“Should enterprises do e-commerce?” This topic is raised again every so often. But if it is reduced to a yes-or-no question, it is easy to reach the wrong conclusion—because the real question most enterprises face is not “whether to do e-commerce,” but “given that we already have stores on third-party platforms, should we build another set of our own?”
This question is not suitable for answering with a stance; it is suitable for answering with costs.
The first calculation: Where does traffic come from
Third-party platforms come with their own traffic entrances. Users arrive with shopping intent, and conversion is relatively direct. The price is promotion costs and changes in platform rules.
A self-built platform has no ready-made customer flow. It requires the enterprise to solve the problem of “how people come” on its own: attracting visitors through search, content, campaigns, or other channels. If this capability is lacking, the actual transactions after the self-built platform goes live will be far lower than expected.
The first step in judgment is to look at whether the enterprise currently has a stable source of its own traffic.
The second calculation: Which side are user habits on
Users who shop on platforms have already formed fixed habits: they need to compare prices, read reviews, and confirm after-sales service in the same place. Getting these users to switch to a single brand’s website to place orders requires a strong reason, and that reason is usually not “this is our official website.”
When products are unique, have strong repurchase attributes, or when customers are themselves corporate buyers, the persuasion cost of self-built channels will be somewhat lower; conversely, for highly standardized and easily replaceable products, self-built channels find it very difficult to compete with platforms.
The third calculation: Who will bear long-term maintenance
A self-built e-commerce platform is not a page, but a continuously operating system: payment, orders, inventory, after-sales, customer service, and security. Each item requires someone to be responsible for it, and also requires ongoing technical investment. What should be included in the evaluation is not the development cost, but the annual maintenance cost after launch.
This point is often underestimated. Project budgets often only account for development, not operations.
A more prudent sequence
Back to practice, many enterprises adopt a two-step approach: first, get transactions running on third-party platforms to validate products and prices; at the same time, use the official website to handle brand display, content accumulation, and customer lead capture, so as to receive people coming from various channels.
Once the volume of owned customers, the repurchase rate, and operational capabilities are all in place, then consider migrating the transaction process to self-built channels. This makes risks more controllable, and avoids paying a development and maintenance cost that is difficult to recover when both capabilities and needs are still unclear.
E-commerce is a trend, but it is not a necessary option for all enterprises, and even less is it the first step for all enterprises.