Valentius Kryptix

Setting up your first ChatGPT Ads campaign in India means seven steps: create and verify your Ads Manager account, structure your campaign, write a context hint, build your creative, set up conversion tracking, submit for review, and launch. Most of it is familiar if you’ve run Google or Meta Ads before — the parts that trip people up are account verification timing, writing a context hint instead of a keyword list, and skipping conversion tracking until after launch. This guide walks through all seven in order.

What you’ll need before you start: a registered business (or a legitimate reason to advertise as an individual, where eligible), a GSTIN if you’re billing from India, a payment method, a website or landing page you control, and — if your category touches health, finance, or another regulated vertical — extra time for additional review.

Time required: account creation and verification typically takes under an hour of active work, but verification itself can take a few business days to clear before you can spend. Budget a week from “I want to start” to “my first ad is live,” longer for regulated categories.

Before You Start — What You Need

Beyond the checklist above, it’s worth deciding two things before you open an account at all, because retrofitting them later costs more time than doing them first. First: what page are you actually going to send clicks to? If the honest answer right now is “our homepage,” pause and read the landing page note in Step 4 before you build anything — a generic destination is one of the most common ways a technically correct setup still underperforms. Second: how will you know if this worked? Decide your success metric — cost per lead, cost per signup, whatever matters for your business — before you launch, not after you’re staring at a dashboard full of clicks with no way to tell if they mattered.

If you haven’t already worked through whether this channel is worth testing for your specific business, our ChatGPT Ads readiness checklist is worth five minutes before you continue — it’s a shorter read than this entire setup guide, and it can save you from setting up a campaign for a category that isn’t actually a good fit yet.

Step 1: Create Your Ads Manager Account

One-line takeaway: sign up at Ads Manager with your business details, and expect a verification step before you can spend real money.

Self-serve access runs through OpenAI’s Ads Manager, reachable from ads.openai.com. Account creation itself is quick — the slower part is verification, covered below.

Business Details & Verification

You’ll be asked for standard advertiser information: legal business name, business website, industry or category, and contact details. If your business falls into a category OpenAI treats as restricted — health, finance, and similar regulated verticals — expect this step to involve additional review rather than instant approval. That’s not a sign something’s wrong with your application; it’s the same category-level review process that governs where your ads are allowed to appear at all. If that’s your situation, our guide to restricted category rules covers what that review actually involves before you apply.

For a standard, non-regulated business, verification is typically a matter of days, not weeks — but “typically” is doing real work in that sentence, and it’s worth applying with a week of runway rather than assuming same-day approval, especially the first time you’re setting up an account with a platform that’s still relatively new to the Indian market.

If verification takes longer than expected or comes back with a request for more information, that’s more often a routine request for clarity (a website that doesn’t clearly match your stated business category, for instance) than a rejection — respond with what’s asked and resubmit rather than assuming the application failed outright. One thing worth avoiding regardless: don’t open a second account hoping it’ll clear faster while your first application is still pending. Ad platforms generally treat duplicate applications as a flag worth extra scrutiny, not a shortcut — it tends to slow things down rather than speed them up.

Setting Currency, Timezone & Billing (India-Specific Notes)

Set your billing currency and timezone at account creation, since these affect how your budgets and reporting display later. Treat this screen as effectively permanent rather than something to fix later — like most ad platforms, the country, currency, and timezone chosen when an advertiser account is created tend to be locked in from that point on, since changing them retroactively has downstream tax and reporting implications. If you’re not certain which legal entity or billing configuration you’re setting this account up for, confirm it before you submit rather than after.

Add your GSTIN to your billing profile at this stage rather than after invoices have already been raised — digital advertising in India attracts the standard 18% GST, and getting your tax details right from the start avoids a billing correction later. For the full breakdown of how GST applies and what it means for your actual spend, see our ChatGPT Ads cost in India guide.

Step 2: Build Your Campaign Structure

One-line takeaway: campaigns contain ad groups, ad groups contain ads and context hints — get the hierarchy right before you write a single line of ad copy.

Campaign → Ad Group → Ad Hierarchy

The structure will feel familiar from any other ad platform: a campaign sits at the top and carries your overall budget and objective, ad groups sit underneath it organised around a specific audience or use case, and individual ads live inside each ad group. If you’re advertising more than one product or targeting more than one customer situation, build separate ad groups for each rather than cramming everything into one — a single ad group trying to serve “growing startups” and “enterprise buyers” at once will struggle to write one context hint that’s genuinely relevant to both.

Choosing an Objective (Reach, Clicks, Conversions)

At the campaign level, you’ll choose what you’re optimising for — broadly, reach (paying per thousand impressions), clicks (paying per click), or a conversion-optimised objective once you have tracking data behind it. For a first campaign, a click-based objective is almost always the right starting choice: it gives you both a cost figure and a traffic-quality signal in the same test, which a pure impressions buy doesn’t. We’ve covered the full cost mechanics of each objective — including realistic budget benchmarks for India — in a dedicated guide, so this article won’t repeat that ground; the short version is don’t default to the conversion-optimised objective before you have a few weeks of real conversion data for the system to optimise against.

