How to Integrate a LINE Official Account: Rich Menus, Member Binding, and Segmented Push — Full Plan with Three Budget Tiers and a 90-Day Roadmap

In March we took on a five-store restaurant chain with 8,200 LINE friends. Their problem was specific: two broadcasts a month, message volume blowing past their monthly plan allowance, roughly NT$9,400 in extra message purchases, a 1.2% click rate, and just over 60 attributable in-store visits. The owner's first instinct was "rewrite the copy." The real issue was that those 8,200 friends existed in the system only as a string of characters beginning with U. Nobody knew who was a regular, who had not visited in six months, or which store anyone last ordered from.

Nearly every failed LINE Official Account looks like this. The account exists; it is simply disconnected from the company's own data. The value of integration was never "a good-looking rich menu." It is connecting the LINE userId to your members, orders, and purchase history.

When It Fits and When It Does Not

Custom integration makes sense when

• You have more than 3,000 friends and monthly message costs have become noticeable.
• You already run a membership system, POS, or e-commerce backend — the data is scattered but it exists.
• Customers repurchase (restaurants, beauty, cram schools, auto service, pet care, self-pay clinic services).
• You want customers to check orders, points, bookings, and renewals inside LINE without installing another app.
• You run multiple branches and need to segment by store rather than blasting everyone.

It does not fit when

• You have fewer than 500 friends and no steady acquisition channel — fix acquisition first.
• Your business is one-off transactions with almost no repurchase (wedding venues, single-move relocations). Push marketing has almost no marginal value.
• You have no usable customer data at all and the POS only prints a paper daily report.
• You only want LINE as a customer service inbox — the built-in chat features are enough, no development required.
• You want to "auto-broadcast to friends' personal LINE accounts." That violates LINE's terms of service. It is a hard line, and we do not do it.

Alternatives Matrix

The cheapest realistic path is usually a blend: run the official console manually for three months to validate content direction, then invest in custom integration once you are past 3,000 friends and broadcasts are demonstrably moving revenue. Building the expensive system first and discovering nobody reads the content is a mistake I have watched more than once.

The Build, Stage by Stage

• Stage 1: Data audit and channel architecture (2 weeks). Deliverables are a customer journey map and a channel mapping table. Draw the journey in Figma, track field mappings in Notion. The single most important decision here is which LINE Login channel will own member binding. userId is channel-scoped, so switching provider or channel turns the same person into a different ID, and merging later means re-binding everyone.
• Stage 2: Rich menu and auto-reply (2–3 weeks). Deliverables are the rich menu design and an FAQ decision tree. Use 2500×1686 or 2500×843 images, create the channel in the LINE Developers Console, then upload via API following the official rich menu setup flow. Remember the menu is permanent screen furniture occupying the bottom half of the phone: fewer tiles beats more, and six is usually the ceiling.
• Stage 3: Member binding and data flowback (3–4 weeks). Deliverables are a binding page (LIFF or a normal page with LINE Login) and a webhook receiver. Laravel or Node both work; what matters is that the webhook is HTTPS and returns 200 within a second. Use ngrok during development. The moment binding completes, write userId, member number, and home store into the same table.
• Stage 4: Segmented push and measurement (2–3 weeks). Deliverables are a segmentation rules table and a performance report. Every push link carries UTM parameters into Google Analytics 4 so you know which message produced actual orders. If payments are involved, fold ECPay's credit card rate (around 2.75% pre-tax for standard members, plus a per-order processing fee) into your per-order margin model at this stage.

Real Costs, Three Budget Tiers

Entry tier, NT$80,000–150,000. Rich menu design and deployment, auto-reply FAQ, welcome message, add-friend source tracking parameters, and a basic webhook receiver with message logging. Suited to brands with 1,000–3,000 friends who want the experience to feel right first.

Standard tier, NT$180,000–350,000. Everything above plus LINE Login member binding, a LIFF member center (orders, points, online booking), tag-based segmentation with a scheduling backend, and per-user rich menus that differ before and after binding. This is where most Taiwanese SMBs genuinely belong.

Advanced tier, NT$400,000–800,000. Everything above plus multi-store data sync, two-way POS or ERP integration, automated journeys (birthday, lapsed customer, post-purchase follow-up), payments and e-invoicing, and a shared customer-tag ticketing system.

The five hidden costs quotes most often omit:

• Plan fees and message top-ups. 8,000 friends broadcast twice a month is 16,000 messages; the High Volume plan covers 6,000, and the overage is billed at top-up rates starting from NT$0.2 per message — roughly NT$2,000 extra a month. Segmentation typically cuts that to a third.
• Rich menu design and revisions. NT$8,000–20,000 per set, usually revised two to four times a year for seasonal campaigns and new services.
• Hosting and HTTPS. NT$6,000–25,000 a year for the webhook service including domain and certificate.
• Premium (vanity) ID. An annual fee. Not necessary, but many owners want it.
• Operations and monitoring. Expired channel access tokens and dead webhooks are the classic silent failures. UptimeRobot monitoring plus a monthly check runs NT$30,000–60,000 a year.

What Clients Expect vs What Actually Happens

• "A friend equals a member." You receive only a channel-scoped userId — no name, no phone number. Getting those requires the customer to actively authorize through LINE Login or a binding form. A 35–50% binding rate is already a good result.
• "Pushing messages is free." Only proactive broadcasts are billed; reply messages sent through the Reply API and auto-responses are not counted toward message fees. So the genuinely economical design is to get customers asking first and answer with replies.
• "We can just edit the menu in the console." True for one shared menu. Showing different menus before and after binding is a per-user rich menu, which only the API can do — the console cannot.
• "LINE support can be fully automated." Auto-reply typically absorbs 40–60% of repeat questions; the rest needs a human. Hide the "talk to a person" button and the complaint goes straight to your Google reviews instead.
• "Integrate once and it's done." Tokens expire, LINE's APIs change, and marketing will want a new menu. This is a system to be maintained, not a one-time renovation.

