"Who are the 15 accounts most likely to close this month?" The owner asks, and the sales manager spends 15 minutes flipping between a spreadsheet, a LINE group, and a phone call to a rep just to piece together an answer. One industrial-parts trading company once tallied it up: at least NT$2 million in pipeline value evaporated in a year because "the day a salesperson quit, 30 active deals disappeared with them." The problem was never a shortage of customers — it was that customer data and sales progress never lived in one shared home.
When It Fits vs. When It Doesn't
A custom CRM isn't for every business. Check which side you fall on first.
| Good Fit | Not a Good Fit |
|---|---|
| Sales team of 5+ with a complex follow-up cadence | 1-2 salespeople with deal cycles under a week |
| A clear staged sales process (inquiry → quote → contract) | One-off transactions with no long-term follow-up (e.g. retail) |
| Needs LINE OA, email, and phone logged in one place | A single, low-volume contact channel (under 30 leads/month) |
| Needs dashboards for owners or investors | Business model is still being validated |
| Spreadsheets/LINE already causing dropped leads or mismatched data | Budget under NT$100,000 with no room to phase spending |
Alternatives Matrix
| Option | Pros | Cons | Cost Range |
|---|---|---|---|
| Excel / Google Sheets | Free, quick to start | No automation, collaboration conflicts, no visual pipeline | NT$0 |
| Global SaaS (HubSpot, Salesforce) | Feature-rich, mature ecosystem | English-first UI, cost scales with seats, low customization, data hosted overseas | NT$800-3,000+ per seat/month |
| Lightweight SaaS (Pipedrive, local CRM tools) | Affordable, ready out of the box | Limited reporting depth; LINE OA/local invoicing integrations often need extra work | NT$300-900 per seat/month |
| Custom CRM (ScriptWalker) | Matches your actual workflow, you own your data, room to grow into an app and dashboards | Higher upfront cost, 4-12 week build | From NT$120,000 |
Core Modules and the Full Build Process
A complete custom CRM typically has four modules: a contact list (people, companies, tags), a sales pipeline (kanban-style stage management), automated follow-ups (due-date reminders, LINE/email auto-sends), and a reporting dashboard (revenue, conversion rate, rep leaderboard). Here's the build process and deliverables:
| Phase | Timeline | Deliverable | Tools |
|---|---|---|---|
| Requirements interviews & data audit | Week 1 | Field spec sheet, data health report | Notion, Google Sheets |
| Pipeline & wireframe design | Weeks 2-3 | Wireframes, database ER diagram | Figma |
| Backend & API development | Weeks 4-9 | Testable admin panel, LINE OA/email automation | Laravel, Filament, LINE Messaging API |
| Testing & data migration | Weeks 10-11 | Test report, import results from Excel/legacy system | Postman, PHPUnit |
| Training & launch | Week 12 | User manual, training recording | Loom, Notion |
Tech Considerations
We typically build the backend on Laravel with Filament for a fast admin interface; MySQL or PostgreSQL both work, but the real priority is getting the customer-contact-deal relationship right early, so you don't have to restructure tables later when reporting needs grow. If reps spend most of their day on the road, pair it with a Flutter mobile app so they can update a deal's stage on-site instead of catching up at the office later. When integrating LINE OA, note that Messaging API pricing is billed by message volume — estimate monthly cost upfront if automated follow-ups will be sending a lot of messages, so you're not surprised after launch. On the security side, build in role-based permissions (reps see only their own accounts, managers see everything) and a basic data-privacy review.
Real Cost Breakdown
| Tier | Scope | Dev Cost | Timeline |
|---|---|---|---|
| Lite | Contact list + basic pipeline board | NT$120,000-180,000 | 4-6 weeks |
| Standard | + automated follow-ups (LINE/email) + reporting dashboard | NT$250,000-400,000 | 8-12 weeks |
| Advanced | + Flutter sales app + multi-role permissions + third-party API integration | NT$450,000-800,000+ | 12-20 weeks |
Hidden costs that often get missed: LINE Messaging API fees (billed by plan and message volume — see LINE's official documentation), email delivery services like SendGrid at roughly NT$300-1,500/month, cloud hosting and database around NT$1,500-5,000/month, SSL certificates (usually free on most cloud platforms, enterprise tiers cost extra), and ongoing maintenance at roughly 15-20% of the build cost per year, or NT$3,000-10,000/month.
What Clients Expect vs. What Actually Happens
- Expectation: Just import the old spreadsheet and you're done. Reality: Data needs cleaning, deduplication, and field standardization first — usually 3-5 extra working days.
