Total cost of ownership

Total cost of ownership covers everything an ecommerce platform costs across its useful life. License fees. Initial build. Ongoing development. Security patches. Hosting bills. The people who keep all of that running. Five years out, the bill looks very different from what the platform vendor quotes in year one. For a headless commerce setup specifically, TCO splits into roughly four buckets. ## License costs License costs cover the commerce engine (Norce, Shopware, commercetools), the CMS (Storyblok), the hosting (Vercel), and any payment, search, or CRM tools. Predictable. Scales with traffic or feature usage. Easiest to compare across vendors and easiest to forecast. ## Build costs Build costs cover the frontend, the integrations, the design system, the data model, the initial migration. A custom headless frontend typically takes 4 to 9 months. A productised frontend like Frntkey replaces most of that with a packaged baseline. The build bucket is where the largest variance between options lives. ## Operations costs Operations costs cover monitoring, security updates, dependency upgrades, CMS user training, and ongoing development for new features. Headless commerce demands more operational attention than a templated platform because more moving parts are involved. Vercel reduces some of this. A well-packaged frontend reduces more. ## People costs People costs cover the internal team or agency partner that keeps the storefront alive. Often invisible until something breaks. Often the single largest line item across five years. Pick a stack your team can hire for. ## The pattern across projects A pattern shows up across replatforming projects. License cost is the smallest of the four, even though it dominates the selection conversation. Build and operations costs dominate the five-year picture. People costs become the gating factor for whether the chosen stack stays healthy. ## Where Frntkey fits Frntkey targets the build and operations lines directly. Most Frntkey storefronts launch in 6 to 12 weeks instead of the 4 to 9 months a fully custom headless build takes. The packaged frontend, pre-installed integrations, and bundled Storyblok CMS compress the build bucket. The structural choice of Nuxt and Vercel keeps operational overhead reasonable for teams that are not running a dedicated platform engineering function. The honest tradeoff: a packaged frontend means accepting product decisions made by someone else. If your storefront needs a fundamentally different architecture (a complex multi-tenant marketplace, an unusual checkout flow), a custom build still wins on fit. For most B2B and D2C merchants on Norce or Shopware, the TCO math favours packaging. When evaluating TCO across alternatives, look beyond year one. The cost of replatforming again in three years if the initial architecture does not scale dwarfs almost any year-one decision. Pick the stack you can still operate in year three with the team you can hire.