Common Traps and How to Avoid Them

• Channel access token embedded in the frontend or app — that publishes the key to your account. Keep the token server-side only, proxy everything through your own API, and rotate on a schedule.
• Opening two or three official accounts on day one (HQ plus each branch) fragments friends and leaves userIds mutually unreadable. Start with one main account plus store tags; split only when a branch genuinely needs independent operation.
• No add-friend source tracking, so three months later you cannot say where friends came from. Give every channel (in-store QR, website, ads, packaging) its own add-friend URL parameter so acquisition cost becomes calculable.
• Broadcast-to-all as the only tactic, with the block rate climbing steadily. Segment before you push, and treat block rate as a core health metric — above 1.5% after a single broadcast means content and frequency need review.
• A slow binding page loses customers at the authorization step. The binding page is the conversion bottleneck; keep LCP under 2.5 seconds, compress images, and drop third-party tracking scripts.
• Pushing ID numbers, medical records, or itemized amounts straight into the chat. Messages should carry notifications and links only; sensitive content belongs behind a login on your own site. This is a hard line.
• Saving money with unofficial broadcast plugins. They violate the terms of service and your account can be suspended, wiping out eight thousand friends overnight. The savings never cover the risk.

Success Metrics and the 90-Day Roadmap

• Days 1–30. Finish the data audit and channel architecture, ship rich menu v1, attach parameterized add-friend URLs to every acquisition channel, and get the webhook receiving and logging events. Milestone: 90%+ of new friends attributable to a source, and weekly rich menu taps at 15%+ of friend count.
• Days 31–60. Launch LINE Login member binding, build the tag taxonomy, and cover the top 20 questions with auto-reply. Milestone: binding rate 35%+, auto-reply resolution 45%+, block rate held under 3%.
• Days 61–90. Ship segmented push and per-user rich menus, wire reporting into GA4. Milestone: segmented push click rate 8%+ (broadcast-to-all typically sits at 1–3%), NT$15 revenue per NT$1 of message spend, and monthly message volume down 30%+ versus the unsegmented baseline.

Four numbers to watch permanently after month three: binding rate, block rate, orders converted from segmented pushes, and revenue per message. The first two govern health; the last two govern return on investment.

Decision Checklist

• ☐ Do you have more than 3,000 friends?
• ☐ Have monthly message fees already exceeded your plan allowance?
• ☐ Do your customers repurchase?
• ☐ Do you have a membership system, POS, or e-commerce backend to connect?
• ☐ Do you know which channel each existing friend came from?
• ☐ Do you have more than one store that needs separate handling?
• ☐ Can you list the top 20 questions customers ask?
• ☐ Is someone responsible for reviewing the dashboard weekly?
• ☐ Can you produce content (offers, new items, education, tutorials) consistently?
• ☐ Do customers need to pay or book inside LINE?
• ☐ Can you accept a binding rate in the mid-thirties at the start?
• ☐ Is there a named contact who can answer spec questions within two working days?
• ☐ Do you have budget next year for menu revisions and content operations?

Fewer than six checked: get more out of the official console first and fix add-friend source tracking. Nine or more: custom integration usually pays back within six months.

Frequently Asked Questions

How much does LINE Official Account integration cost and how long does it take?

A rich menu and auto-reply build runs NT$80,000–150,000 over four to six weeks. The standard version with LINE Login binding, a LIFF member center, and segmented push runs NT$180,000–350,000 over eight to twelve weeks. Integrating POS, ERP, or multi-store sync lands at NT$400,000–800,000 over three to five months. Annual running cost — hosting, plan fees, message top-ups, maintenance — is roughly NT$40,000–90,000.

Why can't I get phone numbers from my 8,000 friends?

Because LINE gives you only a userId scoped to your channel, containing no personal data. To obtain names and phone numbers, customers must actively authorize through LINE Login or complete a binding form, with the purpose clearly disclosed. Incentives — points for first binding, orders visible only after binding — typically lift the binding rate from around 15% to about 40%.

How do I keep message fees from blowing the budget?

Start with the arithmetic: friends multiplied by monthly broadcasts equals message count. 8,000 friends pushed twice is 16,000 messages, while the High Volume plan includes only 6,000. Three levers bring it down: segment so only relevant audiences receive a message; convert anything auto-answerable into unbilled Reply messages; and put key information into the rich menu so customers come to you instead of you pushing outward.

Does the rich menu require development? Can't the console handle it?

One shared menu for everyone works fine in the official console. Development is needed in three cases: different menus before and after binding, different menus by membership tier or store, and tabbed menu switching. All three require the Messaging API; the console cannot do them.

Can I use the LINE broadcast plugins sold online?

Not advisable. Unofficial broadcast or auto-add-friend tools violate LINE's terms of service, and if the account is suspended the friend base you built disappears at once. These tools also tend to store your customer data somewhere you cannot audit. The correct way to reduce message spend is segmentation and unbilled Reply messages, not circumventing the rules.

Ready to Turn LINE Friends Into Actual Members?

ScriptWalker's LINE Official Account Integration service starts at NT$180,000 and covers rich menu design, LINE Login member binding, a LIFF member center, tag-based segmentation with a push scheduling backend, and a segmentation rules table your team can maintain. We start with one session laying out your customer journey and existing data, because what decides this project is whether the data connects — not how the menu looks.

• Email: [email protected]
• Phone: 0916-224-047
• LINE: @ufv9089p