liveBook · Manning

Degradation

As explained in Fast by Default

A condition in which a software product or system performs more slowly or inefficiently than required.

Degradation is a condition in which a software product or system becomes slower or less efficient than required. It often develops without crashes or visible errors: a delayed response, sluggish interaction, or slow-loading layout gradually weakens the user experience. section 2.1 section 2.1.4

The recurring pattern is simple: a technical change appears acceptable in isolation, reaches production, and then compounds across real sessions. Users rarely report the underlying cause; they leave, convert less often, engage less, or switch competitors. Degradation therefore affects money, time, trust, and team productivity rather than performance metrics alone. section 1.2 section 2.1.4

Recognize degradation through recurring symptoms and costs:

  • Measure user and business impact through conversion changes, retention, search or app-store visibility, and support consequences. section 2.0 section 2.6
  • Treat small delays as cumulative: a sluggish interaction or delayed response can determine whether users convert, engage, or leave. section 2.1.4
  • Track internal degradation through reactive investigations, recurring regressions, wasted engineering time, and planned work displaced by firefighting. section 2.3.1 section 2.3
  • Distinguish individual-user speed problems from scalability and resilience problems before selecting architectural remedies. section 5.1

A degradation pattern in practice

A landing page loads only slightly more slowly after a change. That delay seems harmless during review, but across thousands of sessions it reduces engagement and sales. Similar degradation can appear as scripts injected through a tag manager, startup lag on older mobile devices, inefficient database queries, or frame drops in a desktop interface. section 2.1.4 section 2.3.1

How Fast by Default covers it

How deep it goes
Traces degradation from subtle production symptoms through user and organizational costs, but stops short of implementation-level remediation.
Examples
Uses landing pages, APIs, mobile startup, database queries, desktop interfaces, and injected scripts as degradation scenarios.
What you can do after
Evaluate degradation by measuring conversion, retention, productivity waste, visibility, and the cost of a high-value flow.
How it connects
Connects degradation to performance engineering, architecture, scalability, resilience, observability, conversion, retention, and trust.

Also explained in

Compare how 6 books explain degradation →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Fast by Default. The section links point to the parts of the book it draws on.