A short piece in Business Week provides some clues:

Borders’s demise, though, has as much to do with real estate as any metaphysical market shift. During the superstore boom of the 1990s, Barnes & Noble paid close attention to where it put its outlets, which were usually in prime locations. Many of the profitable Borders stores were also centrally located, but numerous industry observers characterized the company as grasping for growth. It had a policy of picking “B locations,” says Fox, and trying to turn these sites into “A economics.” Leases on its stores were also “unproductively long,” adds CEO Edwards. As the company’s fortunes turned, it was difficult for Borders to buy its way out of leases that still had seven and eight years remaining on them.

Analysts predict that Barnes & Noble will have to shrink the number and size of its stores, and it hasn’t tried to gobble up many of the vacated Borders locations—70 percent of which, Barnes & Noble says, were within five miles of one of its outlets. (Barnes & Noble did purchase the remainder of Borders’s Web business.) But so far Barnes & Noble is holding on to its stores, focusing on e-books and filling its outlets with high-profit-margin nonbook items, such as educational toys and games.

The one thing Borders did have going for it was its huge selection, yet even that wasn’t worth as much as the company thought. An average Borders superstore stocked around 140,000 titles at immense cost, but if a customer craves selection, no store can compete with the long tail of the Internet. Maybe more crucially for Borders, the assortment of titles that provided the key to its identity didn’t give it a competitive edge over Barnes & Noble. Mark Evans, a director of merchandising strategy and analytics at Borders until 2009, says that [Borders] surveyed customers to understand why Barnes & Noble, with its slimmer selection, continued to clobber them in terms of year-over-year growth, average sales per store, and even the number of books sold at each location. “Customers didn’t notice our larger assortment of books,” Evans laments. “They didn’t care.”

It’s true — I hadn’t noticed the larger selection of books at Borders.

Also, do you really expect a company to survive whose claim to fame is their “secret sauce”?

The Borders story began in Ann Arbor, where Louis and Tom Borders opened their first store in 1971. Students at the University of Michigan, the brothers developed a then-revolutionary system to track sales and inventory; for years Borders executives called it the company’s “secret sauce.” Their “Book Inventory System” could oversee the flow of a huge number of titles broken into thousands of different subject categories across multiple stores. By evaluating sales data, the system could understand local tastes and predict demand in specific communities. Initially, the brothers hoped to sell the program to independent stores across the country, but bookshop owners proved resistant, asserting that they—and not some punch-card computer—intimately understood their clientele. Instead, Borders opened additional stores, first in suburban Detroit, Atlanta, and Indianapolis, ultimately forcing out many of those reluctant independents. By the 1990s it had stores all over the country, and together with Barnes & Noble controlled 40 percent of the bookselling market.

Perhaps a better answer comes from Mark Evans, who provides his answer of why Borders failed and Barnes & Noble is thriving in this Quora post:

1. Failure to adequately address the internet sales channel and the subsequent ebook market. Specifically, the decision to outsource Borders.com toAmazon.com. To be fair, Borders.com was costing the company millions of dollars in losses each year ($20m I think when they decided to outsource) and one could argue that the outsourcing solution was a case of letting the most efficient etailing organization (Amazon.com) handle the job and turn a big negative into a profitable business. In the short-term, this saved a lot of money. In the long run, the internet is too important to outsource in this manner and Borders’ branding, multi-channel strategy, and customer base suffered. They also dropped the ball on ebooks, but by the time this became an issue they were just trying to figure out how to keep the whole house from burning down around them, so I find it more understandable.
   
2. Poor real estate strategy – Borders leased space that was too large, the storefronts did not compare well to B&N, and they were complacent in picking and relocating existing stores to the best locations. Some of this is subjective as I don’t have great data to back this up – just my own educated assessment based on observation.
   
3. Over-investment in music – while this was a big plus for Borders in the early to mid 90’s, it was a disaster in the long run. This is why the stores were too big once the music business cratered – stores were sized and modeled to provide a large music CD business which largely disappeared. In addition, infrastructure was sized to support this business, including a dedicated warehouse distribution facility. This last part has been addressed over time, but soaked up money, time, and energy. Music was also part of what made Borders a destination for many customers, so when music sales tanked, other product categories’ sales suffered as well.
   
4. Over-reliance on assortment size to compete as opposed to efficient operations – Borders was renowned for its wide and quality assortment of titles. The very large assortment size was an advantage early on before Amazon. However, by its very nature the internet was better at quickly and efficiently connecting customers with obscure titles and bringing the “long tail’ to market. Thus, competing on assortment size was especially vulnerable to internet retailing and Borders suffered disproportionately as the “long tail” customers abandoned them.
   
5. Failure to build efficient systems and processes – While Borders’ legendary “expert system” was considered cutting edge and an advantage early on, the company failed to successfully build upon this foundation and create new, better assortment, replenishment, and supply chain systems and processes to keep pace with the changing state of technology and efficient retail operations. B&N invested considerable time/energy/money through the 90’s in systems and processes. To provide one example, a lower ranked title that sells out in a B&N will be replenished from a central warehouse within 2-3 days. The same process could take up to 16 weeks for Borders. Borders sought to upgrade systems with two large efforts in the 00’s: first one was a home grown effort called Common Systems. Second was a “buy and integrate” project to implement Retek and E3. Both failed spectacularly. The Retek effort dramatically hurt the Walden chain, the only business unit that was managed by the system. With both of these efforts, large sums of money and, perhaps more importantly, human resources and time were squandered.
   
6. Branding failure – In addition to the Borders.com problem, Borders never reached the mindshare that Barnes & Noble did for a variety of reasons. Also, Barnes & Noble secured the exclusive U.S. Starbucks partnership, a major branding and traffic-driving win for them.

Don’t get me wrong. I loved Borders. I still go to bookstores. It’s just that I will return home and then purchase the books I saw on Amazon.