# AI shopping assistants and Black Friday

> In PwC Türkiye's Black Friday 2026 survey, the share of Turkish shoppers using AI assistants for pre-purchase research rose from 13% to 20% in a year. These buyers arrive having already compared products and checked the discount. Getting ready means two things: accurate, consistent price, stock, shipping and return data for assistants to read, and customer service that can absorb order and return questions at peak.

URL: https://corrplus.net/blog/ai-shopping-assistants-black-friday/
Updated: 2026-11-03
Corrplus Team

## How much do AI assistants shape buying decisions?

In Türkiye the November sales season is called Efsane Kasım ("Legendary November"), and Black Friday is Efsane Cuma. The Turkish marketplace Hepsiburada opened its 11th Efsane Kasım on 26 October 2026. The two big days are 11 November (11.11, a Wednesday this year) and Black Friday on 27 November. What is new this year is that a growing share of customers ask an AI assistant before they ever reach a store.

Published results of PwC Türkiye's Efsane Cuma Dosyası 2026 survey show:

- The share of Turkish shoppers using AI assistants for pre-purchase research rose from 13% in 2025 to 20% in 2026.
- Among those using or considering AI, the most common goals are comparing product features and reviews (33%), checking whether a discount is real (29%) and finding a product that fits their needs (27%).
- Only 12% are open to letting AI complete a purchase. Of those, 28% would want a preset budget limit and 24% want to approve the final product and price themselves.
- 54% check past prices to judge a discount; just 3% do not check at all.

The online survey ran from 29 September to 9 October 2026 with 10,000 consumers across Türkiye and Europe, 2,000 of them in Türkiye. The figures describe Turkish shoppers, but the pattern is worth noting for any brand selling into the market: AI is not buying on people's behalf, it is sitting in front of the decision. When the question is "which one is better, and is this discount real", the answer is built from the data on your pages and listings.

## Where do assistants get information about your store?

Google began rolling out AI Mode and AI Overviews in Türkiye on 18 February 2026. Google's Search Central guidance says there are no additional requirements to appear in either: a page needs to be indexed and eligible to show with a snippet. The same guidance recommends making sure structured data matches the visible text on the page and keeping Merchant Center and Business Profile information up to date.

On the ChatGPT side, OpenAI has published a specification for merchants to submit a product feed. It says the standard upload currently targets the US, and other markets need OpenAI's confirmation. For a store serving Turkish consumers today, the most reliable ground is accurate public pages and accurate product data in Google's systems.

The broader discipline is [generative engine optimization (GEO)](/glossary/generative-engine-optimization/), and our [GEO checklist](/blog/geo-checklist/) covers it step by step. During a sales event the point that matters most is consistency. If the same product shows one price on the page, another in structured data and a third on a marketplace, you cannot control which one an assistant repeats.

## Which product data fields should you check before the sale?

Google's merchant listing structured data documentation and OpenAI's product feed specification look at the same core fields. The table below sums up what to check.

| Field | Why it matters | Check |
|---|---|---|
| Price and currency | Google's merchant listings need a price above zero; price is a required field in OpenAI's feed | Do the page, structured data and feed all update at the moment the sale price goes live? |
| Availability | Required in OpenAI's feed; an unrecognised value gets the row rejected | Does a sold-out item show as out of stock within hours? |
| Shipping cost and time | Per OpenAI's spec, an empty shipping price means "unknown", not "free" | Are the free-shipping threshold and delivery time stated? |
| Return policy | Google recommends a standard return policy marked up at organisation level | If sale items have different return terms, are they marked at product level? |
| Review count and rating | OpenAI's spec says both should describe the same set of reviews | Does the rating on the page match the rating in the feed? |
| Title and description | The assistant compares products from this text | Are specs, dimensions and compatibility spelled out? |

Google also sets an order of precedence: product-level data in a Merchant Center feed outranks markup on the page. If prices are managed in more than one place during the sale, agree as a team which source is the master.

Remember that shoppers check discounts against past prices. A struck-through "was" price that does not hold up can be spotted by the shopper and by the assistant they use. Check the current rules on advertising sale prices in each market with your legal team.

## How should customer service prepare for the peak?

A shopper who researched with an assistant still asks the usual questions after ordering: where is my order, how do I return this, why did the voucher not work. During Black Friday that volume rises fast.

