be available and trustworthy and Shelly should be able to immerse herself into a discover > learn > discover cycle.
  2. Learn > Watch: One of Shelly’s primary objective is to decide whether she should watch that movie or not. Shelly should get objective and expert guidance/reviews about a movie/celebrity to help her decide. Shelly should be able to tag or store the information and should be able to access it later.
  3. Watch > Learn: After watching a movie, Shelly would want to learn more about the movie / celebrity.Once Shelly indicates that she has watched a movie, she could be easily shown which of her other friends have seen it, what information have they accessed, how can she use her experience of watching that movie to make decisions about discovering and watching new movies.
  4. Watch > Discuss: Invariably, whenever I go to a movie with friends, the first topic of conversation after the movie is – “Did you like it?”. Shelly is likely to rate a movie, write a review or discuss the movie/celebrity in the message boards. The key would be to highlight key topics of discussion around a movie to Shelly, allow her to share/discuss/compare her review with her friends.
  5. Discuss > Learn: Through conversations, Shelly is likely to discover information about linked movies/celebrities and learn more about it.

IMDb addresses all of the above scenarios, but it has an opportunity to make these experiences more engaging, social and personalized. Based on the affordance of IMDB’s website, today it appears to focus the most on “Learn” and “Discuss” with a long term vision to emphasize on “Watch”. Although, all three experiences are superb by themselves, Shelly’s experience can be enhanced significantly by reducing friction from the transitions.

How would I grow traffic at IMDb?

IMDb has been growing at an enviable rate for its size (almost 40-50% year over year) ans as it continues to grow towards its big goal of streaming all movies and TV episodes, I see strategic opportunities to significantly grow traffic/ad revenues while satisfying the users and advertising customers.

IMDb is an established brand with a large number of loyal users. Taking an overly simplified approach, there are two ways to grow:

  1. Provide “more” services and thereby monetize the existing demographics more
  2. Attract new demographics by offering newer services


1.1 Existing Demographics:

As shown in the tree diagram above, it would be both strategic and tactical to address the users with existing demographics, which can be further segmented as

  • 2.1 Existing Users, who already frequent the website
  • 2.2 New Users who may have similar demographics as the existing users, but do not frequent IMDb

2.1 Existing Users:

  • 3.1 Increase average time per visit:

While integrating videos is likely to increase average time spent on the site, IMDb has several assets (message boards, ratings, reviews, recommendations) which can be leveraged to have more users spend more time on the site. While most discussion on IMDb is facilitated by message boards and comments, there are tactical opportunities to make these conversations more discoverable, frequent and delightful by augmenting them with a social graph and an architecture of reputation for all user generated content. When users have an incentive to discuss movies and celebrities with their friends, they’re likely to spend more time on the site in communicating and reciprocating on the web site, in addition to the time they would spend for their personal movie-exploration.

  • 3.2 Increase frequency of visits:

Existing users typically visit a specific intent to discover/consume some information (and a relatively small percentage of users visit to just explore.) Users in the former category can be prompted to visit the site, when new content which maps directly to their preference is added to the site or when their friends have contributed some content that they may care about. Personalized and social notifications (recommendations, reviews and messages) along with an architecture of reputation could incentivise users to visit the site more frequently.

2.2 New Users:

  • 3.3 Reduce friction to discover the site

In the past three years, a certain class of web-sites and applications have become the sole way to discover, consume and access information on the web. The key in attracting new users is to reduce any friction from their current behavior/method of consuming information.While some optimization can be done through Search Engine Optimization (especially for TV-related information), the highest order bit is to make the information on the site discoverable and available through user’s existing work flow. Detailed integration with facebook and MySpace, further integration with twitter (beyond requesting information about movies) and integration with avenues where people watch movies/TV(for instance: Netflix queues, Blockbuster rental history, Tivo and fandango). This would ensure that new users will discover and use IMDb, without significantly changing their behavior.

  • 3.4 Reduce friction to use/access the site any where and on any device

Requiring a user to be using a computer (desktop/laptop) to be able to fully enjoy the movie-exploration experience may be restrictive. A new set of users could start using IMDb more frequently, if it is pervasive and accessible anywhere from any device. Investments in mobile-optimized website (for instance, trailers which work on iphone), applications on iphone/android/Windows Mobile could reduce friction to access the information. Integration with Netflix-enabled devices (for instance: XBox 360) could be another channel for a pervasive experience.

