The English Question
What 136,215 applications reveal about language in offshore hiring. Six in ten applicants describe their English as fluent or better before any filtering has taken place, and the distinction that actually matters is not the one most employers screen for.
Key findings
- Declared English level is one of the strongest predictors of getting hired. A candidate reporting conversational English is placed at roughly a third of the rate of one reporting fluent English. A candidate reporting basic English is more than ten times less likely.
- But native speakers do no better than fluent ones. Self-declared native speakers are placed at 1 in 488. Fluent speakers, 1 in 469. Self-assessment separates the bottom of the range clearly and the top of it not at all.
- Language does not affect who tries, only who succeeds. The share of candidates who complete a practical skills test is 7.4% at every level from conversational upward. Weaker English does not deter people from attempting the process.
- Language demand follows client contact, not seniority. Sales roles draw the highest share of fluent applicants at 63.5%. Finance roles draw the lowest at 48.9%.
- Two thirds of the pool rates 4 or 5 out of 5 on accent. The gap between the pool and those hired is a quarter of a point, far narrower than declared fluency produces.
What Candidates Report
Self-assessed on a four-level scale at the point of application.
| Level | Applicant pool | Among those hired |
|---|---|---|
| Native speaker | 8.6% | 11.3% |
| Fluent | 51.4% | 70.3% |
| Conversational | 36.3% | 17.9% |
| Basic | 3.7% | 0.5% |
| Fluent or native | 60.0% | 81.6% |
The pool figure is worth sitting with. Six in ten people applying for virtual assistant work describe their English as fluent or better, before any filtering has taken place. This is a widely underestimated feature of the Philippine talent market.
“The language proficiency and execution caliber of the Filipino workforce are underappreciated in global commerce.”
How Much It Matters
| Declared level | Placement rate |
|---|---|
| Native speaker | 1 in 488 |
| Fluent | 1 in 469 |
| Conversational | 1 in 1,301 |
| Basic | 1 in 4,994 |
The gap between conversational and fluent is the significant one
A candidate who describes their English as conversational is placed at roughly a third of the rate of one who describes it as fluent. A candidate describing their English as basic is more than ten times less likely again.
The gap between fluent and native is not significant at all
If anything it runs slightly the other way. Whatever “native speaker level” means when someone selects it on a form, it does not translate into a better outcome than “fluent”.
That has a practical implication for employers. The distinction that matters in this market is between conversational and fluent. Above fluent, further language signalling adds nothing measurable, and screening for it filters out capable people for no gain.
Accent
Self-assessed on a five-point scale, and adjusted by our evaluators where assessment shows the self-assessment to be inaccurate.
| Rating | Share of pool |
|---|---|
| 5 | 19.6% |
| 4 | 45.7% |
| 3 | 30.9% |
| 2 | 3.3% |
| 1 | 0.6% |
| Rating 4 or 5 | 65.2% |
Pool average 3.80. Among those hired, 4.05.
Two thirds of the applicant pool rates 4 or 5. The gap between the pool and those hired is real but narrow, a quarter of a point on a five-point scale, which is a much smaller separation than declared fluency produces.
Language Demand Varies by Type of Work
Share of applicants reporting fluent or native English, by the category of role applied for.
| Work category | Fluent or native |
|---|---|
| Sales and business development | 63.5% |
| Marketing and advertising | 62.2% |
| Real estate | 61.2% |
| Administrative and operations support | 60.3% |
| Social media management | 58.5% |
| Writing and content creation | 58.1% |
| Tech and development | 57.9% |
| Analytics and software engineering | 56.7% |
| E-commerce management | 54.8% |
| Design and creative | 53.9% |
| Finance and accounting | 48.9% |
The pattern follows client contact rather than seniority. The roles that draw the strongest English are the ones where the person talks to the client's customers: sales, marketing, real estate. The roles that draw the weakest are the ones defined by a technical skill: finance, design, e-commerce.
Content writing sits low, at 58.1%
That is below administrative work, which is not what most people would predict. It is also consistent with a finding from our companion report on role types, where content writing takes 27 days to fill against a median of 15 across all roles. Two independent measures point the same way: writing is the hardest category to recruit for.
What we looked for and did not find
English level does not affect engagement with the process. 7.4% of candidates submit a completed skills test at native, fluent and conversational level alike. Only the basic group differs, at 5.1%. Weaker English does not deter people from attempting the assessment. It affects the result, not the attempt.
Native declaration produces no measurable advantage over fluent. Placement rates of 1 in 488 and 1 in 469 are effectively identical. This weakens a claim we could easily have made about the top of our talent pool. It is published anyway, because it is what the data shows and because it is genuinely useful advice for employers.
Accent separates far less than declared fluency. A quarter of a point between the pool and those hired, against a threefold difference in placement rate between conversational and fluent.
We report these because a relationship that turns out not to exist is still a result, and because a report that only contains flattering findings reads as marketing.