A [chatbot](/glossary/chatbot/) can take a large share of those questions, provided that:

- **It reads live data.** Order status and tracking come from the order and carrier systems. If the model guesses, you risk [AI hallucination](/glossary/ai-hallucination/).
- **It answers from approved content.** Return windows, campaign terms and shipping thresholds come from one current source, and that source matches the product pages and structured data.
- **Hand-off rules are written down.** Payment problems, damaged items, complaints and explicit requests for a person trigger a hand-off, and the agent sees the full conversation and the bot's summary.
- **There is a capacity plan.** The bot will not resolve everything. Estimate when handed-off conversations will peak and staff shifts accordingly.

If your customers mostly write on WhatsApp, factor in that channel's rules; we cover them in [AI assistants on WhatsApp Business](/blog/whatsapp-business-ai-assistant/).

## What can you still do before Black Friday?

There is not enough time to build a new platform. There is enough time to fix existing data and launch a narrow customer service automation.

1. For your top 50 products, compare price, stock, shipping and return details on the page with structured data and Merchant Center.
2. Ask ChatGPT, Gemini and Google AI Mode the typical questions in your category and record how your store is described.
3. Write down which system updates sale prices, at what time and who owns it.
4. Put your organisation-wide return policy on one page and state any sale exceptions clearly.
5. Pull the ten most frequent support questions from last November.
6. Route the ones that live data can answer to the bot and the rest to people, and write the hand-off rules.
7. Set up a simple dashboard for wait times on handed-off conversations during sale week.

## Where to start

1. Fix the data first: confirm price, stock, shipping and returns match across every channel.
2. Test how assistants describe your store with your own questions, and keep the results.
3. Start customer service automation with one topic, such as order status.
4. After the sale, review which questions went to the bot and which to people, and plan the next peak around that.

This applies most to [e-commerce and retail](/industries/ecommerce-retail/) businesses. To tailor recommendations to each customer's history after the sale, see [e-commerce personalisation](/solutions/ecommerce-personalization/); to plan stock for the next peak, see [demand forecasting](/solutions/demand-forecasting/).

At Corrplus, discovery means reviewing your product data, support logs and current systems; design means writing down the bot's scope, data connections and hand-off rules. Details are on our [AI chatbot development](/services/ai-chatbot-development/), [AI customer service](/services/ai-customer-service/) and, for visibility in AI search, [customer experience and digital marketing](/services/customer-experience-digital-marketing/) pages. If you are not sure where to begin, an [AI discovery sprint](/ai-discovery-sprint/) is a short way to plan the first step together.

## Frequently asked questions

### What do shoppers use AI assistants for before buying?

According to published results of PwC Türkiye's Black Friday 2026 survey, the most common uses are comparing product features and reviews (33%), checking whether a discount is real (29%) and finding a product that fits their needs (27%). Only 12% are open to letting AI complete the purchase for them.

### How do we make sure ChatGPT or Google AI Mode describes our store correctly?

Make the price, stock, shipping and return details on your product pages match your structured data, your Merchant Center data and your marketplace listings. Google recommends that structured data match the visible text on the page. Conflicting data raises the chance that an assistant quotes the wrong price or return window.

### Can any store send a product feed directly to ChatGPT?

OpenAI's product feed specification says the standard upload currently targets the US, and other markets need OpenAI to confirm the integration and its allowed countries and currencies. This may change, so check the current documentation. Without a direct feed, assistants can still read your public pages.

### Which questions should a chatbot handle during peak season?

Repeated questions that can be checked against live systems: order status, delivery tracking, how returns work and campaign terms. Payment problems, damaged items and complaints should go to a person, along with the conversation history.

## Sources

- [Türkiye'de İş Dünyası, report on PwC Türkiye's Efsane Cuma Dosyası 2026](https://turkiyedeisdunyasi.com/efsane-cumada-indirimlere-yakin-takip-142963/)
- [İndigo Dergisi, Efsane Cuma 2026 survey: consumers now verify discounts](https://indigodergisi.com/2026/10/efsane-cuma-2026-pwc/)
- [DHA, Hepsiburada launches its 11th Efsane Kasım](https://www.dha.com.tr/ekonomi/hepsiburada-11inci-efsane-kasim-donemini-baslatti-2969882)
- [Google Türkiye Blog, AI Mode and AI Overviews launch in Türkiye](https://blog.google/intl/tr-tr/urun-duyurulari/turkiyede-ai-modu-ve-ai-bakisi-donemi-basliyor/)
- [Google Search Central, AI features and your website](https://developers.google.com/search/docs/appearance/ai-features)
- [Google Search Central, Merchant listing (Product, Offer) structured data](https://developers.google.com/search/docs/appearance/structured-data/merchant-listing)
- [OpenAI for Developers, Product feed specification](https://developers.openai.com/commerce/specs/feed)