1.2 New Demographics:

Besides leveraging the users of existing demographics, it would be strategic to invest in newer markets to ensure long-term growth.

  • 2.3 New markets (both geographically and genre-wise)
  • 2.4 New content aimed at both casual users and professional users

2.3 New Markets:

  • 3.5 New geography

IMDb is the number one movie site, but caters primarily to Hollywood movies or acclaimed International movies. India boasts the world’s biggest movie industry (at least in terms of number of movies released per year) and several Indian celebrities are popular across the world (even among non-Indians). While IMDb contains title pages for several Indian movies and celebrities and syndicates news from, the experience for browsing an Indian movie is not as immersive or delightful as that of browsing a Hollywood movie. Through a combination of key partnerships with Indian content providers, SEO and seeding IMDb resumes of Indian celebrities, IMDb has an untapped opportunity to become the number one movie site for Indian movies. The same applies to movies made in rest of Asia (China/HongKong/Taiwan in particular, where the movie is going through a massive growth in the relatively liberal years currently.)

  • 3.6 New genre

While IMDb has successfully rolled out a dedicated section for all things – TV, there are opportunities to increase scope and grow by offering information about video games, user generated videos on YouTube, anime and music.

2.4 New Content:

  • 3.7 Content for casual users

While pro users contribute most of the content, casual users (who mainly seek information about a movie, TV show, trivia, celebrities, gossip, etc.) typically tend to account for the majority of visitors (I haven’t confirmed this for IMDb). Casual users could be interested in interactive games, trivia and polls, which is currently hidden in message boards. There is an opportunity to create a new set of content (games and applications) catered towards the casual visitors. The ideal avenue would be to empower the pro-users to create such content/games, by giving them tools to create, promote and distribute such content. Today, the pro-users already do that through message boards.

  • 3.8 Content for pro-users

Most content on IMDb is user-generated and editor-approved. Pro-users (very frequent visitors of the site, who proactively participate in discussions, write reviews, add biographies, etc.) are responsible for most of the user-generated content on IMDb. There is an opportunity to leverage this loyal base to create more and different content for the casual users to consume, by empowering the pro users with tools to create new, engaging and interesting content. Today, most of the innovation is restricted to message boards on IMDb, which restrains the creativity of the creator and discoverability of the content by a casual user. By providing tools to create games, different visualizations of content (photo albums/collages), movie guides catered towards a diverse set of users (guides for Christians, guides for immigrants, etc.), user-submitted photos, videos and fan sites – a completely new set of content can be created and offered to the users. This approach can be further optimized by adding the notion of “reputation” to incentivize pro-users to contribute more and high-quality content. Today, a user’s profile page on IMDb doesn’t distinguish a registered user who may have written 100 movie reviews from a user who may have not written any.

Decision Matrix – Which ideas to implement?

As a business owner, I would optimize the implementation on the basis of users’ satisfaction and advertisers’ satisfaction. The top left quadrant in the 2X2 matrix below indicates the changes that would add positive value to both users and advertisers while the top right quadrant indicates ideas that are positive for the users but may not be the most optimal for advertisers. Based on the opportunities described above (3.1 -3.8), I see three main themes, which would add positive value to both users and advertisers: social content, personalization, Asia.



