CaseForge
← All posts

From the news · Marketing case interviews

ChatGPT Ads hit $1 billion a year in under 200 days. Should your brand spend on it?

A marketing case from the news: OpenAI's ad business is growing fast and just opened self-serve buying in India. How a D2C marketer should decide, with the break-even maths.

By Aditya Jain · 4 November 2026 · 6 min read

Best sunscreen for oily skin,under ₹600?SponsoredYour brand could be hereLabelled, and kept separate from the answer$1Ba year, run rate<200days after launch

What happened

OpenAI says ChatGPT Ads reached a $1 billion annualised revenue run rate in under 200 days after launch. According to the announcement:

  • Tens of thousands of advertisers use it, and it's available in more than 40 countries.
  • Self-serve buying through Ads Manager has opened in India, Europe, the Middle East and North Africa.
  • Ads are labelled, kept separate from ChatGPT's answers, and OpenAI says they don't influence those answers.
  • Most campaigns now pay per click or optimise for outcomes such as purchases.
  • OpenAI gave one example of an online retailer earning 3× its ad spend back over 28 days.

For an Indian marketer, the line that matters is self-serve buying here. Any brand can now try it, and that's exactly the kind of question that turns up in a marketing case round.

The case

You run marketing for a D2C skincare brand in India. ChatGPT Ads just opened self-serve buying here. Your founder asks: should we move some of our budget into it?

1Clarify

  • What's our contribution margin per order after product cost, shipping, returns and discounts? This sets how much we can pay for a sale.
  • Do our customers ask AI tools for product advice? “Best sunscreen for oily skin” is exactly the kind of question people type.
  • Can we track a sale back to the ad? Without conversion tracking, we'd be guessing.
  • How much can we afford to lose on a test?

2Structure

  1. Audience fit: are our buyers on it, asking questions we can answer?
  2. Economics: can a sale from this channel pay for itself?
  3. Measurement: can we prove it, and compare it fairly with our other channels?
  4. Brand risk: how does an ad next to an AI answer reflect on us?

3Analyse

Read the headline number properly

A run rate isn't a year of sales. It's roughly the latest month multiplied by twelve: $1 billion a year is about $83 million a month right now. That's a fast-growing channel, but not proof that it will work for your brand.

Work out the return you need

Return on ad spend (ROAS) is revenue divided by ad spend. To break even, you need ROAS of at least 1 ÷ contribution margin. The lower your margin, the higher the bar:

ROAS needed just to break even, by contribution margin
60% margin1.7×
40% margin2.5×
25% margin4.0×

Break-even ROAS = 1 ÷ contribution margin. Example margins, not real brand data.

At a 40% margin, OpenAI's 3× example clears the 2.5× bar, but only just. It's also one result that OpenAI chose to highlight, not an average.

Break-even on the first order isn't the whole story either. A customer who buys again is worth far more, so compare channels on cost per repeat customer, as in our worked marketing case.

4Recommend

I recommend a small, time-boxed test rather than moving budget: about ₹2 lakh over four weeks, on two or three product questions our customers actually ask.

The audience fit is strong and self-serve buying makes it cheap to try. But the only evidence of returns is the platform's own, and our margin means we need roughly 2.5× ROAS just to break even.

The risk is paying for clicks that don't turn into loyal customers. So I'd set up conversion tracking first, and scale only if cost per customer beats our search campaigns and repeat buying holds up.

Sources