- Expectation: Reps will faithfully log data on their own. Reality: Without a simplified workflow and reminders, logging rates often fall below 30%.
- Expectation: The dashboard will be accurate on day one. Reality: Dashboards usually need 4-8 weeks of clean data before the numbers mean anything.
- Expectation: Switching to a CRM will improve revenue by itself. Reality: A CRM is just a tool — without a follow-up SOP and manager review, the impact is limited.
Common Pitfalls and How to Avoid Them
- Pitfall: Trying to cram every department's workflow in from day one. Fix: Focus first on sales follow-up and pipeline, and roll out other modules in phases.
- Pitfall: Too many required fields, so reps avoid filling them in. Fix: Cap required fields at 5 or fewer; default or auto-populate the rest.
- Pitfall: Web-only backend, so reps can't update anything while out visiting clients. Fix: Pair it with a Flutter app, or at minimum a responsive (RWD) interface.
- Pitfall: Migrating messy legacy data as-is. Fix: Schedule a "data cleanup week" before launch, with someone assigned to dedupe and standardize formats.
- Pitfall: Generic, robotic automated follow-up messages get low reply rates. Fix: Write different templates per deal stage, and A/B test them regularly.
- Pitfall: No role permissions or audit log, so deleted data can't be traced. Fix: Add tiered role permissions and an audit log for key actions.
Success Metrics and the 90-Day Post-Launch Roadmap
Launch isn't the finish line — it's when the data starts to mean something. Here's what to track at 30/60/90 days:
- Days 1-30: Goal is "data gets in." Track: new contacts logged daily, rep logging completion rate (target above 80%), duplicate rate.
- Days 31-60: Goal is "the process gets used." Track: days spent per pipeline stage, automated message open/reply rates, weekly update frequency per rep.
- Days 61-90: Goal is "the dashboard drives decisions." Track: close rate, average deal cycle, rep leaderboard — then adjust follow-up scripts and stage definitions based on the data.
Decision Checklist
- ☐ Sales team of 5 or more people
- ☐ Sales process has clear, distinct stages
- ☐ More than 30 new leads or deals per month
- ☐ Have lost customer data before due to a rep leaving or being on leave
- ☐ Owner/manager needs weekly revenue reports
- ☐ Have LINE OA or email that needs automated follow-up
- ☐ Reps travel frequently and need mobile updates
- ☐ Existing Excel/legacy system already causes dropped leads or mismatched records
- ☐ Budget can accommodate NT$120,000+ with phased payments
- ☐ Willing to invest at least a week in data cleanup and migration
- ☐ Someone internally can own CRM training and rule maintenance
- ☐ Willing to review dashboards and adjust process regularly through the first 90 days
Frequently Asked Questions
What's the difference between a CRM and Excel? When should we switch?
Excel works fine when customer volume is low and the process is simple. The signal to switch is when you start seeing multiple people editing the same sheet, data disappearing when a rep leaves, or the owner needing real-time numbers that only exist after someone manually compiles them.
Is a custom CRM or a global SaaS like HubSpot the better value?
With fewer than 5 salespeople and a simple process, lightweight SaaS is usually more cost-effective. With a larger team, a need for local integrations like LINE OA or e-invoicing, and a 3+ year horizon, a custom CRM's total cost of ownership is usually lower since it doesn't keep scaling with seat count.
We only have 5 salespeople — do we really need something this complete?
No — start with the Lite tier (contact list + pipeline). Once you've confirmed the process is actually being used, add automated follow-ups and reporting in later phases to reduce the upfront risk.
How soon after launch will we see revenue improve?
Most clients see improved conversion or follow-up efficiency within 60-90 days. The first 30 days are mostly about building data and habits — don't judge the system by revenue numbers in month one.
Can we import old data from Excel or a legacy system? Is it a hassle?
Yes, but it needs field mapping and cleanup first — usually 3-5 working days. If the data volume is very large or the format is especially messy, we'll estimate additional hours separately.
Closing Thoughts and Next Steps
A CRM's value isn't in how many features it has — it's in whether your sales team actually opens it. When we plan a build, we cut required fields to the bare minimum first, automate follow-up reminders as much as possible, and only then talk about reporting and expansion. If scattered customer data and invisible revenue trends are causing you pain, let's talk about where you are now — we can help you figure out which module to start with for the best return.
- Email: [email protected]
- Phone: 0916-224-047
- LINE: @ufv9089p