  1. Social Content

    1. Social Graph: I’m not proposing a “me too” approach of copying “Flixster”. Instead, I’m proposing to highlight and leverage the already existing social activitiy happening through message boards on IMDb.
      1. Instead of relatively static message boards, IMDb has an opportunity to increase the reach of discussions to a larger number of users by integrating with other portable social graphs (Facebook Connect, My Space Open API, etc). It would be strategic to extend the notion of “Friends” on message boards to IMDb’s own social graph in the long run, thereby enabling scenarios for easily sharing friends’ reviews, ratings, notifications, etc. (today limited sharing is available through exchange of static links, which could be a barrier for some users)
      2. The conversation can be further extended to a conversation between celebrities and users, as opposed to amongst just users. Ashton Kutcher has proven the value of a constructive dialog between a celebrity and fans. IMDb has an opportunity to be a social interface between celebrities and fans.
      3. IMDb has an opportunity to incentivize users who add significant value to other users by implementing a notion of “reputation” based on a user’s activity
    2. More user generated Content:
      1. IMDb has an opportunity to leverage the loyal user base to create/source innovative content around movies, TV and celebrities. Empowering users to create their “fan sites” or blogs around a particular celebrity or a movie could lead to more traffic and a stronger sense of community.
      2. Enable user to comment, rate and review individual content pieces as well. For instance enable users to comment on a photo, instead of generic comments on the title page.
    3. Platform Approach:
      1. A majority of innovation could be crowdsourced to the long tail of movie fans and devlepers who would find creative ways to build new experiences around IMDb’s content. Today most content is not easily accessible (developers need to download the text files locally and pay licensing fees.) IMDb has an opportunity to selectively expose some content through a hosted API, and let movie fans innovate. There’s a minor risk of sabotaging existing licensing fees (earned from the content), but it could prove to be a strategic long term move
  2. Personalization:
    1. Instead of restricting personalization to visual appearance, IMDb has an opportunity to leverage user’s browsing habits, ratings, reviews, friends, message board participation and My Movies lists to create a very personalized movie-experience. Personalized notifications through RSS feeds about new content that may interest the user could incentivize the user to visit the site more frequently.
    2. Partnerships with Netflix, Tivo and other similar services could help automate personalization.
    3. Enable the user to fine-tune the personalized recommendations by allowing them to vote on the recommendations.
  3. Asia Readiness
    1. Focus on India: Position IMDb as the most comprehensive destination for information on Indian movies and celebrities (it is NOT today), by seeding information about Indian movies and celebrities.
    2. Raise awareness amongst the Indian audiences through SEM and partnerships with leading content providers in India.
    3. Repeat in China.

IMDB has been successful as the most comprehensive source for movie, TV and celebrity content. Fostering easily accessible and shareable conversations around this content could accelerate and nurture rapid growth.

There’s no business like show business!


Amazon Kindle College: rent textbooks – my elevator pitch to Jeff Bezos

One of the most annoying anomalies during my college education was the incredibly high and progressively increasing price of textbooks (often around $100 for most Computer Science and Economics classes). More than the price itself, the most irritating factor was the artificially-induced price-inelastic demand for text-books (no matter how high the
prices were, almost all students including myself were forced to purchase the textbooks). These high prices can be attributed to the existing business model and delivery mechanism of text-books which is dominated by a small set of publishers and distributors. While renting physical printed books does address some of the issues (I admire Chegg : Text book rentals), it doesn’t solve the problems of high cost, out-of-date content and logistical inefficiences.


From my vantage point, Amazon Kindle’s adoption, usability and content-distribution mechanism makes it far more superior both functionally and experientially for the college student. I believe firmly that Amazon Kindle for college students has an opportunity to make a substantial dent in the $4.9 billion text books market by shifting the focus from “used books” to affordable and up-to-date new digital book rentals. If I ever bump into Jeff Bezos in the next few months (you never know – Seattle’s a small city), here’s my elevator pitch:


  • Amazon Kindle College could offer college students with rental textbooks at dramatically lower costs than used text books, thus substantially reducing the student expenses on textbooks by more than half. College students who are often cash-strapped are frustrated by the ever-increasing costs of text-books and the hassles associated with re-selling them back.
  • Kindle could enable to publishers to gain substantially larger aggregate margins for a textbook unlike contemporary model where the publishers get margins/profit-share only from the first sale of a new book, even though an average textbook may be sold multiple times over in the used books market.
  • Kindle could leverage technology and a significantly better customer-experience to eliminate the inefficiencies associated with the used books market, which is estimated at a $1.9 billion per year.


Textbooks are expensive and prices keep increasing:

Textbooks account for a major expense for college education, ranging up to $1000 per year per student. Since the supply chain (publishers and distributors) is an oligopoly (just five major publishers – Thomson, McGraw-hill, Wiley, Houghton-Mifflin and Peason, and just four major distributors/wholesellers – Follet, Barnes and Noble, Nebraska and College Bookstores of America), price of new textbooks is not controlled by market forces. The demand for textbooks is almost inelastic (see the graph above). Nonetheless, students are forced to purchase textbooks.

The market for used books is inefficient and wasteful

Students waste significant mindshare and resources at the beginning of every semeseter/quarter to sell their textbooks back (either to other students or to a bookstore), and are often forced to do so due to financial constraints. The “used books” market is typically driven by bookstores (online and offline) and is often full of inefficiencies related to pricing and logistics. The “used books” market is optimized for the benefit of the bookstores/distributors while publishers as well as authors are kept out of the loop.

Renting physical printed textbooks has shown positive results, but is neither scalable nor optimal.

Owing to large initial investment requirements and conflicting interests with existing profit streams of bookstores, bookstores/universities often don’t rollout rental programs. In the last few years, there have been some progressive startups like Chegg, who rent textbooks for typically 1/3rd price of the cost of a new book. But these models work only for beginner classes (with a large and predictable enrollment of students). These models also keep the publishers out of the loop of sharing margins of book-rentals, thus giving them incentives to publish new books more frequently.


Amazon Kindle for College could rent textbooks to students for a semester/quarter at a substantially lower price than that of used books. Incentivize publishers and authors by sharing revenues from each book rental. Shift the cost savings from “not having to move around used books” and “not paying hefty margins on used books at bookstores” to the customers, while providing them with an incredibly superior “Textbook” experience.


Overall the market textbooks in the US is estimated at $4.9 billion per year. Used books account for about $1.9 billion per year market annually. There are bout 17.5 million college students in the US (addressable market size.)

Benefits of innovation:

  • Students get a superior and up-to-date textbook experience for a significantly lower price
  • Publishers get a share from every book consumed by a student, thereby increasing their aggregate profit.
  • Authors get a royalty for every book consumed by a student and can publish updated content seamlessly at a marginal cost.

Distribution and go to market:

  • Offer students with an offer they cannot refuse. Price the books such that, the cost of a Kindle College and first semester rentals would be less than what students would pay otherwise for textbooks for one semester.
  • Add value to the Kindle College experience by opening the Kindle platform to developers to develop value-added applications

Napkin Numbers: These are some numbers that I came up with based on my college experience and some napkin math. These numbers, may not be exactly accurate, but should be in the ballpark. I had purchased a Computer Science textbook “Introduction to Algorithms” by Cormen for about $100 and am sure that it is still being circulated in the used-books market, benefitting the bookstores.

After being forced into buying a new book for $100, I could have sold it for $40 to the local bookstore, who would resell it for $70 and the cycle could continue, until the publisher publishes a newer version of that book. Although the book would be experienced by multiple readers, the publishers and authors wouldn’t get any royalties/margins from the subsequent purchases.



Rentals on Kindle may eat into profits of existing “used books” business of Amazon in the short term, but it is THE right thing to do for customers in the long run.

Most numbers about the market and supporting data are quoted from a comprehensive study done by Dr. James Koch (smart man) which can be found here.

On a personal note, this has truly been one of the most irritating things about my college experience and I had attempted the following means at various times to deal with it:

  • Bought the same textbooks at significantly cheaper prices from India (typically for the spring semester, whenever I went to India for Christmas break.)
  • Attempted to rent my textbooks to juniors.
  • Sold my used books on Amazon zShops.
  • Did not buy a text book for a class (as an experiment) and managed to get a decent grade (had almost bought the textbook on the eve of a midterm exam)
  • Mostly ended up buying used books on or (they offered the cheapest used books)

Let’s fix this!

Berkshire Hathaway letter – what I enjoyed

Warren Buffett does it again. Here are some verbatim statements, that I enjoyed in his latest letter to the investors of Berkshire Hathaway.

By year end, investors of all stripes were bloodied and confused, much as if they were small birds that had strayed into a badminton game.

In God we trust; all others pay cash.

In poker terms, the Treasury
and the Fed have gone “all in.”

In good years and bad, Charlie and I simply focus on four goals:
(1) maintaining Berkshire’s Gibraltar-like financial position, which features huge amounts of excess liquidity, near-term obligations that are modest, and dozens of sources ofearnings and cash;
(2) widening the “moats” around our operating businesses that give them durable competitive advantages;
(3) acquiring and developing new and varied streams of earnings;
(4) expanding and nurturing the cadre of outstanding operating managers who, over the years, have delivered Berkshire exceptional results.

