Why Is ChatGPT Looking Up Prices and Reviews?

What the October study actually examined.

Ahrefs published its analysis on October 12, using 66.6 million free, logged-out ChatGPT responses and their research queries from August and September. It observed increased use of price and review terms in that sample. The model, interface and prompts constrain what the findings establish.

One important distinction: a query used to look for information is not proof that a particular page was cited in the final answer. Read the study as a clue about information needs, rather than a recipe for guaranteed recommendations.

The buying questions are familiar.

A person comparing providers still wants to know the likely cost, what is included, whether the product fits and whether the claims can be checked. An assistant searching on that person’s behalf can encounter the same gaps.

Inspect your own offer before writing another broad AI explainer. If price depends on scope, describe the variables. If a feature has limitations, make them visible. If you mention a customer outcome, link to evidence you are authorized to publish.

Replace an empty price answer with useful conditions.

A fixed public price does not fit every business. But “contact us” without context leaves the buyer unable to understand whether the next step is sensible. Explain the minimum information needed for a quote and what the quote will cover.

The example below is a fictional service page. It provides a concrete answer without inventing a market price or promising that a particular wording will be selected by ChatGPT. The buyer can tell how the scope affects the decision.

Keep reviews and official facts in their proper roles.

Your own product page is an appropriate place for specifications and current terms. Independent feedback can help a reader assess experience, but a review is not a substitute for the current product documentation.

Avoid fabricating reviews, buying undisclosed endorsements or presenting your own comparison as independent research. Explain your comparison criteria, include relevant limits and record the review date. This helps readers judge the evidence even when they disagree with your conclusion.

Turn the observation into a small content review.

Choose one pricing, product or comparison page. List the questions it leaves unanswered, verify the missing facts and publish the most useful correction. Save the date so later traffic or citation observations can be interpreted in context.

Compare repeated questions and sources across engines instead of treating one answer as a ranking. Read our pricing-page guide for worked structures, and citation tracking method for an observation record.

Sources and editorial scope

Announcement and documentation facts are attributed to the primary sources linked below. Availability can change. The visual frameworks and trial methods are Arrow AI editorial guidance; examples are illustrative, not measured product tests or customer results.

• Ahrefs: How ChatGPT Does Query Fan-Out, October 12, 2026