Remote ≠ Remote: The Skills That Actually Open Global Work to Indonesians
· Loker Dollar Research
From 12,891 remote listings in our corpus: a finding that runs against the usual advice. The highest-paid coding skills (Python $125k) are the most geo-locked for Indonesia-based applicants — only ~1 in 7 are open. The doors that actually open reward leadership, sales, and communication.
Aggregate dataset · Pay by skill/seniority + geo-eligibility, with confidence intervals
remote-geo-gap-indonesia-2026-07.csv (free, CC BY 4.0)Columns: section, metric, n, value, unit, ci_lo, ci_hi. Aggregate figures from our own corpus only — no per-row listings, company names, or individual salaries.
Key Figures
- Only 5.7% of listings explicitly welcome global applicants; of the listings that state an eligibility rule, 66.2% lock out Indonesia-based applicants (95% CI 64.2–68.2).
- Median remote pay ~$78k/yr — roughly 17.2× the Indonesian local median.
- The high-paid coding skills are the most locked: Python $125k (14% open), React $76k. The open skills: leadership $135k (53% open), sales, negotiation, communication.
- Remote is not an entry-level market: only 9.9% entry/junior.
The Headline: The Open Door Isn't the Coding Door
The advice Indonesian job-seekers hear most is “learn to code for remote work.” The data says otherwise. We pair each skill's median pay with how often listings requiring that skill are actually open to Indonesia-based applicants. The result splits into two clear groups.
High pay & OPEN (people/business skills)
High pay but LOCKED (coding skills)
Embed this chart
Copy the snippet below to embed this chart on your site (no JS, no tracking). The source link is baked in.
| Skill | Median pay | % open to ID | n |
|---|---|---|---|
| leadership | $135k | 53% | 32 |
| sales | $90k | 48% | 60 |
| negotiation | $82k | 46% | 31 |
| customer service | $73k | 48% | 30 |
| communication | $73k | 46% | 185 |
| python | $125k | 14% | 41 |
| node.js | $84k | 14% | 25 |
| javascript | $78k | 14% | 33 |
| react | $76k | 10% | 34 |
The takeaway: for many Indonesians the realistic remote path isn't becoming an engineer at a US company (that door is ~90% locked by work-authorization rules), but roles that reward leadership, sales, and communication — which also pay well.
The Skill Pay Ladder
Remote skills span nearly a 5× pay range: from leadership/Python/AI ($135k) down to social media & copywriting ($27k).
Embed this chart
Copy the snippet below to embed this chart on your site (no JS, no tracking). The source link is baked in.
| Lowest-paid skills | Median pay | n |
|---|---|---|
| social media management | $27k | 25 |
| tiktok | $33k | 33 |
| copywriting | $36k | 26 |
| adobe premiere pro | $38k | 32 |
| $40k | 28 | |
| meta ads | $52k | 27 |
| motion graphics | $54k | 38 |
| storytelling | $54k | 31 |
Seniority & the AI Premium
Embed this chart
Copy the snippet below to embed this chart on your site (no JS, no tracking). The source link is baked in.
The curve rises steeply: entry $62k → lead $140k. And the “AI-skill” premium? Smaller than the hype: $84k with an AI skill vs $78k without — about 8% more. AI is not an automatic golden ticket.
Want the list of remote jobs that are actually open to Indonesia?
Create a free account and our Hunt assistant filters the openings that match your skills — then helps you apply automatically.
Start free →Methodology
Population & snapshot date
A single cross-sectional snapshot of our production corpus on 2026-07-06: 12,891 remote listings (5,543 active). Not a random sample of the entire global remote market — it reflects the boards we ingest. No time-trend claims (ingest volume rose from a pipeline ramp, not a market signal).
Geo-eligibility classification (validated)
For each listing, a deterministic, reproducible classifier decides whether an Indonesia-based applicant is eligible: "restricted" when an explicit location/work-authorization rule excludes them (e.g. "US only", "authorized to work in the US", "Remote in Europe"); "id_friendly" when a strong global-openness signal is present (worldwide/anywhere/APAC/Indonesia); otherwise "unknown". Because the classifier is strict, ~83.2% of listings state no signal — exclusion rates are computed over the subset that states one (2,165 listings), and the whole-corpus share is reported too.
Classifier accuracy (measured on a 120-row hand-labeled gold set):
| Class | Precision | Recall |
|---|---|---|
| restricted | 90.9% | 87% |
| id_friendly | 100% | 100% |
The gold set is stratified from prior labels, so precision/recall describe the classifier's behaviour on that set. A pre-registered audit found the prior geo labels non-reproducible; this classifier is the corrected re-classification.
Salary: definitions & rules
Salary is annualized to USD (yearly ×1, monthly ×12, hourly ×2080, weekly ×52); the fixed period and null currency are excluded; plausibility band $5k–$600k; value = midpoint of min/max. Median + IQR always, never the mean; per-cell n≥5. Only ~26% of listings publish pay, and disclosure is not random — the salary figures describe disclosing listings.
Local comparison
The Indonesian local median is computed from IDR-salaried listings in our market corpus (Glints/SejutaCita, n=487): ~Rp6.2M/month (~$5k/yr). A local-vs-global skills comparison is NOT reported: local skill tags contain non-skill requirements (age, vehicle) and a different extraction method, so a direct comparison would be invalid.
Statistics
Every proportion uses a Wilson 95% confidence interval. Cells with n<5 are suppressed; skills are reported at n≥25 (pay) or n≥20 (geo-openness). Observational — no causal claims.
Limitations
- Signal coverage: ~83.2% of listings state no geo rule; exclusion rates are over those that do.
- Listing ≠ hire. This is a demand signal, not a hiring outcome.
- Salary disclosure bias: only some listings publish pay.
- Cross-sectional: one snapshot; no longitudinal claims.
Questions or corrections: contact us. Every figure regenerates from the aggregate CSV above.