Linked: The New Science of Networks
Barabási's book introduced the science of networks to a popular audience: scale-free networks, preferential attachment, hubs, the small-world property, and the mathematics that explains why the internet, social networks, disease transmission and cellular metabolism share the same structural patterns.
The book is the accessible version of research that reshaped how we understand interconnected systems.
For product direction the frameworks are directly applicable — platform dynamics, viral growth, network effects, and the vulnerability of systems with concentrated hubs are all applied network science.
Read alongside Castells for the sociological layer and Shapiro and Varian for the economic implications.
Barabási writes clearly; the science has aged well.
Central argument
Barabási argues that most real-world networks — from the internet and social graphs to cellular metabolism — are not random but share a common structural signature: a scale-free topology generated by preferential attachment, where new nodes disproportionately connect to already well-connected nodes. This mechanism produces hubs, a small number of nodes with vastly more connections than the average, which explains both the small-world property (short path lengths between any two nodes) and the asymmetric resilience of these systems — robust against random failures but acutely vulnerable to targeted attacks on hubs. The central finding is that this pattern is universal across biological, social, and technological networks, meaning the same mathematical laws govern phenomena as different as viral epidemics and hyperlink structures.
Critique
The scale-free model and preferential attachment have faced significant empirical revision since 2002: subsequent research has questioned whether true power-law degree distributions are as ubiquitous as Barabási claimed, with more rigorous statistical testing finding that many networks fit alternative distributions equally well or better. There is also a determinism in the framework that can obscure the role of institutional, historical, and intentional design factors — treating network structure as an emergent inevitability risks underweighting the degree to which platforms and infrastructures are shaped by deliberate choices, regulatory environments, and incentives that preferential attachment alone cannot explain.
Why it matters for product
The hub dynamics of scale-free networks map directly onto platform strategy: a CPO designing for network effects needs to understand that early preferential attachment is not self-correcting — the nodes that accumulate connections first tend to dominate permanently, which means decisions about onboarding, API openness, and ecosystem incentives in early product stages have compounding structural consequences that are extremely difficult to reverse. The vulnerability of hub-concentrated systems also has organisational implications: product architectures or partner ecosystems built around a small number of high-dependency integrations are brittle in exactly the way Barabási describes, making hub identification a legitimate input to technical and strategic risk assessments.