There is no magic number — there is a stop rule
Employers often ask how many maids to interview before deciding. The useful answer is not “always five” or “always twelve.” It is: interview until you have two credible finalists who clear your care, language, and rest-day bar — or until you learn your brief is wrong and must rewrite it. More calls after that usually add noise.
This page is an employer hiring briefing for migrant domestic worker (MDW / FDW) search. General guidance only. Soft MOM: mom.gov.sg. SmartMaid.com.sg is a marketplace for agency-published biodata; licensed EAs place and process Work Permits. Platform ≠ agency.
Neighbouring hubs:
- Biodata first: How to read a maid biodata (unique #15)
- Language check: Checking English or Mandarin communication clearly (unique #56)
- References: Reference checks employers can request (unique #55)
- Stop a bad fit: When to walk away from a mismatch (unique #54)
- This article: how many interviews, and when to decide
A practical range (not a quota)
For most Singapore households:
| Stage | Typical range | Purpose |
|---|---|---|
| Biodata shortlist | 5–15 profiles browsed | Filters, not interviews |
| First interviews (video/phone) | 3–7 helpers | Find 1–2 real finalists |
| Deep second round | 1–2 helpers | Duties, language, spouse join |
| Offer path via EA | 1 | Paperwork + soft MOM / EA process |
If you are past ~8–10 first interviews with zero finalists, stop interviewing and fix the job brief (care load, language, rest day, budget talks via EA — no invented $ rates on this site).
Quality beats volume
Interviewing twenty similar biodata with the same vague brief produces twenty vague maybes. Better:
- Write must-haves (infant nights / elderly / cooking / pets / language)
- Shortlist only profiles that claim those must-haves
- Run the same scorecard every call
- Stop when two clear the bar — or none do
Browsing more on SmartMaid is cheap; unfocused interviews are expensive in time and hope.
Build a one-page scorecard (reuse every call)
Score 1–5 on:
- Duty match (what she actually describes vs biodata)
- Language safety (English / Mandarin check)
- Rest-day / phone boundaries realism
- Attitude to learning your home’s rules
- Gut risk flags (contempt, total yes-machine, major story conflicts)
Add a binary: spouse veto? Yes / No / Need second video.
When to stop early (after 1–2 strong interviews)
You can decide sooner when:
- Care needs are narrow and one helper clearly matches
- Transfer timing is tight and EA confirms process
- Co-employers both watched and agree
Do not invent delays “just to see more people” if your finalist already clears the bar and you are ready for EA next steps.
When to keep going (without doom-scrolling)
Keep interviewing when:
- Finalists fail language or duty echo tests
- You only have “nice chats,” not duty walk-throughs
- Spouses disagree and need one more shared video
- Brief changed mid-search (new baby, new elder)
Cap yourself: set a calendar window (e.g. two weeks of interviews) then decide or rewrite the brief.
Spouse / co-employer rules that cut interview count
- Both watch the same finalist video (or same summary notes)
- One person owns scheduling so you do not double-book chaos
- Agree the walk-away criteria before call three (walk away guide)
- No secret “I’ve already promised the agency” without the other
Misalignment is the #1 reason households burn ten interviews and still cannot choose.
Agency process vs more biodata
Sometimes you do not need more interviews — you need a clearer EA:
- Replacement / guarantee questions answered (replacement policy questions)
- Realistic timeline for transfer vs overseas
- Honest “this profile’s language is fair, not good”
A chaotic EA plus endless interviews is still a mismatch. You may switch agencies on the marketplace and interview fewer, better-supported profiles.
Transfer urgency vs overseas timelines
Transfer helpers can move faster; popular ones will not wait forever. That pushes some employers to decide after 2–3 solid interviews. Overseas / new profiles often allow a slightly longer shortlist — still use a stop rule, not infinite browsing.
After you pick: stop interviewing others cleanly
Once you instruct the EA to proceed with one biodata:
- Pause other interview slots
- Decline waitlisted helpers politely through the EA
- Focus on onboarding prep (house rules, safety, food prefs)
Keeping five “backups” warm with weekly interviews usually signals you have not decided.
Common traps that inflate interview count
- Changing the job description after every “no” without rewriting the brief
- Interviewing profiles outside your filters “just in case”
- Re-interviewing the same helper three times hoping the answer changes
- Using interviews as couple therapy instead of fixing spouse veto rules first
Notes template (30 seconds after each call)
Date / biodata name / EA
Duty score · Language score · Rest-day notes · Spouse view · Next: proceed / second video / decline
Keep notes factual. They feed references and final compare — and later onboarding — without becoming gossip files.
Soft MOM reminder
Interview count is a household choice inside licensed EA hiring. Soft MOM guidance on EAs and Work Passes: mom.gov.sg. Marketplace ≠ agency. Emergency: 995 / 999.
Copyable employer checklist
- [ ] Written must-haves before interview one
- [ ] Same scorecard every call
- [ ] Target: two finalists or a rewritten brief — not endless maybes
- [ ] Language + duty echo tested (#56)
- [ ] Spouse alignment before offer
- [ ] EA process questions clear enough to proceed
- [ ] Decline others cleanly once you choose
- [ ] No invented $; platform ≠ agency
Sample decision scripts
Enough — proceed: “We’ve interviewed enough. Please proceed with [name] on the next EA steps.”
Need one more: “Please arrange one more video with a stronger [infant / Mandarin / elderly] match this week — then we’ll decide.”
Reset brief: “We’re pausing interviews. We’ll send an updated duty/language brief, then shortlist again.”
Bottom line
How many maids to interview before deciding should follow a stop rule: reach two credible finalists (or fix a broken brief), using a shared scorecard and language/duty checks — not an endless tally. Browse on SmartMaid, interview with intent, decide through your licensed EA.
