Maintenance: Of Everything, Part One
Stewart Brand — the man behind the Whole Earth Catalog, the Long Now Foundation, and How Buildings Learn — turns his attention to the vast, invisible labour of keeping things working.
His argument is that maintenance, not innovation, is what sustains civilisation, yet it receives almost no prestige, funding, or serious study.
Brand draws on examples from infrastructure, software, buildings, and biological systems to show that neglect compounds silently until failure becomes catastrophic.
For anyone directing a digital product, this is a direct confrontation with the tension between building new features and preserving what already exists.
The book extends the thesis of How Buildings Learn into every domain where human-made systems must endure.
It reframes technical debt, operational investment, and platform reliability as questions not of engineering trade-offs but of institutional values.
Central argument
Brand argues that maintenance — not innovation — is the foundational activity that sustains civilisation, and that its near-total absence from institutional prestige, funding, and serious intellectual study is itself a civilisational risk. Drawing on infrastructure, software, buildings, and biological systems, he demonstrates that neglect does not produce visible decay but rather silent compounding failure, which erupts catastrophically only after the point where intervention was still affordable. The book reframes How Buildings Learn's core insight — that systems must absorb change over time to survive — and extends it universally: what we choose not to maintain reveals what we actually value, not what we claim to.
Critique
Brand's argument, compelling as it is, risks overstating maintenance as a corrective by treating it as an almost unconditionally virtuous activity — but not everything warrants preservation, and the harder institutional question is how to distinguish systems worth sustaining from those that should be deliberately retired or replaced. A CPO reading this could absorb the thesis and use it to justify defending legacy architecture or inherited product surface area that genuinely ought to be abandoned. The book's breadth of examples across domains also means it occasionally trades analytical precision for rhetorical sweep, making it easier to feel persuaded than to operationalise the argument within a real organisational context.
Why it matters for product
For a product director, Brand's thesis reframes technical debt and operational investment not as engineering concerns that periodically surface in planning — but as a direct expression of organisational values, which means the decision to defund maintenance is a strategic choice with compounding consequences, not a neutral resource trade-off. Concretely, it challenges the common product org structure where reliability and platform work sit beneath roadmap velocity in prioritisation hierarchies: Brand's framework would read that structure as an institution systematically choosing prestige over sustainability. It also gives product leaders a non-technical vocabulary to argue for maintenance investment in executive and board conversations where engineering framing tends to lose.