Skip to content

How to write a ChatGPT insurance ad that fits the card

Alex Moura Co-Founder and CEO of Digital Star Marketing. Fractional CMO for insurtech and MGA teams, with hands-on experience in paid media and HubSpot attribution. Published · Updated
A copywriter in a lavender sweater writes by hand in a notebook at a café window, a cup beside her, concentrating. Two people trim printed headlines with scissors on a cutting mat, short strips of paper arranged in a column, bright studio.

Write the ad the way your customer talks, with one specific detail, one reason to trust you and one clear next step. Then check it on a small phone, which only shows about 24 characters of the headline and 48 of the description. Test your logo against a picture of the moment, and test sentence-style hints against keyword-style ones. And turn off creative enhancements, the setting that lets AI rewrite your approved copy.

The formula

  1. The asker’s words. Say “first apartment,” not “renters coverage solutions.” Pull the words from the prompts in Chapter 3.
  2. One specific. A price floor, a time, a coverage amount, a state. “From $9.” “2 minutes.” “$250k.”
  3. One trust signal. “Ohio licensed.” “Since 1985.” For a regulated purchase, use the license line. Some states require more, such as a license number on certain ads, so check your state’s rules.
  4. One next step. What happens when they tap. “Quote in 2 minutes.” “Apply online.” “Bound in a day.” Not “learn more”: the desktop card already adds its own Learn more link.
LineHeadlineDescriptionCount
RentersRenters policies from $9Quote in 2 minutes, Ohio licensed24 / 33
Auto, young driverTeen driver quotesTexas rates in 2 minutes, online18 / 32
PetPet insurance in minutesHelps pay vet bills, dogs and cats24 / 34
Term lifeTerm life, no exam$250k from $19 a month, apply online18 / 36
Small businessRestaurant insuranceBound in a day, 20+ states20 / 26
BundleAuto and renters bundleOne bill, licensed Ohio agents23 / 30

Every number in those examples is a placeholder for yours. Do not run a price you have not filed, and put its conditions on the landing page. Claim “no exam” or “bound in a day” only if it is true for the people who will see the ad.

Check the length on a small phone

The headline field accepts 50 characters. The description accepts 100. Ads Manager’s counter suggests 32 and 54, and past 32 it warns: “Copy may be truncated in some placements.” On a small phone (360 pixels wide) the headline cuts near 24 characters and the description near 48. Write to 24 and 48 and the full line shows on every phone today.

Vital Design, which runs campaigns on the platform, says ads kept inside those cuts doubled their click rate.

Many insurers write past the cut. The median insurance headline in the 405-creative sample is 25 characters. The median description is 54. On the smallest phones, about half of them lose the end. OpenAI can change the card at any time, so check the length before each launch. On desktop, the wider card now spans the full answer column, so a 50-character headline fits there. Phones still cut at about 24 and 48, so write to the phone limits.

On an iPhone, headlines up to about 30 characters fit. The cut falls by width, not by character count. Of 15 cards seen on an iPhone, 5 headlines, 6 descriptions and 4 advertiser names were cut. Never start the headline with your brand: the card already shows it, and one warranty ad spent its only headline line on its name. Put the price in the image, as ERGO NEXT does (“from $29/month”): the phone cuts text, never the picture. Use only a filed price a typical applicant can get, with its conditions on the landing page.

Two ad cards: one written to the 50 and 100 limits and cut mid-word, one that fits in 24 and 48; below, three example pairs with their character counts
What a small phone shows today. The counts are ours; the doubling is Vital Design’s claim.

What should the image be?

Three public image tests exist, and they point in different directions.

TestWhat wonHow solid
Workshop Digital, 69,630 impressions, three image typesIndustry-specific image 1.2% click rate, lifestyle photo 1.1%, generic graphic 0.9%Measured, small, no significance test
Vital Design, its own campaignsLogo beat product or service imagery by 14%Their claim, sample not shown
Symphonic Digital, relaying another advertiserA real human headshot instead of a logo, click rate near 5%Secondhand, unverified

What they agree on: the image is tiny, so detail turns to mush and a clear mark reads first. What they do not settle: whether that mark should be your logo, a picture of the moment (a front door, a car, a dog), or a face.

Test your logo against a picture of the moment, with two ads per ad group and a readout on day 14. Use a square image at least 640 by 640 pixels. It shows at about 118 pixels on the card, so skip small text. If you test a price in the image, keep it to one large line, the way ERGO NEXT does. Skip anything that looks like stock.

Multi-image ads, with 4 to 6 square images, are now in open beta for all advertisers. Keep a first test to single-image ads so the two-ad comparison stays clean.