During 2008 I did some dumb things in investments. I made at least one major mistake of commission and several lesser ones that also hurt.

Furthermore, I made some errors of omission, sucking my thumb when new facts came in that should have caused me to re-examine my thinking and promptly take action.

As we view GEICO’s current opportunities, Tony and I feel like two hungry mosquitoes in a nudist camp. Juicy targets are everywhere. First, and most important, our new business in auto insurance is now exploding. Americans are focused on saving money as never before, and they are flocking to GEICO.

A promise is no better than the person or institution making it. That’s where General Re excels: It is the only reinsurer that is backed by an AAA corporation. Ben Franklin once said, “It’s difficult for an empty sack to stand upright

Ajit came to Berkshire in 1986. Very quickly, I realized that we had acquired an extraordinary talent. So I did the logical thing: I wrote his parents in New Delhi and asked if they had another one like him at home. Of course, I knew the answer before writing. There isn’t anyone like Ajit.

I described it as involving “borrowers who shouldn’t have borrowed being financed bylenders who shouldn’t have lent.”

capitalism at its best..

on poverty elimination: scope and my approach

Poverty-elimination is one of my life-goals. Recently, I took some time to scope the problem of poverty-elimination and attempted to come up with an initial approach. I approached it in the same way, I would approach any new product launch – by segmenting the market and understanding each segment's role. Here's a snapshot of my early analysis. I would love to hear your thoughts:

Question: What can I do to eliminate global poverty in the next 35 years?

Answer: Assuming that microfinance works, most obvious thing to start with would be to make microfinance more accessible in a measurable and leverage-able way.

More on the answer..(I don't know answers to the questions in red, but the search is on..I'll post the answers as I find them..)

Segmentation: In an effort to understand the market needs, following segmentation of the addressable market seem to make sense to me, but there are some fundamental questions (listed in red.)

1. Poor versus Ultra poor: At a very high level, needy poor can be divided into two categories:

1. People living just above the poverty line (say $2 per day) and

2. People living below poverty line (ultra poor)

Out of the two billion+ poor in the world, how many fall in #1 vs #2?

I’m assuming that microfinance typically addresses the needs of the economically active poor (i.e. those who live above the poverty line). With sophistication of the industry, more commercial institutions (banks, commercial lending groups, etc.) are likely to enter the market, but would serve the needs of the poor who are living above the poverty line. NGOs are most likely to serve the ultra poor in these circumstances. Would the NGOs be able to meet the demands of the ultra poor in terms of scale and type of services? Also, when resources-laden commercial banks enter the market how would the existing MFIs differentiate themselves/maintain a competitive advantage?

2. Willingness versus readiness versus access to microfinance:

Another way to segment the poor could be:

  • ·         Those who are not willing to borrow, due to social beliefs, fear of debt, etc.
  • ·         Those who are not ready to borrow, due to lack of markets, political/geographic instability, etc.
  • ·         Those who don’t have access to microfinance, due to lack of access to MFIs or lack of enough MFIs in the market.

    Is this segmentation apt and  how many of the poor fall in “not willing” vs “not ready” vs “don’t have access” category?

Innovation: I believe that three types of entities have a bigger opportunity as well as higher probabilities to innovate and add value in measurable and leverage-able ways:


Type of Innovation

Potential value-adds



Service Innovation

a.       MFIs could design and implement newer delivery mechanisms of microcredit to make it more accessible to a larger set of poor and decrease the number of  (not willing + not ready + don’t have access).

b.      MFIs could design and implement newer ways to offer micro-credit at a cheaper rate than today, by introducing technical/process innovation (for instance: automatic credit approval and lending, auto-payment options, etc. to reduce manual inefficiencies from the microfinance value chain)

c.       MFIs could design  and implement new services that span beyond just microcredit  (for eg: money management, health insurance, credit rating, etc.)



Social Innovation

d.      Test and deliver high cost/riskier programs to address the needs of the ultra-poor.


Social Enterprises

Product Innovation

e.      Design and distribute products that directly impact the bottom line of occupation/work that the poor are typically involved in (for instance, Grameen Phone initiatives in Africa, Cheaper irrigation products by Paul Polak, etc.)

This is a first step in doing something measurable and leverage-able towards poverty-elimination.

