A 40-person trading company. Its only engineer resigned at the start of the month and leaves at the end of it. The ERP integrations, the website, the automated shipping scripts - all of it lives in his head: no documentation, no backup, passwords stored in his browser. The owner has 30 days. Many SMEs run their entire systems on 1-2 people exactly like this. This article is a 30-day handover playbook you can follow step by step.
Industry Myths, Debunked
- "Handover means handing over the code" - code without environment configs, secrets, and a deployment process is a half-finished product. You can't even run it.
- "One month is enough for a full handover" - verbal knowledge needs structured extraction. In 30 days you can only complete the "life-saving layer": access, deployment, scheduled jobs, and emergency response.
- "Hire a replacement engineer first" - recruiting takes 3-6 months on average. The right order is stop the bleeding and take inventory first, then decide between hiring and outsourcing.
- "We'll write the docs later" - once the person is gone, there is no later.
Core Framework: The 30-Day, 6-Step Rescue Playbook
| Timeline | Step | Key actions |
|---|---|---|
| Day 1-3 | Freeze and inventory | Freeze all non-essential changes; list every system, every account, every piece of work in progress |
| Day 4-7 | Reclaim access and assets | Transfer domain, servers, DB, and third-party service ownership to company accounts; move GitHub repos per the official transfer guide and rotate API keys per the Stripe key management docs |
| Day 8-14 | Documentation sprint | Deployment runbook, architecture diagram, scheduled-task list, recorded video walkthroughs |
| Day 15-21 | Evaluate the successor | In-house hire vs outsourcing, scored with the KPI scorecard below |
| Day 22-30 | Shadow run | The successor performs one real deployment and handles one real ticket while the departing engineer is still present |
| Day 90 | Review | Check incident count, response time, and documentation coverage; decide the long-term model |
The order cannot be reversed: documentation written before access is reclaimed may become a door you can no longer open once the person leaves.
Three Typical Scenarios Compared
| Company | System profile | Recommended route |
|---|---|---|
| 10-person e-commerce | Small footprint, mostly SaaS | Move straight to an outsourced monthly retainer; handover takes about 2 weeks |
| 40-person trading firm | Custom-built ERP integrations | Have an outsourcing partner take over for 3 months to stabilize, then decide between hiring and continuing |
| 120-person manufacturer | Has an IT department, but the leaver is the only person who understands one system | Run a knowledge-extraction and documentation project so the existing team can take over |
The Complete Hidden-Cost List
- Replacement recruiting takes 3-6 months; a mid-level engineer costs roughly NT$900K-1.4M per year, plus a headhunter fee of 20-30% of annual salary.
- Rebuilding an undocumented system often costs 30-50% of the original development fee.
- Downtime risk from an incomplete handover: each incident takes 8-24 hours to resolve, with external emergency support billed at roughly NT$2,500-4,000 per hour.
- Opportunity cost: managers are forced to act as human password vaults, burning an extra 5-10 hours per week.
KPI Scorecard for Evaluating a Takeover Vendor
| Dimension | What to look for |
|---|---|
| Tech-stack match | Consistent with your existing systems (e.g. Laravel) |
| Takeover methodology | Can they produce a written takeover SOP |
| Documentation output | Do they commit to delivering a runbook and architecture diagram |
| SLA response time | How many hours to respond to an emergency |
| Confidentiality and contract | NDA and access-return clauses are complete |
| Past takeover cases | At least 2 verifiable takeover engagements |
| Pricing transparency | Health check, retainer, and emergencies priced separately |
| Knowledge-transfer plan | Scheduled interview sessions with the departing engineer |
| Emergency contact channel | Reachable on weekends and at night |
| Exit clause | How documents and accounts are returned at contract end |
Score each item 0-2; below 14 total, find another vendor.
ScriptWalker's Offering and When We Are Not a Fit
We handle this transition under a monthly retainer model:
- System takeover health check: a code plus infrastructure inventory report, from NT$25,000, delivered in 5-7 working days.
- Post-takeover maintenance retainer: NT$15,000-30,000 per month, covering monitoring, fixes, and monthly documentation updates.
We are not the right fit if:
- You cannot produce the source code or account access - no vendor can rescue that; asset ownership must be resolved first.
- You expect a "complete" takeover in 1 week - even 30 days only covers the life-saving layer; anyone promising faster is selling.
- You only want a free assessment to shop around with.
- The tech stack is a complete mismatch (e.g. a large pure-.NET system) - we will tell you straight that it is not our strength.
Day 31-90: From Life-Saving to Stable
Once the life-saving layer is done, spend months 2-3 raising documentation to the point where a newcomer can deploy by following it, and set up monitoring and alerting. At day 90, review three metrics: incident count, average response time, and documentation coverage. Only when all three pass is the system truly taken over.
Decision Checklist: Do You Need to Start the 30-Day Playbook?
- ☐ Are all your systems maintained by only 1-2 people?
- ☐ Are the top-level credentials for servers, domain, and DB on personal accounts?
- ☐ Is there no deployment documentation or runbook at all?
- ☐ Can nobody produce a complete list of scheduled jobs?
- ☐ Are keys and passwords not centrally managed?
- ☐ Could nobody redeploy the system after the leaver is gone?
- ☐ Is the departing engineer the owner of your third-party services?
- ☐ Have you not yet frozen in-progress system changes?
- ☐ Have you not decided between hiring and outsourcing for the takeover?
- ☐ Is there no agreed handover schedule with the leaver?
- ☐ Is there no recorded or written operational walkthrough?
- ☐ Is there no external emergency contact when things break?
If you checked 4 or more, start the playbook before the leaver's last working day.
FAQ
What if the departing engineer will not cooperate?
Start with what does not depend on them: audit external invoices and reclaim access to company-paid services. Then trade a handover-completion bonus for cooperation - more legal and more effective than withholding pay.
Is 30 days really enough?
Enough for the life-saving layer, not for full knowledge transfer. Complete documentation usually takes another 2-3 months, which is why the day 31-90 plan matters just as much.
Should I hire a new engineer first, or find an outsourcing partner first?
Outsource first to stop the bleeding. Recruiting takes 3-6 months and your systems cannot wait; recruit at your own pace during the takeover, then hand documents and access to the new hire.
What do I get from the system takeover health check?
An inventory report: code quality, infrastructure and account lists, risk grading, and a recommended takeover route with pricing - usable for comparison with any vendor.
Next Step
If your only engineer has already resigned, every day of delay means recoverable knowledge slipping away. We offer a free 30-minute consultation to help you decide which step of the playbook to start from:
- Email: [email protected]
- Phone: 0916-224-047
- LINE: @ufv9089p