做APP、微信小程序、H5还是网站？2026老板选型指南

关键工程指标 (Key Engineering Metrics)

• 业务型小程序参考报价 (Business mini program price)：3万–8万
• 业务型MVP常见工期 (Typical business MVP timeline)：4–8周
• 跨端可节省重复开发成本 (Cross-platform savings)：40%–50%

工程落地核查清单 (Architecture Checklist)

• 先判断客户主要从微信、搜索、广告还是应用商店进来
• 按使用频率区分：低频交易先小程序，高频留存再做APP
• 列出是否必须用蓝牙、推送、相机、定位、硬件连接等手机能力
• 先用MVP验证核心闭环，再根据真实数据决定是否开发原生APP
• 要求报价单写清模块、工期、里程碑、源码归属和后续新增需求规则

一、先看获客渠道和使用频率，别一上来就做APP

1. Start with acquisition channels and usage frequency, not an app

结论：靠微信成交、用户使用不频繁的业务，优先做小程序；靠搜索获客和建立信任，先做官网；需要每天打开、留存和强运营，再考虑APP。很多老板把“做个软件”想得太笼统，实际上渠道决定入口，频率决定形态。用户从公众号、朋友圈、微信群、微信支付进来，小程序路径最短；从百度、谷歌、必应搜索进来，网站更适合承接内容和信任；投放活动页、裂变页、表单预约，H5更轻。建议不要一开始押注大而全平台，先用业务型MVP跑通登录、列表、支付、后台等核心闭环，常见4–8周，再根据真实使用数据决定是否加APP。

Choose a mini program for WeChat-driven, low-frequency transactions; a website for search acquisition and credibility; and an app only for daily use, retention, and heavy operations. The channel determines the entry point, while frequency determines the product form. WeChat traffic converts faster in a mini program, search traffic is better served by a website, and campaigns or forms can use lightweight H5 pages. Validate the core workflow with a business MVP, usually in 4–8 weeks, before investing in a full app.

• 微信私域、扫码、支付、分享：优先小程序
• 百度/谷歌搜索、品牌展示、外贸获客：优先官网
• 活动报名、广告落地页、短期营销：优先H5
• 高频使用、会员留存、复杂运营：再考虑APP

二、再看手机能力：蓝牙、推送、硬件和动画会影响成本

2. Check device capabilities: Bluetooth, push, hardware, and animation affect cost

结论：只要涉及蓝牙、硬件连接、强动画、复杂推送或系统级能力，就不能只按普通H5或简单小程序估预算。普通展示、预约、下单、会员、支付，用小程序或H5通常够用；但如果要连接设备、稳定推送、调用复杂传感器，APP往往更合适。需要注意，Flutter或uni-app跨端开发通常可省40%–50%重复开发成本，但强动画、蓝牙、硬件场景仍可能需要原生适配，不能简单理解为“一份代码完全不用原生”。报价时要把账号和第三方费用单独看清：Apple开发者账号99美元/年，微信小程序认证300元/年，域名、服务器、SSL证书、短信按服务商实际价格另计。

Do not estimate a project as ordinary H5 or a simple mini program if it needs Bluetooth, hardware connection, rich animation, sophisticated push notifications, or system-level features. Mini programs or H5 are usually enough for display, booking, ordering, membership, and payment, while an app is often better for devices, reliable push, and complex sensors. Flutter or uni-app can reduce duplicated development cost by 40%–50%, but advanced animation, Bluetooth, and hardware may still require native adaptation.

• 普通展示、下单、预约、支付：小程序/H5通常够用
• 蓝牙、硬件、传感器、强动画：提前评估原生适配
• 跨端能省钱，但不能假设所有能力都一份代码解决
• Apple账号、小程序认证、服务器、短信等费用要单独列明

三、预算和工期不确定，就按MVP分阶段做

3. When budget and timeline are uncertain, build an MVP in stages

结论：预算有限时，不要把官网、小程序、APP一次全做，先做能验证成交的最小版本。2026年参考价看，展示型小程序一般1万–3万；含登录、支付、列表、后台的业务型小程序3万–8万；带多角色、营销、消息、数据看板的8万–20万；原生APP加后台通常15万起，复杂平台30万以上。低于8000元却承诺“全功能”的，大概率是模板套壳。工期上，业务型MVP常见4–8周，复杂平台2–4个月；APP参考4–8周起，小程序参考2–5周起。建议先梳理需求，拿到逐模块报价，再按3–4个里程碑付款，每期看到可运行成果再付下一笔。

If budget is limited, do not build a website, mini program, and app all at once. Start with the smallest version that can validate transactions. In 2026, a display mini program is typically RMB 10k–30k; a business mini program with login, payment, lists, and admin is RMB 30k–80k; advanced multi-role and dashboard projects are RMB 80k–200k; a native app with backend usually starts at RMB 150k, with complex platforms above RMB 300k.

