BPO and Call Centre English Tests

Every cell below comes from the owner's own page, read on 21 October 2026. Where the owner does not publish a figure, the cell says so; we do not fill gaps with numbers from prep sites. Versant appears twice because Pearson currently describes two spoken tests: the Versant English Test sold on its India store, and the Versant by Pearson English Speaking and Listening Test in its 2024 official guide. Your invitation tells you which one you have.

Sources: the Pearson India Versant store (which also says that test "is for practice purposes only"), the Versant Speaking and Listening official guide (2024, PDF), SHL SVAR, Harver Spoken Language Assessment and Berlitz Philippines testing. Two of our older guides go deeper on single tests and stay where they are: the Versant English Test guide walks through each section with samples and a two-week plan, and SVAR test questions and answers has 12 sample questions with model answers. What transfers between them: repeating a sentence exactly, reading aloud at an even pace, and answering in one clear sentence. The fees and results page covers buying, retakes and how scores reach the employer.

Employers rarely say in a job advert which English test they use, and most "company X uses test Y" lists online carry no source. The fastest answer is in your invitation. A Test Identification Number (TIN) and a Pearson link mean Versant; Pearson's 2024 guide says you get the TIN "from your test administrator". An SHL link, or the word SVAR, means SVAR. A Harver link means Harver. A call or link from Berlitz means Berlitz. Below is what we could confirm on official pages on 21 October 2026, labelled honestly.

Philippines

Berlitz Philippines sells testing to employers "in over 40 languages, 24 hours a day, seven days a week, online or over the phone", with a Speaking Skills test it can customise for the employer and a SOPI taken online; it names no clients. Harver names two contact-centre customers on its own page, KellyConnect and IntouchCX, without saying which modules they use. Concentrix's careers FAQ says "Harver is a virtual assessment tool that allows us to objectively evaluate our candidates" and lists language proficiency among the areas tested. It does not say which Harver module tests your English; that it is Harver's Spoken Language Assessment is reported by candidates only. The Berlitz page and the Harver page take each in depth.

India

SHL names customer service and contact centres, technical support, telesales and front-office roles as SVAR uses, but no employers. People search "svar test wipro", "accenture communication assessment" and "tp versant test"; that is demand, not proof, and none of those companies names its English test on a page we could read. The Concentrix FAQ that names Harver is the company's global careers FAQ and does not single out India or any other country.

Three of the four owners say a machine scores the spoken test. Pearson's 2024 guide says the Versant Speaking and Listening test "is autoscored using Pearson's highly trusted, state-of-the-art technology". SHL says SVAR uses "AI-powered scoring" of "pronunciation, fluency, grammar, vocabulary, and comprehension". Harver says its assessment is "validated by internationally recognized spoken language standards, using official IELTS Examiners". Berlitz does not say how its speaking test is marked. What the machines listen for is published, and it is narrower than most applicants think.

Three rules in Pearson's guide apply to every machine-scored test. "If you do not start speaking within a certain time, the test will assume you're not going to answer and move on": that is 6 seconds on repeat and short-answer items and 8 on the conversation and passage questions. You "should not correct yourself". And "you can only move forward", so you cannot re-record. In our judgement, a silence, a restart and a self-correction each cost more than an accent does. The practice bank trains exactly these: every item is timed and says which trait it scores.

Each test reports on a different scale, and none of the four owners publishes a pass mark. Versant reports an overall score from 10 to 90 on Pearson's Global Scale of English, with sub-scores, per the India store page; the 2024 guide adds that score reports are given "against internationally recognized scales (Pearson's Global Scale of English, CEFR)". SHL says SVAR was "developed using globally sourced Common European Framework (CEFR)-aligned data and rated by CEFR experts", but its product page does not show the report scale. Harver reports a percentile with IELTS and CEFR references. Berlitz Philippines publishes no scale and says results are "provided within two business days".

The cut-off is the employer's decision, and no employer we could read publishes it. Forums and prep sites circulate numbers for named BPOs. None shows a source, so we do not repeat them here as facts; the scores page explains what each band means in plain English and labels anything candidates report as exactly that. A fair question for your recruiter is which level the role needs. Two practical points follow. First, two of the four sub-scores on the Versant English Test are Fluency and Pronunciation, so pace and clear sounds carry real weight on the report; a short week is better spent there than on vocabulary lists. Second, a low score on a practice test is information, not a verdict: the store practice test is a fixed form, so a second sitting of it measures memory as much as English.

We expect most readers of this hub to speak Tagalog, Bisaya or another Philippine language, Hindi, or a South Indian language at home. None of the owners publishes results by first language, so this is TalkDrill teaching judgement, based on how those sound systems differ from English, not test data. For Tagalog and Bisaya speakers, the habits that change words are f said as p, v said as b, and short and long vowels merged (ship and sheep; in Bisaya, pin and pen). "Po" in an English sentence is an extra word in a repeat item. For Hindi, Marathi and South Indian speakers, the costly ones are v and w merged, a vowel added before s clusters ("iskool") or after final consonants ("and-u"), equal stress on every syllable, and speaking fast to sound fluent. Bengali speakers may turn s into sh and v into bh.

Every one of these maps to a trait in the table above: a changed word lowers pronunciation and repeat accuracy, an added vowel slows the pace and breaks the rhythm, and a rising tone at the end of a statement shows up in pitch. Harver says its assessment "can also be tailored for use in different regions by identifying words that are especially difficult to pronounce by candidates in a particular language group", so your single-word list may be built around exactly these sounds. The first-language page has a table for each language, a fix and a 60-second drill for every habit, and ten self-test lines to record tonight.

If your invitation came with a deadline, plan backwards from it. Concentrix's careers FAQ warns that "once you open the module, the clock starts ticking". So do not open the link until you have a quiet room, a working headset and twenty uninterrupted minutes. Pearson's guide asks for "a headset with a microphone boom" and a room with no background noise, adding that "even loud fans and air conditioning can cause interference". Then spend the days you have on delivery, in this order.

1. Day 1: record the ten self-test lines on the first-language page and one read-aloud passage. Pick your two worst habits.
2. Days 2 to 5: thirty minutes a day. Ten repeat sentences at full speed, one read-aloud paragraph, one sound drill for your two habits, one open question answered in 40 seconds.
3. Day 6: a full timed mock from the practice bank, recorded, with no restarts.
4. Day 7: light practice, check the headset, sleep.

The English test is one part of hiring, not all of it. Concentrix's FAQ also lists customer support and typing tests, and you will still face an interview, so the last ten minutes of each day belong to the mock call and interview page. The official resources page lists every owner link we verified.

Thirty minutes a day, one action a day, a full timed mock on day 6. Test a week away? Start at week 2. Two weeks away? Start at week 1, which builds the base: your target, your first-language sounds and your first mock calls. Week 3 is for after the machine test, when the interview and the mock call round come. Every day links to the page or the practice mode it uses, and the item codes (r01, p01, o01 and so on) are from our practice bank. The reasons behind each day are on the 7-day plan page.

Each item has four parts. The prompt is the task as you would meet it, with the timing. Where Pearson publishes a timing we quote it; where nobody does, the timing is ours and says so. The counterpart is what the scorer listens for, so you know what to fix before you start. The model is a strong response written by TalkDrill, with a coaching note for the short items. The examiner note names the trait the task scores. Say every item aloud, record it, then compare. For repeat and short-answer items, ask a partner or the TalkDrill AI tutor to say the item once, so you hear it before you see it. None of these items is copied from a real test; they are written in the style of each task.