Let's eliminate poverty!

Extra: I had asked similar questions to Paul Polak – "Can we eliminate global poverty in the next 35 years?" His answer was Yes (with conviction). I liked how he explained pragmatically : Considering that half of the world's poor rely on agriculture for their livelihood, desiging innovative and affordable products to make effective agriculture accessible to the poor farmers – can solve half of the problem. Similarly product innovation can help with the other remaining half as well…

on Competition: knowing the other players

"As an entrepreneur, if you think that you don't have any competition, then it means one of two things:

  1. What you're working on is not worth working upon or
  2. You don't know how to use Google"

Guy Kawasaki had once told this to Peter Panas and me during a global student business plan competition years ago, after we had just pitched the business plan for Securamed at the Stanford global e-challenge in Singapore. Competitiveness is the essence of any business and you need to know the competition to be competitive.

It must be obvious that you need to know who your present competitors are, but it is even more important to know the following three facets about your competitors:

  1. How will your competition react to your strategies?
  2. What is your strongest point and your competitors' weakest point?
  3. Who is likely to be your competitor in the short term vs the long term?

In this this post, I'll share my thoughts on #1 – proactively predicting the competitor's moves/reactions:
Just knowing the names and basic offerings of your competitors is never enough. Basic information about the obvious competitors is typically very easy to find. If it is a large company it is pretty straight forward to find the basics from the company's website. If the competitor is a public company, then it is even better. For large companies, which are private, you may find the basic information from the Dunn And Bradstreet reports. For large companies, which are publicly listed, most information is available from the EDGAR reports or the annual reports. Hoover's database (easily accessible from any decent library) is also very comprehensive. It starts getting trickier as the size of a competitor starts decreasing. Not only can a smaller company harder to spot, but their strategies/reactions are harder to predict. Most effective small companies are fairly agile and can adapt easily to changing market conditions.

Despite the size of the company, there are two proven ways to learn the most about a competitor.
One, be a customer of the competitor. Eat the competitor's dogfood. Empathize with the experience offered by your competitor to their customers. Use the competitor's product. Talk to their customer support personnel. Post a question on their help forum (if one exists.) I'm not suggesting espionage. There are legal ways of purchasing and using a competitor's products. Don't hide your identity. It is OK to say that you work for your company and disclose who you are. The competitors assume that you know everything about their product, anyway.The goal of this exercise is not to find the good things about a competitor and copy them. The key is in understanding the guiding principles of the competitor's execution strategy. Is the competitor focusing a lot on customer service? Is the competitor's product sold on the merits of its user interface? How has the competitor's product matured over time? Has the backend being made more robust or has the competitor focused more on improving the user interface over time? What are their customers/users liking/disliking about their product (can be found easily through newsgroups/blog posts, etc.)? Buy the best possible version of the competitor's product and renew your service subscriptions. Keep up with the competition as a pro-customer and get a pulse of their business.

Second and most important aspect is to really really know the people involved in the competition. Is the CEO of the competition a strategiest or a people manager? Is he/she a technologist or a marketer or a deal maker? What does the DNA of the competition's leadership look like? Is it dominated by technocrats or sales people? Where have most people worked before (this gives a pretty clear picture of their work-style or their approach?) Who is likely to succeed the current leadership? The goal of this exercise is to really understand how your competition is likely to react, when you execute your strategies. For instance, if you introduce a new product, is the competitor likely to introduce a more technically robust product, or is it likely to strike a deal with another competitor to combat you or is it likely to reduce the price of their offerings. These things can only be found out by investing heavily in understanding the minds of the key leaders at your competitor. All decisions and strategies are implemented by people and not organizations/companies. It is important to be able to understand their value system, their guiding principles and their likely reactions. If you know the likely reaction from a competitor, you can always be prepared for it or even use it to deceive the competition. Data about people, their past experiences and their workstyles can be easily found on LinkedIn/facebook. Be aware that the competitor would also  be trying to sum you up, so control how you project yourself through LinkedIn/facebook/blog posts/replies to newsgroups, etc.

If you play any competitive sport, these things may seem like "common sense", because it is common sense. Remember to use it. Tennis and chess provide a very straightforward way to simulate one:one competitive environment and practice your strategies. My most favorite board game – RISK, is the best way to simulate competition amongst multiple players and try out strategies.