• 展示型小程序：1万–3万
• 业务型小程序：3万–8万
• 复杂多角色平台：8万–20万
• 原生APP加后台：通常15万起，复杂平台30万以上
• 先MVP验证，再按数据决定是否扩展APP

四、小程序验证后再做APP，优先考虑uni-app或Flutter复用

4. Validate with a mini program first, then reuse with uni-app or Flutter

结论：路线上建议“官网获客＋小程序成交＋APP服务高频用户”，不要反过来先花大钱做APP。小程序验证通过后，如果用户确实每天使用、需要推送或设备能力，再做APP更稳。技术上可评估uni-app或Flutter，把页面、业务逻辑、接口、后台尽量复用，减少40%–50%重复开发成本；但蓝牙、硬件、强动画要预留原生适配。合作时还要把底线写进合同：每周给一个可运行版本，源码、数据库脚本、接口文档和部署文档验收结清后100%归客户；上线后30天免费修Bug，非新增需求；需求冻结后新增内容走变更报价，避免中途模糊加价。

A practical route is “website for acquisition, mini program for transactions, app for high-frequency users,” not the reverse. Once the mini program proves demand, build an app if users need daily access, push notifications, or device capabilities. Evaluate uni-app or Flutter to reuse pages, business logic, APIs, and backend services, potentially reducing duplicated work by 40%–50%. Keep Bluetooth, hardware, and rich animation budgeted for native adaptation, and put source-code ownership, weekly runnable builds, milestones, and change-request pricing in the contract.

• 先用官网和小程序跑通获客与成交
• 高频、强推送、硬件需求明确后再做APP
• uni-app/Flutter可复用页面、逻辑、接口和后台
• 每周要有可运行版本，源码归属必须写进合同

常见问题解答 (FAQ)

问：我是第一次做软件，到底先做APP还是小程序？ (This is my first software project. Should I build an app or a mini program first?)

答：多数老板应先做小程序。你的客户如果在微信里沟通、支付、分享，小程序入口更短，开发和推广成本也更低。等用户使用频率、留存和设备能力需求被验证后，再做APP更稳。 Most business owners should start with a mini program, especially if customers communicate, pay, and share inside WeChat. It has a shorter path and lower validation cost. Build an app after frequency, retention, and device needs are proven.

问：H5、小程序和官网有什么区别？ (What is the difference between H5, a mini program, and a website?)

答：官网更适合搜索获客和品牌信任，H5适合活动页、投放页和表单，小程序适合微信内交易和复购。简单说，陌生客户先看网站，微信内成交用小程序，短期活动用H5。 A website is best for search traffic and credibility, H5 for campaign and landing pages, and a mini program for transactions and repeat purchases inside WeChat. Use a website for discovery, a mini program for conversion, and H5 for campaigns.

问：做uni-app或Flutter，是不是就能便宜一半？ (Will uni-app or Flutter cut the cost in half?)

答：通常可以减少40%–50%重复开发成本，但不是所有项目都能直接减半。普通业务页面复用度高；蓝牙、硬件、强动画、复杂推送仍可能需要原生适配，报价时要单独评估。 They can often reduce duplicated development by 40%–50%, but not every project is cut in half. Standard business pages reuse well, while Bluetooth, hardware, rich animation, and complex push may still require native work.

问：预算不多，怎样避免做一半加钱？ (With a limited budget, how can I avoid surprise price increases midway?)

答：先做逐模块书面报价，需求冻结后新增需求一律走变更报价。付款按3–4个里程碑走，每期看到可验证成果再付下一笔，并要求每周给一个可运行版本，合同写清源码归属和Bug维护期。 Get itemized written pricing first, and handle all new scope after the freeze through change quotes. Pay across 3–4 milestones only after verifiable results, require a weekly runnable build, and put source-code ownership and the bug-fix period in the contract.