Context hints that match the ad

Hints are free text at the ad group level. OpenAI’s pattern: offer, who it helps, the situation, one fit detail. Their example: “Same-week air-conditioning repair by licensed technicians for homeowners in Dallas.” OpenAI’s developer docs add one rule: put location in targeting, not in the hint. The insurance version:

  • Renters insurance for people moving into a first apartment, starting around $9 a month.
  • Auto insurance quotes for parents adding a 16 to 19 year old driver to their policy.
  • Pet insurance for new dog and cat owners comparing vet bill reimbursement plans.
  • Restaurant and bar insurance for owners opening a first location, general liability and liquor liability, bound within a day.

Workshop Digital’s one test found short keyword-style hints beat sentence-style hints, 1.2% to 0.9% click rate, so test both. On a $25 budget, do it in the second round: one ad group with sentence hints, one with keyword hints such as “renters insurance first apartment,” same ads in each.

Keep one need per ad group, so the ad and the hint say the same thing. Keep the list short; Symphonic relays reports that long, comprehensive hint blocks stopped delivery. Don’t write instructions (“only Chicago,” “not for students”). The hint guides matching and enforces nothing.

The answer after ChatGPT’s quiz

ChatGPT often answers with a short quiz: two or three quick questions with buttons, such as “What monthly budget would you prefer?” When the person presses its button, ChatGPT sends their situation as one detailed message, such as their state, budget, family and what they own.

That answer carries the richest context in the chat, so write some of your context hints in the quiz’s own terms:

  • First apartment, budget $15 to $25 a month, laptop and furniture.
  • Parent with two kids and a mortgage, working out how much term life to buy.
  • Solo landscaper with a personal truck and trailer, buying a first general liability policy.

The landing page

Send the click to a page that answers the question the ad was matched to, with the quote form above the fold, not to the homepage.

Workshop Digital found headlines and descriptions cut off earlier than the stated limits. So the page has to finish what the card started. The license line and the state belong on it. OpenAI reviews the landing page with the ad, and its crawler has to reach it, so no sign-in walls and no blocked bots.

Use ERGO NEXT’s ChatGPT landing page in Chapter 3 as a model. Copy its shape: a dedicated URL, an industry picker, a rating with the review count, a customer count, the same starting price the ad promised with its footnote, and every button going to the quote flow. Then add what it lacks: the issuing carrier, your license line and a short form on the page.

The card shows your account name and no web address. Put that same name at the top of the landing page, or the person thinks they landed in the wrong place. One travel ad’s name appeared nowhere on the site it linked to.

A fast page keeps more of the people who tap. Chapter 6 covers tracking them.

Turn off creative enhancements

Creative enhancements let the platform rewrite your headline and description to match the conversation. OpenAI’s webinar recommends leaving them on.

As of Friday, October 30, 2026, OpenAI turns creative enhancements on for new and existing campaigns unless you have opted out. They can rewrite and translate your text, resize your image and change your landing page address. Enhanced ads “can run automatically without requiring you to review or approve each version.” Turn them off for the account in Ads Manager, under Manage settings. Then check each campaign, because a campaign’s own setting overrides the account default. Opt-outs you set before October 30 carry over.

For state-filed language, rate disclosures, and anything your compliance officer signed, leaving them on is the wrong trade. A changed landing page address can also skip the page that carries your state disclosures. Write two versions yourself and let the auction choose.

Check what ChatGPT already recommends

Before you spend, ask ChatGPT your customers’ questions on a free account and read what it says about your category and your brand. If the answer calls you expensive or names your rival, an ad under that answer pays for a hostile introduction.

Fix the answer first. The six steps are in our guide to getting recommended.

FAQs

Can I put the price in the headline?

If it is a filed, real starting price, yes. 11.1% of insurance creatives mention a price or a savings figure. “From $9” is a specific. “Cheap” is not.

Should the headline be a question?

8.4% of insurance creatives ask one. The prompts are already questions. The card that answers reads better than the card that asks another.

How many ads per ad group?

Two to start, differing in one thing: the image, or the specific. $350 buys about 70 to 117 clicks at $3 to $5 a click, and about 1 to 6 started quotes. That is too few quotes to pick a winning ad, so compare the two on click rate. Add a third, and even click rate cannot tell them apart in 14 days.

Sources, checked October 27, 2026:

Want it run for you? Talk to an insurance marketer who runs ChatGPT ads.

Alex Moura
Written by

Alex Moura

Alex Moura is Co-Founder and CEO of Digital Star Marketing and a fractional CMO for insurtech and MGA teams, with hands-on experience in paid media and HubSpot attribution.

Make your ChatGPT ads count quotes, not clicks.

Tell us what lines you write and where you are licensed. Alex Moura will check your tracking, your quote form and your state plan, then map the first test with you.

Bobby Touran

★★★★★

“Digital Star is hands-on, creative and thoughtful at every step!”

Bobby Touran, CEO and Founder at Rainbow