Hiring: Content Writing Intern · Remote (India) · Paid

We are looking for
one exceptional writer.

Baweja Media is taking on a single content writing intern. You will write ad copy, video scripts, long-form articles, landing pages and email for brands across the UK, US, India, Canada and Australia — work that is measured in revenue, not in views. This is a paid, full-time, remote internship. The application below is harder than the job interview at most agencies. That is deliberate.

Sit the Entrance Test
$125M+
Ad spend behind the copy you’d write
18
Questions. No calculator. No second attempt.
1
Seat. We would rather leave it empty than fill it badly.
Paid
Full-time · Remote (India)
The Work

Writing that has to
do a job.

There is no content calendar to fill here for the sake of filling it. Every piece you write exists to move a number someone is watching.

Ad copy and video scripts

Primary text, headlines, hooks and full scripts for Meta ads running live spend. You will see the click-through rate on what you wrote, usually within forty-eight hours.

Long-form and SEO

Articles and landing pages that have to rank and then convert. Research first, structure second, sentences third.

Social and carousels

Instagram and LinkedIn copy for the agency and for client brands. Short form is harder, not easier.

Email and newsletters

Sequences and broadcasts where a subject line is worth more than a paragraph.

You will be edited hard, in writing, with reasons. If that sounds unpleasant rather than useful, this is the wrong role.
The Bar

What we are actually
screening for.

We are not screening for someone who can produce words on demand. Software does that now, badly, and at scale. We are screening for the things it cannot do.

Comprehension under load

Read something dense once and extract what it actually claims — not what it seems to claim. Section I of the test is exactly this.

Reasoning that holds

Spot the assumption an argument is resting on, and the evidence that would break it. Six questions, and every wrong answer is one a careless reader would pick.

Numeracy

A writer who cannot tell the difference between a ten-point drop and a twenty-five percent drop will eventually publish something false. Four questions, no calculator.

Command of English

Grammar you could defend to someone holding a copy of Wren & Martin, and the judgement to know when precision beats elegance.

Storytelling

The ability to make a stranger care in two sentences. We ask you to demonstrate this, not describe it.

Full commitment

7+ hours a day, every working day, through India business hours. Not alongside a job, a degree in progress, or another internship.

On AI. You may use it; most people do. What you may not do is send us its output and call it yours. Every written answer below is timed, and pasted text is recorded. If a sample reads like a model wrote it, the application is closed — not marked down, closed. We would rather read four honest paragraphs with a flaw in them than four flawless ones you did not write.
How It Goes

Four steps.
Most stop at the second.

Apply

Roughly twenty-five minutes. Two written answers and an eighteen-question entrance test.

We read it

The test is scored automatically. The writing is read by a human, properly, every time.

A live brief

Shortlisted candidates get a short paid brief with a deadline. Written work, on the clock, on a real topic.

Offer & start

Stipend discussed on the call. Then your first live piece, in week one.

The Entrance Test

Three pages.
Do not rush the second.

Page one is the basics. Page two is where you write — that is the part a human reads hardest. Page three is eighteen questions: reading comprehension, critical reasoning, data interpretation and grammar. There is no way to bluff it, which is the point.

Step 1 of 3 · The Basics
Step 2 of 3 · The Writing
Both boxes below are timed, and anything pasted into them is recorded. Neither has a word limit. Write until you have said the thing, then stop.
Step 3 of 3 · Eighteen Questions
Elapsed00:00
Answered0 / 18
Section I

Reading Comprehension

Read the passage once. Then answer without scrolling back if you can.

Economists have long distinguished between what consumers say they value and what their purchases reveal they value. The doctrine of revealed preference, formalised by Samuelson in 1938, holds that choice data are epistemically superior to survey data: a shopper who reports caring about sustainability but consistently buys the cheaper, less sustainable option has, on this view, told us what she actually prefers. The appeal of the doctrine is its parsimony — it requires no access to interior mental states, only to receipts.

Yet the inference is weaker than it first appears. Revealed preference identifies what a consumer chose given a particular choice set, at a particular price, in a particular moment of haste or depletion. It does not establish that the same ranking would survive a different context. A shopper constrained by budget may buy the cheaper item while genuinely preferring the other; the choice reveals the constraint at least as much as the preference. Critics press a further objection: the doctrine risks circularity, since preference is inferred from choice and choice is then explained by preference.

The practical upshot for anyone researching a market is not that surveys should be trusted over behaviour. It is that neither instrument interprets itself. Stated preference tells us what a person believes about themselves — which is a fact worth having in its own right, since it predicts what they will defend in public, if not what they will buy in private.

1. The passage is primarily concerned with

2. The example of the budget-constrained shopper functions in the argument to

3. As described in the passage, the objection of circularity is that

4. The author would most likely agree that stated-preference data are

Section II

Critical Reasoning

Each argument has exactly one answer that does the job asked of it.

5. A publisher finds that articles whose headlines contain a numeral receive 30% more clicks than articles whose headlines do not. It concludes that adding a numeral to a headline causes more clicks. Which of the following, if true, most weakens the conclusion?

6. A brand halved its advertising budget and revenue stayed flat over the following quarter. The marketing director concluded the advertising had never been working. The argument depends on assuming that

7. Every one of our strongest writers reads fiction daily. Therefore, if we hire only people who read fiction daily, the standard of our writing will rise. The reasoning is most vulnerable to the criticism that it

8. Every applicant who scored above 15 was invited to interview. No applicant who submitted after the deadline was invited to interview. If both statements are true, which of the following must also be true?

9. A newsletter moved from weekly to daily sending. The open rate per email fell from 40% to 25%, and the editor now wants to revert. Which of the following would be most useful to establish before deciding?

10. A company removed all customer testimonials from a landing page, and the conversion rate rose. Which of the following, if true, best explains that result?

Section III

Data Interpretation

No calculator needed. All four are doable in your head.

11. A campaign spends ₹80,000 and produces 3,200 landing page views and 64 completed applications. What is the cost per completed application?

12. Of those 64 applications, 25% are shortlisted. One in four shortlisted candidates is hired. How many hires does the campaign produce?

13. An email open rate falls from 40% to 30%. By what percentage has the open rate fallen?

14. Article A is read by 10,000 people, 2% of whom click through. Article B is read by 40,000 people, 7% of whom click through. What is the click-through rate across both articles combined?

Section IV

Language & Precision

Grammar first, then the harder question of what precision actually means.

15. Which sentence is free of a misplaced or dangling modifier?

16. Neither the editor nor the writers ______ satisfied with the draft.

17. The candidate ______ we interviewed on Tuesday has withdrawn her application.

18. Which revision of “The campaign did really well and got a lot of good numbers” is the most precise?

Please complete the highlighted fields before continuing.

Your answers go directly to the hiring team. No spam, ever.

✓

Test submitted.

Your two written answers are now in front of a person, not a filter. If you are shortlisted, you will hear from us within 48 hours with a short paid brief and a deadline.

However it goes — that was a genuinely hard test, and you finished it.