Scale Black Friday Customer Service with an AI Chatbot

The short answer: the bot doesn't replace the team, it filters the crowd in front of it. Set up properly, it closes repeat questions instantly, and brings the customers who truly need a person to an agent along with the chat history.

When is Black Friday 2026, and when does the rush start?

This year Black Friday is on Friday 27 November 2026 and Cyber Monday on Monday 30 November 2026. In Turkey this period is also called Efsane Kasım ("Legendary November"), Efsane Cuma or Kara Cuma, and campaigns often start in the middle of the month. On the support side the load isn't squeezed into a single day; it spreads over about three weeks, from the pre-announcement to parcel delivery.

So the date your bot needs to be ready is not 27 November. It's the day you send the first campaign message.

What questions come in during the rush?

Sort the messages that arrive in campaign periods into a few groups and the picture gets clearer. Each group suits a bot to a different degree.

Two rows in the table stand out: orders and shipping. These are the most repeated questions of the busy period, but the answer is different for every customer. We don't leave them to the bot to "guess". We cover them separately below.

What information does the bot answer from?

The quality of an AI-powered customer service bot can't exceed the information you give it. The AI in Chatopya answers common questions from the business's own information; business details and site pages are added to this knowledge base. When a question comes in that has no match in the knowledge base, the bot doesn't invent an answer; it hands the chat to an agent.

For Black Friday, this means the real work of preparation is the knowledge base. Before the campaign starts, get these written down:

• Campaign terms: start and end date, which categories it applies to, whether coupons stack, the minimum cart value.
• Shipping and delivery policy: the estimated preparation time during the campaign period, the free shipping threshold.
• Return and exchange rules: the exchange period for discounted products, the return steps.
• Size and product guides: size chart, material, usage information.
• Payment options: whether installments and cash on delivery are available.

A warning. If you write the campaign terms one way on the site and another way in the knowledge base, the bot will tell the old one. If a term changes, update the knowledge base in the same hour.

How current is the stock and price information?

In Chatopya, stock and price information is updated from the connected store at regular intervals; it is not instant. On normal days you don't notice the difference. On Black Friday, though, a size can sell out within minutes.

The practical fix is simple: the bot says "it shows as in stock" and points the customer to the product page or the payment link; it passes a customer who wants firm confirmation to an agent. A promise like "definitely in stock, I've reserved it" should not appear in a bot answer.

Product recommendations and the catalog

In campaign week, customers often don't quite know what they're looking for. Chatopya's AI can ask about the need and recommend a suitable product from the catalog, list products when a category is asked about, send a PDF catalog or an image, and direct the customer to an order or payment link. This is close to what a sales advisor does in a store when they say "how can I help you?" The difference is that it talks to a hundred people at once.

Order and shipping questions: solve them with notifications, not the bot

In the week after Black Friday, the most common question the support team gets is "Where is my parcel?" Answering it with a bot looks tempting. We suggest a different path: keep the question from coming in at all.

In Chatopya, order notifications are triggered from the integration; when the shipping status changes, an automatic WhatsApp notification goes to the customer. Ready-made integrations exist for Shopify, Trendyol and Hepsiburada; other platforms are connected with a custom integration, and the scope is settled in the quote. When customers learn on their own that their order has shipped, without asking, they have little reason to ask the same question.

Someone will still ask. In that case the bot recognizes the question and hands the chat to an agent; the agent looks up the order and answers. Think of the setup like this:

1. Order received → automatic notification.
2. Shipped → automatic notification.
3. If the customer still asks → the bot hands over to an agent, and the agent replies with the chat history in view.

We cover order notifications in detail in the article on WhatsApp ecommerce order tracking.

Returns and exchanges

Explaining the return policy suits a bot. Processing the request itself is the agent's job. In Chatopya's drag-and-drop flow designer you can build a flow for the return scenario: the bot asks for the order number and the reason, gathers the details and hands the chat to an agent. The agent doesn't have to ask everything again.

Handover to a person: when should the bot step back?

What separates a good chatbot from a bad one is when it goes quiet. In the busy period you need to write the handover rules in advance, or else everything lands with an agent or the angry customer argues with the bot.