A lot have been written on competition by both academicians and practitioners. Works of Michael Porter and Sun Tzu (The Art of War) remain the most admired. Read these works, if you haven't.


Why I blog?

A friend recently asked me – why I blog? Simple answer - I get to connect with several like-minded people by sharing my thoughts. I've met several folks from around the world who are now my very good friends. Here's one of the recent mails I received from a reader. Messages like this makes it worthwhile to blog and share ideas.


Hello Kintan,

I am currently pursuing Masters in Information Management from University of Washington, Seattle. I have been a follower of your blog for a very long time and I really appreciate your taking time out to write about Program Management.

Your blog was one of the reason, I applied for the PM position at Microsoft and I actually crossed the initial screening. I will be on Microsoft Campus very soon for the on site interviewing.

I need your suggestion on designing in particularly "software". I understood the concept of designing the "physical objects" like a kitchen, or a building. But while thinking of designing a software, I wanted to know whether the interviewer means the UML design (like the way it is given in OO Design Patterns by Gang of 4) of the software or something else. And what modifications will you suggest in your "design template" specifically for Software.

Looking forward to your reply.




back to business after a “think month”

Every year, I set aside some time to "think". Typically, it has been during the month of December, but this year it ended up being the month of October. While throughout the year, mind is always on a relentless pursuit of thought, followed by action, I make conscious efforts to prioritize thoughts, share thinking points with friends/mentors and select three concrete areas to work on.

I've found this process very helpful and in fact addictive over the last six years. Several people I know and respect have been doing it for years. For me, the "think" day/week/month entails three things : 

1. Emptying your mind and organizing all thoughts in the mind: Open up a journal (I prefer paper journal, but OneNote works as well) and write down all thoughts that have been lingering around in the mind – things you've wanted to learn, things that have been bothering you, things that you've wanted to get done, people you've wanted to meet, etc. After listing down all thoughts – sort, bucketize, prioritize these thoughts/themes.

2. Deliberately add new thinking points to the mind: Since our mind is always receiving information/thinking points from a gamut of sources, "think time" is a good opportunity to put a filter on the information sources and seek out to learn more about a few specific topics. Seek out the experts on these topics, read books, read blogs and take notes.

3. Executing: The entire exercise is not worth it, if it is not acted upon. Prioritize three or four themes/areas to focus on, create milestones and action items, between now and the next "think time."

None of these points are new and most of us use this (or an even more complex/scientific process) for work-related projects. I've gotten myself to do several insightfully satisfying things and thereby increase the portfolio of my passions through these "think times". Starting companies, painting, blogging, getting passionately involved in poverty-elimination/Unitus, travelling to specific places, some very close friends – have been results of my past "think times." Pixtoria is one of the results of my most recent "think time".

A wide variety of themes dominated this year's "think month":

  • relationships: family, friends, personal vs professional circle of influence
  • opportunities/industries: unified communications, social networks, mobile, microfinance, healthcare

  • poverty elimination vs environment protection, micro econ vs macro econ
  • living in Seattle vs San Francisco
  • portfolio of passions: painting vs photography, car racing, travelling

I typically don't suspend my regular activities/work during this "think time" (in fact, I try out things that I've been wanting to do for a long time, but haven't gotten a chance to do so – for instance, sky diving, reading books that were on hold for a long time). I don't blog during this time.

The duration of "think time" does not have to be a month, but at least a day or a weekend. It is hard to go in detailed depths, if you don't invest sufficient time. I'll share some of my thinking points during subsequent posts.

let's think!

Let ideas PoP (Panel on Poverty 2008)

Global poverty is a big and a hairy problem. To eliminate global poverty is an even bigger, harier and more audacious goal. Every year, few friends of mine and I organize the annual Panel on Poverty in Redmond, where we invite the leaders from global organizations working towards poverty-elimination. Typically “who’s who?” of poverty-elimination/microfinance in particular gather on the same platform to openly discuss key trends, solutions and challenges of poverty-elimination.

If you missed this year’s POP, here are the video recordings:

I believe firmly that global poverty can be solved in our lifetime. Generations to come will learn about poverty only through”poverty museums.”

Let’s eliminate global poverty.


Technorati tags: Panel on Poverty , Poverty