ChatGPT is charging 4% for Instant Checkout. Now what?

In or out: your choice.

Author

Pietro Saccomani
January 28, 2026

The rate of change with AI right now is blowing my mind.

It seems like every other week there’s some major news dropping.

The latest (in terms of ecommerce) news bomb: OpenAI is charging a 4% fee when customers buy products directly inside ChatGPT using Instant Checkout.

That might not sound dramatic on its own. Four percent isn’t Amazon money. It’s not even TikTok Shop money.

But it matters. 4% is an extra 4% on top of your already tight margins.

And it’s also completely voluntary. It’s not like Amazon or Walmart, where the referral fee is the cost of being on the platform.

The choice is whether to bet 4% of each sale on the value of cutting out a few extra clicks.

So what actually changed?

Perhaps not a change so much as an announcement.

If a customer buys your product directly inside ChatGPT using Instant Checkout, OpenAI takes 4% of the transaction.

Shopify still handles the underlying payment processing as usual, so this is an additional fee; not a replacement.

Your products can still appear in ChatGPT results even if you don’t enable Instant Checkout. And you’ll need to opt in to Instant Checkout (not enabled by default).

The difference:

  • Opt in, and customers can buy without ever leaving the chat.

  • Opt out, and they’ll click through to your site like any other referral.

Compared to existing marketplaces, the economics seem reasonable (on first glance, at least).

Amazon sellers give up 15% in referral fees (and then spend heavily on ads just to stay visible). TikTok Shop and Etsy charge around half of that.

By comparison, 4% looks good.

But that’s if you look at ChatGPT as a marketplace, and compare it to such platforms.

If you’re looking at it as an organic search channel (like Google), the picture is a lot different.

Does ChatGPT even matter?

For all the hype, LLMs still drive a pretty tiny share of ecommerce traffic. For most stores, it barely registers.

That’s what a lot of people jump to when we bring up LLMs/AI: “Why worry about a channel that sends almost no revenue?”

That’s a fair point - for now.

But Google search didn’t become dominant overnight. Social didn’t. Mobile didn’t. Early usage always looks trivial, up until it isn’t.

And I have a hard time believing that LLMs & ChatGPT are not going to become a serious shopping surface.

The potential for improvement on the existing search experience is just too strong.

So I’m going to go out on a limb and say that a lot of people will search and buy products through ChatGPT.

That’s not the biggest question, though.

To Instant Checkout, or not to Instant Checkout

This is the real question.

Not “should we care about LLMs?”

Not “is AI shopping big yet?”

It’s how important it is to let customers complete their purchase inside ChatGPT.

If you opt in, the buying flow stays in the chat. No redirects, no site visit, minimal friction. I think it’s safe to assume higher conversion rates if you opt in.

But the tradeoff is real. You’re giving up 4% of the sale, on top of everything else you already pay. And you’re handing over part of the buying experience to someone else’s interface.

For brands with tight margins, that alone is enough to make this a hard no.

If you don’t opt in, nothing dramatic happens (at least not yet). Your products can still appear in results. Customers can still discover you. They just click through to your site to finish the purchase.

OpenAI have said that they won’t give preferential treatment to products available via Instant Checkout… though the history of the internet may give you reason to be skeptical about that.

But still, for now, the question is whether you think the reduced friction is worth giving 4% of your sales.

The logical answer would be to test it - but with volume still so low, and LLMs like ChatGPT still being a total black box for reporting, it’ll be hard to get an significant data.

So you’re not deciding based on perfect data. You’re deciding based on how you think buying behavior is going to evolve.

My current take (for now)

I’d say, for DTC brands, Instant Checkout probably isn’t worth turning on yet.

Four percent is not trivial. Especially if your margins are already under pressure (as most are), paying a new toll to give up control of the checkout experience is a tough sell.

And right now, it’s hard to say whether the upside is worth it.

It’s worth noting at this point that other LLMs - Gemini, Perplexity, Copilot - don’t charge an additional fee to let customers buy through their platform.

These platforms are all pretty far behind ChatGPT. Though we’re not an ecommerce site, 75% of our LLM traffic comes through ChatGPT (Perplexity 15%, Gemini 5%. We get 10x more traffic from DuckDuckGo than Copilot).

I would imagine the split for brands would be relatively similar.

But the play could be to enable the Instant Checkout equivalents on these sites, and see if there’s any meaningful difference from in-chat checkout.

And with ChatGPT, I get the feeling it’s something brands need to constantly monitor this year.

Today it might not be worth 4% to be part of Instant Checkout.

But once customers get used to buying things with just a couple of taps through ChatGPT, those still sending them through the archaic process of Prompt → Website → Checkout may end up falling behind.

Take control of your customer relationships

If you don’t want to rely on a thousand algorithms to stay in front of your best buyers, you need a mobile app.

It’s a direct connection to your best customer. And finally lets you own the relationship (no middlemen controlling your reach — no more being forgotten).

MobiLoud turns your existing site into an app, with no rebuilding, nearly no work to maintain, and full feature parity with your website.

Quick Hits

ChatGPT Ads Are Not Going to be Cheap

Ads are coming to ChatGPT - and OpenAI is betting they’ll be a hot commodity. A report claims these ad slots will be priced at $60 CPM, which, for reference, is 3x the typical cost of Meta ads.

The kicker? You’re not even going to get the same kind of performance data you get from Meta and Google. Seems like a tough sell, but at the same time it’s hard to argue these slots won’t be valuable.

How AI Shopping Agents Are Reshaping the Buying Journey

This post does a great job of breaking down how ecommerce is changing with AI. We really are in the middle of a massive shift - and it’s worth remembering it’s still relatively early. People may not be comfortable with agents buying on their behalf now, but come 2030? Who knows.

Cracking the Code: Livestreams for the Luxury Market

Livestreaming has the potential to be the next breakout channel for ecommerce (in the West, at least. It’s already huge in Asia).

And it might not be cheap trinkets and beauty products that benefit the most. This article suggests it could be a shot in the arm for the luxury market.

DTC Brands Reaffirm Value of Brick‑and‑Mortar

DTC brands like Mejuri and Beyond Yoga are doubling down on physical retail. The interesting part? 60% of in-store buyers are actually new customers. Stores aren't just serving existing fans, they're driving acquisition as digital CAC keeps climbing.

The Challenge with Viral TikTok Success

TikTok Shop is now forecast to hit $20B+ in sales this year; and brands like Pacsun and Tarte are seeing huge wins (11K jeans sold on Black Friday alone from one influencer post). But now profitability is the new challenge, as TikTok pulls back subsidies and the platform "turns demand planning on its head."

What Bootstrapping to $50M Looks Like

Sean Frank breaks down the math of what it takes to bootstrap an ecom brand to $50M in 6 years - showing that it’s possible, but you’ve got to expect a tough road to get there.

That’s all for now.

I’ll be back in touch next week, with more on how successful brands are doing CX and retention right.

If there’s any topic you’d like to see us dive into, for either the newsletter or the podcast, just shoot me a message here.

Until next time,

Pietro and The Retention Edge Team

PS: want to boost retention, revenue and profitability? If so, launching your own app could be the best move you make this year.

See how: go to our website to get a preview of your app for free, or shoot me a DM on LinkedIn to talk about it.