Why AI Gets Your Product Pricing Wrong

ChatGPT and Perplexity are telling customers the wrong price for your products. Here's why it happens and what ecommerce brands can do about it.

Mersel AI Team

December 28, 2025

8 min read

Ask ChatGPT how much your product costs. There's a good chance the answer is wrong.

Not slightly wrong. Wrong enough to lose you a sale. Wrong enough that a potential customer moves on to a competitor whose pricing the AI got right.

This is already happening at scale. As AI referral traffic to retail grows 4,700% year-over-year and 58% of consumers use AI for product recommendations, pricing accuracy in AI answers is becoming a real revenue problem for ecommerce brands.

Here's why it happens and what you can do about it.

How AI Reads Your Pricing (Badly)

When someone asks ChatGPT "How much does [product] cost?", the AI doesn't visit your site and look at the price tag the way a human would. It reads the raw HTML of your page and tries to extract a number that looks like a price.

That process breaks in five common ways.

1. JavaScript-Rendered Prices

Most modern ecommerce platforms render pricing client-side. The price you see on screen doesn't exist in the initial HTML that AI crawlers read. Instead, the HTML contains an empty container that gets filled by JavaScript after the page loads.

AI crawlers often don't execute JavaScript. They see the empty container. No price.

This is the single most common reason AI gets pricing wrong. Shopify, WooCommerce, and most headless storefronts all have this problem to varying degrees.

2. Dynamic Pricing and Variants

Your product has five sizes and three colors, each at a different price. The page shows "From $29.99" with a dropdown that reveals the full price range. AI sees "From $29.99" and reports that as the price. The $89.99 premium variant? Invisible.

Worse, some stores show different prices based on location, login status, or A/B tests. AI might crawl a version of your page that shows a price no customer will ever actually see.

3. Promotional Pricing Confusion

Your product is on sale. The page shows a crossed-out $79.99 and a current price of $49.99. AI crawlers often extract both numbers without understanding which is current. The result: AI tells customers your product costs $79.99, or it reports a confusing range, or it shows whichever number appears first in the HTML.

Seasonal sales, coupon codes, and bundle discounts make this worse. Every pricing variation is another chance for AI to get confused.

4. Currency and Regional Pricing

You sell internationally. Your site shows USD to American visitors and EUR to European visitors. AI crawlers typically see the default or server-side version. If that's USD but the user asking the AI is in Germany, the answer is technically correct but practically useless.

Some stores show no currency symbol at all in the raw HTML, leaving AI to guess.

5. Missing Schema Markup

Even when the price is in the HTML, AI might not recognize it as a price. The number "49.99" on a page could be a price, a rating, a weight, or a model number. Without structured data (specifically Product and Offer schema markup), AI is guessing.

Only a fraction of ecommerce sites have complete Product schema. Partial schema is almost worse than none because it gives AI enough to make a confident but wrong statement.

Why This Costs You Sales

When AI confidently states the wrong price, three things happen.

Customers don't check. 80% of consumers rely on AI-generated answers for 40% or more of their searches. Most accept the AI's price as fact and make buying decisions based on it. They don't visit your site to verify.

Price comparison goes wrong. When someone asks "Is [your product] cheaper than [competitor]?", AI compares whatever prices it extracted. If yours is wrong but the competitor's is right, you lose the comparison even if you're actually cheaper.

The wrong price scales. Unlike a single bad Google listing, an incorrect price in AI gets repeated across every conversation where your product comes up. ChatGPT alone handles over a billion searches per week. One bad extraction multiplies fast.

And here's the part that stings: ChatGPT referral traffic converts at 15.9% compared to 1.76% for Google organic. The visitors you're losing to pricing errors are the highest-converting traffic on the internet.

How to Fix It

The fix has two parts: making your prices machine-readable, and making sure they're accurate when AI reads them.

Implement Complete Product Schema

This is the highest-impact fix. Add Product and Offer schema markup to every product page with:

price (the current selling price)
priceCurrency (ISO currency code)
availability (InStock, OutOfStock, PreOrder)
priceValidUntil (for sale prices)
highPrice and lowPrice (for variant ranges)

Validate with the Google Rich Results Test. If your schema is incomplete, AI fills in the gaps by guessing from your HTML. That's where errors come from.

Server-Side Render Your Prices

If prices only appear after JavaScript executes, AI crawlers miss them entirely. Make sure the current price is in the raw HTML that gets served on initial page load. Right-click your product page, select "View Page Source", and search for your price. If it's not there, AI can't see it either.

Handle Variants and Sales Explicitly

For products with multiple prices, include AggregateOffer schema with clear lowPrice and highPrice values. For sale prices, use priceValidUntil so AI knows whether a discount is current.

Don't rely on visual strikethroughs or CSS to communicate price changes. The raw data needs to be unambiguous.

Create a Pricing Summary Page

Build a dedicated, crawlable page that lists your products and prices in plain text with structured data. This gives AI a single reliable source for your pricing instead of piecing it together from individual product pages that may have rendering issues.

Monitor What AI Actually Says

Regularly ask ChatGPT, Perplexity, and Gemini about your product pricing. Compare the answers to your actual prices. When you find errors, that tells you exactly where your structured data or rendering has gaps.

The Bigger Picture

Pricing is the most visible symptom of a broader problem: most ecommerce sites are invisible to AI because they were built for human browsers, not AI crawlers.

If AI can't read your prices correctly, it probably can't read your product features, reviews, or inventory status correctly either. The fix for pricing is the same fix for overall AI visibility: structured data, server-side rendering, and clean machine-readable content.

The brands that get this right now, while most competitors are still ignoring AI search, build an advantage that compounds as AI-driven product discovery grows. The ones that don't will keep losing sales they never knew they missed.

Mersel AI fixes this automatically. We make your pricing, product data, and reviews readable by AI crawlers — no code changes required. Book a free AI visibility audit and we'll show you exactly what AI gets wrong about your products today.

Frequently Asked Questions

How do I check if AI is showing the wrong price for my product?

Ask ChatGPT, Perplexity, and Gemini "How much does [your product] cost?" and compare the answers to your actual prices. Do this for your top 5 products. If any prices are wrong, your structured data or rendering has gaps.

Will adding Product schema fix my pricing in AI answers?

In most cases, yes. Product schema with price, priceCurrency, and availability gives AI a clear, unambiguous source for pricing data. Without schema, AI guesses from raw HTML and often gets it wrong.

My prices change frequently. How do I keep AI pricing accurate?

Use priceValidUntil in your Offer schema for sale prices so AI knows whether a discount is current. Server-side render your current price so it's always in the initial HTML. For products with many variants, use AggregateOffer with clear lowPrice and highPrice values.

Does Shopify handle AI pricing automatically?

Not fully. Shopify renders some pricing client-side, which means AI crawlers may not see it. You'll need to verify that prices appear in your page source (View Source, not Inspect Element) and add complete Product schema if it's missing.

Related reading: