The three types, by what they actually own
Front-end developers own everything a user sees and touches: layouts, interactions, accessibility, performance in the browser. Back-end developers own everything a user never sees: servers, databases, authentication, integrations, the business logic that moves money and data. Full-stack developers work across both, which makes them efficient for common projects and shallow for hard ones, because the stack is two professions deep and nobody masters both ends equally. The hiring overview covers the types in breadth; this page is the decision rule.
The decision rule: hire where the risk lives
Every project has one side where failure hurts most, and that side picks the specialist. A brand relaunch where the design must land perfectly → front-end lead (with a designer, the design-vs-development guide untangles that overlap). A portal moving sensitive data with integrations and permissions → back-end lead. A straightforward marketing site or a store on a mature platform → a competent full-stack developer is the efficient choice, because neither side carries unusual risk. The mistake buyers make is hiring full-stack for projects whose risk is entirely on one side, paying generalist rates for the side they did not need and hoping on the side they did.
The full-stack trap, stated fairly
"Full-stack" is the most inflated title in the industry: the phrase is honest when it means "competent on both ends," and inflated when it means "I followed a bootcamp that touched both." The vetting tell is depth probes: ask a claimed full-stack developer about browser rendering one minute and database indexing the next, real ones answer both with specifics, shallow ones redirect to buzzwords. The vetting guide has the full probe list. For genuinely two-sided projects (a custom design AND a complex backend), the honest answer is two specialists or one senior full-stack with a proven track record on exactly that shape, the dedicated-model page covers how that engagement works.
What each type costs, briefly
Rates cluster by scarcity and risk: back-end and full-stack with real data experience price above pure front-end at the same seniority, because the failures they prevent cost more. Regional rates and the full context live on the pricing page and the region pages; the calculator prices your specific shape. Bring the project description and where you think the risk lives, the scoping call either confirms it or moves it, free of charge either way.