Our suggestions for rules:

• Sentiment signal: Chatopya's sentiment analysis instantly tags an angry or anxious customer. Take this customer to an agent without making them wait.
• No match in the knowledge base: the bot doesn't invent an answer; it hands over.
• Actions involving money and personal information: returns, payment problems, address changes.
• When the customer asks: don't keep a customer who says "I want to talk to an agent" in the bot.
• Banned words: Chatopya sends a notification to the manager when a banned word appears; assign someone responsible to these chats.

On handover, the chat history goes to the agent. The customer doesn't explain their problem a second time; this alone prevents a good share of complaints.

Working hours

Campaign messages are often sent in the evening, and questions continue until midnight. In the Bot Center you can define working hours and the handover rule. Outside working hours the bot keeps answering common questions; for chats that need an agent, state clearly when the customer will hear back. Saying "we'll get right back to you" and then waiting until morning angers people more than not answering at all.

Meta's rules: which bots are allowed and which aren't?

In the WhatsApp Business Solution terms, Meta banned general-purpose AI assistants: that is, bots where the AI itself is the main product and chats about everything. The rule started to apply to new users on 15 October 2025 and to existing users on 15 January 2026. Business bots for customer support, sales and notifications are outside this ban.

What does this mean for you? Your bot has to stay focused on your business: your products, your campaign, your policies. A customer service bot with a clear subject, grounded in a knowledge base, fits this definition exactly. Stay away from the idea of "a general assistant that answers every question".

The 24-hour window and the service message fee

When a customer writes to you, a 24-hour customer service window opens. In this window you can reply without a template. One thing has changed: since 1 October 2026, these service messages are also charged. For each business number the first 1,000 delivered service messages a month are free, charging starts with the 1,001st, and unused allowance doesn't carry over to the next month.

In Black Friday month this limit is easy to exceed. Every reply given in the window, by the bot or an agent, counts as a message. To work out the budget before the campaign you can use the WhatsApp cost calculator; the details of the rule are in the article on the WhatsApp 24-hour rule. Meta's message fee is separate from the Chatopya plan price and is passed on at Meta's unit price; Chatopya adds no commission. On the Chatopya side, AI bot usage runs on credits, and the credit package is separate.

Team plan: nobody is out of work once the bot is in

When the bot takes the repeat questions, agents are left with fewer but harder chats. A simple plan for campaign week:

1. Build shifts around the campaign message. The two or three hours after the bulk message goes out are the busiest slot. Put the most people in that slot.
2. Make one person the "handover queue" owner. Let that person distribute the chats that drop out of the bot. In Chatopya several agents can use the same number, and chats can be assigned and handed over.
3. Write ready-made reply templates in advance. Have text ready for agents to copy for situations like a shipping delay, a stock sell-out or the campaign ending.
4. Use one inbox. If WhatsApp, Instagram DM, Messenger, Telegram and Web Chat land on the same Conversations screen, nobody runs between channels.
5. Rehearse. Use the pre-launch test in the Bot Center to ask the bot the campaign questions. Catching a wrong answer on 20 November is much cheaper than catching it on 27 November.

For the details of teamwork, see the article on one number, many agents.

What should you measure?

When the campaign ends, saying "it was very busy" isn't enough; you need numbers for next year. Chatopya's analytics dashboard shows response time, resolution rate and satisfaction data; history and a performance record per agent are kept as well. At the end of every conversation a short satisfaction (CSAT) survey goes to the customer.

The indicators we suggest tracking:

• First response time: the time between the customer's first message and the first reply. With the bot on duty, this time should stay low during campaign hours too.
• Resolution rate: how many of the chats were resolved.
• Handover rate and reason: on which questions did the bot hand over to an agent? If the same question repeats often, the knowledge base has a gap.
• CSAT: how does satisfaction compare between bot and agent conversations?

Speak from your own data here. Another company's "we saved this much percent with a bot" figure doesn't describe your customer base. We collected other ways to shorten first response time in the article on how to shorten first response time.

A short pre-campaign checklist

• Campaign terms, return and shipping policy are in the knowledge base, the same as on the site.
• Discount wording follows the 10-day rule.
• Handover rules are written: sentiment signal, unanswerable question, money/personal information, customer request.
• Working hours and the out-of-hours message are defined.
• Order and shipping notifications are on, tested with a test order.
• Ready-made reply templates are with the team.
• The bot has been tried with the campaign questions in the pre-launch test.
• The service message and AI credit budget has been worked out.

If you're curious how to use WhatsApp on the sales side on Black Friday, you can start with the guide to increasing sales with WhatsApp on Black Friday.

Frequently asked questions