For journalists
How to cite this data
Attribute figures to “VA Masters, The English Question 2026” and link to this page. Always carry the data period with the figure, for example “60.0% of applicants report fluent or native English, February to August 2026”. Charts and tables may be reproduced with attribution.
Media contact
Interviews, additional data cuts, and methodology questions:
[email protected]
+1 (312) 766-0301
Press room & full dataset
Press coverage & media kit · State of the Filipino VA Industry 2026
Boilerplate
VA Masters is a global recruitment agency founded in 2021 that places rigorously vetted Filipino virtual assistants with businesses across the US, UK, Australia, Canada and Europe. Over 1,000 placements to date.
Company facts
Founded: 2021 · Founder & CEO: Alon Pearl · Placements: 1,000+ · Vetting: 6 stages
Questions & Answers
Common questions about English proficiency in the Filipino virtual assistant workforce.
How good is the English of Filipino virtual assistants?
Better than most buyers assume. 60.0% of applicants describe their English as fluent or native-level before any filtering has taken place, and among those hired the figure is 81.6%. Two thirds of the pool also rate 4 or 5 out of 5 on accent. Figures cover applications received February to August 2026.
Does English level affect whether a candidate gets hired?
Substantially. A candidate reporting conversational English is placed at 1 in 1,301 against 1 in 469 for fluent, roughly a third of the rate. A candidate reporting basic English is placed at 1 in 4,994.
Should employers screen for native-level English?
The data suggests not. Self-declared native speakers are placed at 1 in 488 and fluent speakers at 1 in 469, which is effectively identical. The distinction that matters in this market is between conversational and fluent. Above fluent, further language signalling adds nothing measurable and filters out capable people for no gain.
Which virtual assistant roles need the strongest English?
The ones with client contact. Sales roles draw the highest share of fluent or native applicants at 63.5%, followed by marketing at 62.2% and real estate at 61.2%. Finance and accounting draws the lowest at 48.9%. The pattern follows customer contact rather than seniority.
Do candidates with weaker English apply less often?
No, and they attempt the assessment at the same rate. 7.4% of candidates submit a completed skills test at native, fluent and conversational level alike. Only the basic group differs, at 5.1%. English level affects the result, not the attempt.
How is accent measured in this data?
On a five-point scale, self-assessed by candidates at the point of application and adjusted by our evaluators where assessment shows the self-assessment to be inaccurate. The pool averages 3.80 and those hired average 4.05, a narrower gap than declared fluency produces.
What is this analysis based on?
136,215 applications received between February and August 2026, from the VA Masters applicants database exported 2 September 2026. English level is self-reported on a four-point scale at the point of application. This report describes what candidates declare and what that declaration predicts.
Can I cite these figures?
Yes. Attribute figures to “VA Masters, The English Question 2026”, link back to this page, and carry the data period with the figure. For additional cuts or an interview, contact [email protected].
Methodology & Definitions
Source
The VA Masters applicants database, exported 2 September 2026. 136,215 applications received between February 2026 and August 2026.
Window
February 2026 is the first month in which both language fields reach complete coverage in our system. Earlier records are substantially incomplete, and including them would produce comparisons distorted by our own record keeping rather than by the candidates.
What is being measured
English level and accent are self-reported by candidates at the point of application, on a four-point and five-point scale respectively. Where an evaluator finds during assessment that a self-assessment is inaccurate, the record is corrected. This report describes what candidates declare about themselves and what that declaration predicts. It is not an independent language test, and no claim is made here about the overall accuracy of self-assessment.
Role categories
Applications were joined to roles by exact match on role name, which succeeded for 90.5% of applications in the window. Categories with fewer than 2,000 applications are not reported.
Placement rates
Calculated over applications received in the window that have reached a hiring decision. Applications received late in the window may still be in progress, which affects all levels equally and does not change the comparison between them.
Exclusions
Candidate names, individual applications, roles opened by VA Masters for its own internal team, absolute application counts by category, and any pay or rate expectation data.
About this data
This report describes what passes through VA Masters, not the Philippine talent market as a whole. Every dataset of this kind reflects the business that produced it. Our applicant pool is shaped by where our roles are advertised and by what those roles ask for, and our client base is concentrated in English-speaking markets, which raises the language bar across the board relative to a market serving clients elsewhere.
We think the sample still says something real. 136,215 applications over seven months is among the larger structured datasets on this workforce, and the patterns are consistent across role categories rather than driven by any one of them. But a reader should treat it as a detailed view of one large pipeline, not a measurement of the national talent pool. Where a finding is unusually specific to our client mix, we say so in the section itself rather than leaving it to this note.
Get the full report & data pack
Enter your details below and the complete PDF is yours — every breakdown, the full methodology, and all the figures behind this report, ready to cite.
- The complete State of the Filipino VA Industry 2026 report as a PDF
- All the data behind the charts, cleared for citation with attribution
- Optional: request a custom data cut or an interview with our leadership team
Journalists and researchers welcome. Prefer email? Reach us at [email protected].