Step 3: Write Your Context Hints

One-line takeaway: describe the customer’s situation, not a keyword list — this is the single step most advertisers coming from Google Ads get wrong on their first attempt.

Each ad group carries a context hint: a short, written description of the conversations, needs, or situations where your product might genuinely be relevant. A weak one reads like a keyword (“accounting software”). A useful one reads like a sentence describing an actual customer’s situation (“accounting software for early-stage startups that need invoicing, expense tracking, and basic financial reporting as their books get more complicated”). This is a big enough topic that it has its own dedicated guide — our full breakdown of how ChatGPT Ads targeting works includes a worked example, a formula for writing strong context hints, and a side-by-side table of weak versus strong versions. Read that before you write your first one; it’ll save you a week of underperformance while you figure out by trial and error what this guide already explains.

Step 4: Build Your Creative

One-line takeaway: the ad format is small and plain by design — a clear, benefit-led headline matters more than clever copy.

The current ad format is deliberately simple: advertiser name and logo, a title, a short line of copy, a landing page link, and in some placements an image. There’s no room for a long pitch, so the job of your copy is narrow — state the benefit plainly enough that someone mid-conversation immediately understands whether this is relevant to them.

The landing page row deserves more attention than advertisers usually give it. A context hint that describes “growing startups needing invoicing and expense tracking” sets up a specific expectation — if the click lands on a generic homepage that makes the visitor re-explain what they need, you’ve spent budget earning attention and then thrown away the reason it was earned. Build (or at minimum adapt) a landing page for the specific situation your context hint targets before you launch, not after a disappointing first week of numbers.

Don’t stop at one ad per ad group, either. A single creative gives the system very little room to find what actually resonates, and gives you no way to learn anything from the test. Two or three genuinely different angles on the same underlying offer tend to outperform five near-identical rewordings of the same line — for a payroll product, that might mean one ad leading with ease of setup, a second leading with compliance confidence, and a third leading with time saved for the finance team. Different angles can win in different conversations even when they’re selling the exact same thing, and you won’t know which resonates until you’ve actually let more than one compete.

For creative specs specifically — exact character limits, image dimensions, supported formats — check Ads Manager’s current guidance directly when you’re building the ad rather than relying on any article’s numbers, this one included. Format specs are exactly the kind of detail platforms revise as the product matures, and getting them from the source avoids submitting a creative that’s already outdated by the time you launch.

Step 5: Set Up Conversion Tracking Before You Launch

One-line takeaway: do this before you spend a rupee, not after — you can’t retroactively measure a week you didn’t track.

Pixel & Conversions API

OpenAI’s measurement tools support tracking conversions through a pixel and/or a server-side Conversions API, feeding metrics like impressions, clicks, spend, CTR, CPC, CPM, and conversions back into Ads Manager. Install whichever tracking method fits your site before your first ad goes live — retrofitting tracking after a campaign has been running for two weeks means those two weeks are permanently unmeasurable, which defeats the entire point of running a controlled test in the first place.

UTM Parameters

Alongside platform-level tracking, tag every landing page URL with consistent UTM parameters, and make sure whatever analytics tool you already use (GA4 or otherwise) is capturing them correctly. You’ll want your own independent view of what happened, not just Ads Manager’s version of events — every ad platform’s own dashboard has some incentive to look generous, and a second, independent source of truth is the only reliable way to catch that.

Step 6: Submit for Review

One-line takeaway: every ad goes through review before it can serve — build in a buffer day before your intended launch date.

Once your campaign, ad groups, and creative are built, submission triggers a review before anything actually starts serving. Standard categories typically clear this faster than regulated ones, but treat “faster” as relative rather than instant — don’t schedule a campaign to launch the same day as a product announcement or sale without a buffer, since a review delay landing on launch day is an entirely avoidable problem.

If your campaign shows as approved but isn’t actually serving, resist the instinct to immediately rebuild it from scratch. Work through the boring checklist first: is the account fully verified, is the billing profile and payment method actually complete, are the campaign dates and budget correct, and did the ad itself clear creative review. Most “why isn’t this running” situations trace back to one of those four rather than anything more exotic.

Step 7: Launch & Monitor the First 30 Days

One-line takeaway: resist judging results in the first week — most test budgets don’t produce enough volume to mean anything that fast.

Once live, resist the urge to draw conclusions from the first few days. A campaign running near the platform’s ₹725/day minimum produces only one or two clicks daily — nowhere near enough to judge creative, targeting, or conversion performance. Give a properly funded test its full planned window before deciding whether it’s working, and use the first month as a genuine learning period rather than trying to scale or judge it immediately. A useful way to structure that first month:

• Week 1 — Delivery. The only question worth asking is whether the campaign is actually running: are impressions, spend, and clicks showing up at all, consistently. This is a sanity check, not a performance review — don’t draw conclusions from a handful of clicks.
• Week 2 — Relevance. Which ad groups are getting delivery, and which creative angles are getting clicked? You’re starting to see which of your context hints and angles the system considers relevant, even before you have enough data to talk about conversions.
• Week 3 — Conversion quality. If tracking is live (it should be, per Step 5), this is where cost per lead or cost per conversion starts becoming a real number rather than a guess. Compare it against the break-even CPC or acceptable CPA you worked out before launching, per our cost guide.
• Week 4 — Optimisation. Now, and only now, act on what the first three weeks showed: pause the weak creative angles, put more budget behind the ad groups that are actually working, and fix landing pages that got clicks but no conversions.

