Are ChatGPT ads worth it?
For most advertisers, not yet. Here is how to tell, in five minutes, whether your business is one of the exceptions.
ChatGPT ads arrived with plenty of hype and very little data. That has changed. A study of 1,898 US online stores now gives a clear average, and 21 advertisers have published what happened when they tried the channel with their own money.
The short version: paid clicks from ChatGPT turned into orders about a third as often as Google Search clicks, and 8 of the 21 public tests got no sales or leads at all. Yet 7 made money. The difference between those groups wasn't luck. It was arithmetic.
01The arithmetic that decides it
Every ad channel comes down to one question: can a sale pay for the clicks it took to get it? For ChatGPT ads, the most you can afford per click is simple to work out.
Take a store with a $120 average order and a 45% margin. At the study's 0.37% conversion rate, it can afford about $0.20 a click. OpenAI suggests starting bids of $3 to $5. That is fifteen times too expensive, and no amount of clever ad copy closes a gap that size.
02Who it works for, and who should wait
The 7 public tests that made money had something in common: each sale was worth a lot relative to the cost of a click.
It can work if
- Your orders are large, so one sale pays for many clicks.
- Your margin is healthy, leaving more of each order for ads.
- Your product answers questions people already ask ChatGPT.
- You can count real orders in your own store or CRM.
Wait if
- Your break-even is well under the $3 to $5 starting bid.
- You sell low-priced products on thin margins.
- You can't absorb a test that brings zero sales.
- You'd judge it on the ad dashboard alone.
8 of 21public ChatGPT ads tests got zero sales or leads. Most of them were B2B lead generation, often paying $7 to $15 a click.
That last point on the list matters more than it looks. In one public test the dashboard reported 36 conversions, and only 2 real inquiries came in. In another, the dashboard showed no sales when 3 real ones had arrived. Whatever you decide, count orders where you trust the numbers.
03Decide in five minutes
- Find your profit per order. Multiply your average order value by your gross margin.
- Apply a realistic conversion rate. Start with 0.37%, the study average, unless you have your own ChatGPT data.
- Compare with what clicks cost. If clicks cost more than twice your break-even, skip for now.
- If you test anyway, set the rules first. A hard budget cap, a stop rule in clicks, and a judgment based on real orders.
The free calculator does the first three steps for you. For the evidence behind these numbers, see all 21 public tests and the full benchmarks.
Will it pay for itself at your price?
The free calculator turns your order value, margin and cost per click into a break-even number and a verdict. Want it done on your real numbers, with a stop rule? That's the Pre-Flight Report, $34 instead of $49 this launch week.