How to Build a Customer-Service Ticketing / After-Sales Repair System (Costs + 90-Day Roadmap)

A home-appliance repair business with six stores had every service request buried across three LINE groups, one support phone and a handwritten notebook. Last month an air-conditioner warranty job was missed because "the message got scrolled away," the customer complained to consumer authorities, and cleanup cost two days and a free replacement unit. The owner asked us: "Is there a system where no request ever disappears and every technician knows which homes to visit today?" That is exactly what a customer-service ticketing / after-sales repair system solves.

When It Fits vs When It Doesn't

Good fit:

• More than 20 requests/complaints a day, scattered across LINE, phone and forms
• Dispatch needs: assign cases to different technicians/agents and track progress
• SLA or warranty management: different cases have different response deadlines
• Reconciliation and KPIs: volume per person per month, average time to close

Poor fit:

• Fewer than 5 requests a day — a shared inbox is enough; do not run a system for it
• Cases need no assignment or history — a LINE OA auto-reply suffices
• You have not defined what "closed" means — fix the process before the system

Alternatives Matrix

Full Process Breakdown (Stage + Time + Deliverable + Tools)

• Stage 1 · Requirements & process mapping (3-5 days): map the request lifecycle (intake→dispatch→handle→close→follow-up); deliver a flow diagram (Figma/Whimsical) and field list.
• Stage 2 · System design (1 week): data tables, permissions, SLA rules, LINE integration design; deliver a prototype (Figma).
• Stage 3 · Development (3-4 weeks): build the ticket core in Laravel, a responsive or Flutter console for technicians, integrate the LINE Messaging API for intake and notifications; deliver a testable build.
• Stage 4 · Test & launch (1 week): run real cases in parallel for a week, train staff; deliver the live system plus a manual.

Real Cost Breakdown

• Build: from NT$120,000 (single-store basic); multi-store / SLA / reports NT$220,000–350,000
• LINE official account: verification and message-volume fees (per LINE rates)
• Hosting: VPS ~NT$800–2,000/mo
• SSL: Let's Encrypt free; domain ~NT$400/yr
• SMS notifications (optional): ~NT$1–2 each
• Maintenance: recommended NT$6,000–15,000/mo for updates, backups, small changes

Reality vs Client Imagination

• Client thinks "a live system means no missed tickets"; reality: misses come from undefined process (who responds by when); the system only amplifies existing discipline.
• Client thinks "technicians will use it"; reality: field staff resist typing; the UI must be "three buttons to close," or they revert to LINE.
• Client thinks "get it perfect at launch"; reality: fields and notification rules always need tuning in the first two weeks — reserve an adjustment window.

Common Traps & How to Avoid Them

• Trap: too many fields, techs skip them → Fix: cap required fields at five.
• Trap: no SLA reminder, cases silently overdue → Fix: auto-escalate to a manager on breach.
• Trap: LINE and system data out of sync → Fix: the system is the single source of truth; LINE is only a notification gateway.
• Trap: no "closed" definition, KPIs uncomputable → Fix: define close and reopen rules explicitly.
• Trap: loose permissions, techs see all cases → Fix: scope visibility by store/role.

Success Metrics + 90-Day Roadmap

• Day 30: every request enters the system (channel-consolidation > 90%), missed tickets at zero.
• Day 60: average first-response and time-to-close have a baseline to watch.
• Day 90: use data to find bottleneck techs/stores, tune SLA and dispatch rules, start on automation (e.g., auto-dispatch).

Decision Checklist

• ☐ More than 20 requests/complaints a day?
• ☐ Cases scattered across 2+ channels?
• ☐ Need to assign cases to different people and track them?
• ☐ Warranty/SLA deadlines to manage?
• ☐ Owner wants per-person/per-store volume and close times?
• ☐ Can you clearly define "closed"?
• ☐ Will field staff adopt a new tool (with training)?
• ☐ Will stores/headcount grow in 3 years (affecting SaaS seat cost)?

FAQ

Why not just use Zendesk or Freshdesk?

Few stores, few users, standard flow — SaaS is the fastest answer; start on the free/low tier. But as headcount grows (per-seat fees rise linearly) and you need deep LINE and dispatch integration, a self-built one-off cost wins. Compare on three-year total cost, not month one.

Must it integrate LINE?

In Taiwan, almost always. Customers report via LINE; integrating the LINE Messaging API pushes requests straight into the system and replies via LINE — the key step to turn scattered group messages into trackable tickets.

What if technicians can't type in the field?

Simplify field use to the extreme: photo upload, dropdown status, one-tap close. Tap over type wherever possible — this decides whether the system actually gets used.

How long to launch?

Single-store basic ~6–8 weeks; multi-store with SLA and reports ~10–12 weeks. The real variable is not development but whether the process can be defined first.

Call to Action

ScriptWalker builds customer-service ticketing / after-sales repair systems from NT$120,000, including LINE integration and a technician console. Want a free process-mapping and quote first? Contact us:

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