That sequence — delivery, then relevance, then conversion quality, then optimisation — is deliberately in that order. Trying to optimise in week one, before you have enough signal to know what’s actually working, is how a perfectly good test gets shut down for the wrong reasons.

Putting It Together: A Worked Example

It’s easier to see how the seven steps fit together with an actual example than as a list in isolation. Say you run a payroll software company selling primarily to mid-sized Indian businesses.

After account creation and verification (Step 1), you’d likely build two separate ad groups rather than one (Step 2) — say, one for growing startups scaling their headcount, and a separate one for established companies expanding across multiple states with more complex compliance needs. Cramming both into a single ad group with one context hint would force a compromise that serves neither audience well.

For the startup ad group, a context hint (Step 3) might read: “Payroll software for growing startups automating salary processing for the first time as they scale past a founding team.” For the multi-state ad group: “Payroll software for companies managing compliance and salary processing across multiple Indian states as headcount grows.” Two different situations, two different hints — not two variations of the same keyword.

Each ad group gets its own creative (Step 4): a title and one line of copy matched to that specific situation, sent to a landing page built for that audience rather than a shared homepage. Before either goes live, conversion tracking (Step 5) is installed so you can actually tell, three weeks later, which ad group produced qualified leads and at what cost — not just which one got more clicks. Both submit for review together (Step 6), and once live, both run their full planned test window before you draw any conclusions about which audience, if either, is worth scaling (Step 7).

That’s the shape of a properly built first campaign: not one generic ad group hoping to catch everyone, but a small number of specific ones, each telling a coherent story from context hint through to landing page.

Pre-Launch Checklist

Before you submit for review, a condensed version of everything above, worth running through once as a final check:

• Account fully verified, billing profile complete, GSTIN added
• Country, currency, and timezone confirmed correct (these are effectively locked in once set)
• Campaign objective matches what you’re actually trying to learn (see Step 2)
• Each ad group covers one clear situation, not multiple unrelated ones
• Context hints describe a customer situation, not a keyword list (see Step 3)
• At least two genuinely different creative angles per ad group, not near-duplicates
• Landing page matches the specific situation each ad group targets — not the homepage
• Landing page is publicly reachable and doesn’t block crawlers that need to access it for review or ad delivery
• Pixel and/or Conversions API installed and firing correctly
• UTM parameters added and confirmed showing up in your own analytics
• Success metric (cost per lead, cost per signup, whatever matters) decided and written down before launch

Common Setup Mistakes to Avoid

A few patterns worth watching for, pulled from the mistakes that show up most often across ad platforms generally, not just this one:

• Treating context hints like a keyword list. Piling on variations of your product name doesn’t improve targeting the way it might on Google — see Step 3 and our full targeting guide for why.
• Sending every click to your homepage. Covered in Step 4 — it’s the single most avoidable way to waste a well-targeted click.
• Launching without conversion tracking installed. Covered in Step 5 — a week of untracked spend is a week you can never go back and measure.
• Testing at the platform’s minimum budget and calling it a real test. ₹725/day produces almost no usable data. See our cost guide for what an actual test budget looks like.
• Judging results after three or four days. Covered in Step 7 — give the test its planned window before you decide anything.
• Building one ad group to cover every customer type. Covered in the worked example above — a single generic context hint trying to serve two different audiences tends to serve neither well.

DIY vs. Hiring an Agency

Nothing about this setup process is technically beyond a capable in-house marketer — the account creation, campaign structure, and review flow are all things a self-serve platform is explicitly designed to let anyone do without help. Where an agency’s experience actually earns its cost is less in the mechanical setup and more in the judgment calls: writing a context hint that’s specific enough to work without being so narrow it never delivers, building (or knowing when to build) a dedicated landing page rather than defaulting to the homepage, and reading early results correctly instead of either panicking at day three or ignoring a genuine problem for three weeks.

If your team has the bandwidth to work through this guide properly — including actually building a matched landing page, not just the ad itself — DIY is entirely reasonable for a first test. If you’re weighing that decision against your team’s time, or you’d rather have someone who’s already made the common mistakes on someone else’s budget, that’s exactly the gap an agency fills.

A reasonable way to decide: if you already have in-house experience running Google or Meta Ads, understand your own conversion economics, and have design or dev resources to build a dedicated landing page on short notice, DIY is a genuinely good option — you have most of what this platform requires already. If any of those three is missing, the setup itself will still work, but the campaign that comes out of it is more likely to be a generic one, and a generic ChatGPT Ads campaign is close to the worst version of this channel, since its entire advantage depends on specificity.

Setting this up correctly the first time saves weeks of trial and error. Valentius Kryptix can build and manage your ChatGPT Ads campaign end to end — account setup, context hints, landing pages, and tracking. Book a free setup consultation →

Frequently Asked Questions