Usual mix
Customers and members use a WeChat or Alipay Mini Program; employees use an APP or WeCom; campaign pages use H5. The back office is still one set. WeChat or Alipay. See WeChat Mini Program or Alipay Mini Program.
The main entry is not “which looks more premium.” It is which door people will walk through every day. Mini programs skip install and share easily. An APP icon stays, with stronger push and device features. A matrix beats a vote. Making an APP the main entry just to have one is the failure in Why shipping an APP just to “have one” fails. The boundary for going native. See When you actually need a native APP.
| Look at | Mini program as main entry | APP as main entry |
|---|---|---|
| Install cost | Scan and use. Fits customers, stores, suppliers | Must pass a store. Internally you force distribution |
| Frequency | A few times a week: book, status, redeem | Many times a day: report, approve, deliver |
| Push | Subscribe messages are limited and easy to miss | OS notifications are stabler. Approvals die less often |
| Device and offline | Camera scan is enough. Offline is weak | Bluetooth, background location and offline queues are fuller |
| Reach and review | Links, posters and official accounts can enter | Listing, updates and many channels all need care |
Do not run two main entries
Asking one role to download an APP and also use a mini program teaches nobody. There can be only one main entry. The other is a capability add-on. When H5 alone is enough. See Is H5 enough on its own.
Lock the role, then fill this table
Employees, customers, stores and suppliers are not the same row. Store redemption needs a fast scan—a mini program is often enough. Field work offline all day wants an APP. Suppliers who occasionally take orders and settle usually win with a mini program. See How supplier orders and settlement should run. Employee feature scope is in Employee features people will actually use. Fill the matrix and the main entry is no longer a taste question.
DaXi does not default to “build an APP first.” We fill the table by who uses it, how many times a day, and whether there is a network. To pick together, go to the Apps & Mini Programs and name the roles. Brief-to-launch order is in From brief to launch: how a mobile build runs.
Do not fill the matrix once. When work moves from “customers book occasionally” to “riders take orders all day,” the main entry may move from mini program to APP. Conversely, if internal approvals already happen at high frequency in WeCom, another standalone APP as main entry splits people. How you write member home depends on whether the main entry is a mini program. See What belongs on a member mini program home.
One company can have two main entries if they are split by role: employee APP, customer mini program. Do not let the same role finish the same job in two containers. Training, posters and store copy teach one path so the funnel does not leak on “which one do I download.”
Once the main entry is set, teach only that path
Store shots and training teach only that path. Teaching both is no main entry. The download funnel is in What to do when nobody downloads the APP. A mini program as main entry may still need WeCom or SMS to cover weak push. That is not a reason to build another APP unless frequency and device features already pay the install cost. Write “must be native” as acceptance items: offline queue, Bluetooth, background location. If you cannot write the items, do not default to native.