How Barnes & Noble made a comeback, with CEO James Daunt

James Daunt left finance in 1990 to open the first Daunt Books, immediately facing the Gulf War and a recession that shuttered neighboring businesses and left him unable to cover rent. Despite being financially underwater, sales grew every single month from opening, moving from roughly 10 percent below breakeven to flat to profitable over three to four years. That experience shaped his conviction that a well-run physical bookstore can survive almost any external pressure.

He later took over Waterstones, a large British chain in serious financial difficulty in the late 2000s and early 2010s, after being rejected by every private equity firm he approached. A Russian entrepreneur who loved books backed the deal after roughly 35 to 40 minutes of pitching, with the purchase price described as approximately half what a Premier League club had just paid for a single player. The Waterstones recovery became the template he would apply to Barnes and Noble.

Barnes and Noble was in a comparable position before its turnaround, with declining sales and stores Daunt described as really awful and getting worse on every visit. He chose to pursue keeping the roughly 300 existing stores intact rather than letting the chain disappear and opening fewer new locations himself, because he believed the physical bookstore offered a discovery experience Amazon could not replicate. That bet on scale rather than retreat defined the entire subsequent strategy.

The core of the turnaround at both chains is replacing centralized data-driven merchandising with local store judgment. Daunt argues that Barnes and Noble's predecessors used data to allocate retail space and got it systematically wrong, giving too much room to board books and too little to young readers. He calls data-driven merchandising decisions a death knell for bookselling and instead empowers individual stores to make their own assortment and presentation choices, with data reintroduced only after a culture of local judgment is firmly established. The sequencing matters to him: autonomy first, analytics second.

Barnes and Noble operates with zero marketing dollars and no marketing department, a deliberate choice Daunt maintains even though he acknowledges a thin central team amplifying organic content would likely improve profitability. He resists it specifically to protect the decentralized culture, arguing that BookTok and similar phenomena work because of their authenticity and that management involvement would undermine that. He also notes that BookTok activity is largely local rather than national in scale, which naturally fits the store-level autonomy model and makes centralized amplification less valuable than it might appear.

Daunt expected the Barnes and Noble turnaround to be faster than Waterstones given his prior experience, but the timelines have been almost identical. He attributes this to culture changing slowly regardless of the leader's experience, comparing it to a rubber band that snaps back the moment pressure is released. The implication is that the decentralized model requires sustained, active protection rather than a one-time structural change.

On AI-generated books, Daunt said Barnes and Noble will sell them if customers want them but that the company has a deep natural antipathy to books not written by real authors. Barnes and Noble actively filters AI-generated books from its online catalog and intends to ask publishers to indicate where books are AI-assisted or AI-written so the retailer can communicate that transparently to customers. Daunt predicted AI will take over certain categories such as home study guides but said it is very unlikely to produce the next great novels, and that the personal connection customers feel toward authors is something he does not expect to change.