The Short Answer

Instagram's own robots.txt file shuts out every crawler it does not name, and that includes ChatGPT's search crawler, so a brand that exists only on Instagram gives ChatGPT's search nothing it may read. As of 8 November 2026, OpenAI's crawler page says sites closed to that crawler "will not be shown in ChatGPT search answers". Keep Instagram as your shop window and put the facts on pages you own: one page per best seller with the item named plainly, the price in Naira, sizes, fabric, delivery areas and how to order, plus answers to the questions customers send you in direct messages.

Short on time? Ask an AI to summarise this post:

A woman in Abuja asks ChatGPT where she can buy a ready-to-wear adire dress from a Lagos designer who delivers within a week. ChatGPT gives her a few names. Your brand has a busy Instagram page, good photos and happy customers in the comments, and it is not one of the names.

For a brand that sells only through Instagram, that result is likely, and the reason is written in a public file on Instagram's own website. This post explains the file, what it means for ChatGPT, and what a fashion brand in Nigeria can do about it without leaving Instagram.

The buyers are there. According to DHL's E-Commerce Trends Report 2026, 46% of Nigerian shoppers use AI chat tools when shopping online, fourth of the 29 markets DHL surveyed.

Why can ChatGPT not see my Instagram page?

Because Instagram tells ChatGPT's search crawler to stay out. Every website can publish a file called robots.txt that tells crawlers which parts of the site they may read. As of 8 November 2026, Instagram's file names a long list of crawlers one by one and ends with a rule for every crawler it has not named: "User-agent: *" followed by "Disallow: /". That rule closes the whole site. OpenAI's search crawler, OAI-SearchBot, is not on the named list, so the closing rule covers it. OpenAI's training crawler, GPTBot, and Perplexity's crawler, PerplexityBot, are named, and each of them gets "Disallow: /" too.

Folded lengths of bright ankara print fabric stacked on a market stall, in purple, green, blue and orange patterns, the kind of stock a fashion brand photographs for Instagram
Photo by dodmichaela on Pixabay

OpenAI's own crawler page says what that means for ChatGPT. It describes OAI-SearchBot as the crawler "used to surface websites in search results in ChatGPT's search features", and it says sites opted out of it "will not be shown in ChatGPT search answers", though they can still appear as navigational links. Perplexity's crawler page says PerplexityBot is the crawler that surfaces and links websites in Perplexity's results, and Instagram shuts that one out by name.

The comments at the top of Instagram's file are plain about the reason: collecting data from Instagram by automated means is "prohibited unless you have express written permission". That is Meta's decision about its own platform. For a brand that lives there, the result is that your captions, your price replies and the delivery details in your highlights sit behind a door ChatGPT's search does not open.

Two honest limits. A famous brand can still be known to ChatGPT through other sites that write about it, such as news articles, blogs and directories, so a big name may come up with no website at all. A small brand rarely has those. And the file can change. We read it on 8 November 2026, and anyone can read it today at instagram.com/robots.txt.

Does Google see my Instagram page?

Partly. Instagram's file treats Google's crawler differently: Googlebot is named and allowed into most of the site, with some sections closed, such as comments and direct messages. So an Instagram profile can turn up in Google Search. Google's documentation for its AI features says a page has to be indexed and eligible to show a snippet before it can appear in an AI Overview or in AI Mode, with no special optimisation needed beyond that. Whether a given Instagram post meets that bar is Google's call.

The second problem applies to every engine. A grid of photos gives a machine very little text to quote, and OpenAI's crawler page says nothing about reading what is inside a picture. The buyer types "white linen kaftan for men, size XL, delivery to Port Harcourt". The answer can only match those words to a page where those words are written down.

What does ChatGPT read instead?

When ChatGPT searches, it reads pages its crawler is allowed to visit and builds the answer from them. For a fashion brand that means your own website if you have one, plus any page that writes about you and lets the crawler in: a press feature, a fashion blog, a directory listing, or a marketplace page if that marketplace allows the crawler. If none of those exist, there is nothing about you for the answer to use, and the names it gives belong to brands that have them.

What should a fashion brand's website say?

Your website does not have to be big. It has to say in words what your Instagram says in pictures. Six steps, in order.

1. Ask ChatGPT what your customer asks

Open ChatGPT and type the question a real buyer would type, with the item, the size or occasion, and the city. Leave your brand name out. "Where can I buy a ready-to-wear aso ebi dress in Lagos that delivers to Abuja?" is a useful question. Write down who is named and where each link points. Do the same in Perplexity and Gemini if you have ten more minutes.

When this step is done you have a dated baseline, and you know which brands you are up against.

2. Get one page live on your own address

If you have no website, start with one page on your own domain. Say your brand name, what you make, where you are based and how to order, and type your WhatsApp number out as text so a machine can read it. Ask whoever builds it to confirm that the site's robots.txt file allows OAI-SearchBot, PerplexityBot and Googlebot. OpenAI's crawler page says a change to that file can take about 24 hours to reach its systems.

When this step is done, there is a page about your brand that ChatGPT's search is allowed to read.

3. Write a page for each best seller

Give every item a customer asks about its own page, written in the words a buyer would type. Name the item plainly ("Men's linen kaftan, white") instead of a collection codename. Then write down:

  1. The price in Naira.
  2. The sizes, and whether they run small or large.
  3. The fabric and the colours it comes in.
  4. Whether it is ready-to-wear or made to order, and how long made-to-order takes.
  5. Where you deliver and how long delivery takes.
  6. How to order, with the WhatsApp number written out.

Put the photos beside the words. The photos sell to the person; the words answer the machine.

When this step is done, a question like "white linen kaftan Lagos delivery" has a page on your site whose text matches it.

4. Answer the questions customers send you in direct messages

Read back through a month of direct messages and WhatsApp chats. The same questions come up again and again: do you deliver to Abuja, can you make it in two weeks, do you do alterations, what happens if it does not fit. Put each one on your site with a two-sentence answer in your own words. People now type these same questions into ChatGPT, and a public page that answers them is the page that can be quoted.

When this step is done, the questions in your inbox have public answers with your brand name on them.

5. Make every listing say the same thing

Your Instagram bio, your website, your Google Business Profile if you have a showroom or a pickup point, and any marketplace listing should carry the same brand name, the same phone number and the same one-line description of what you make. When two sources disagree, an engine has to pick one or hedge.

When this step is done, every place that describes you describes the same brand.

6. Get written about where the crawler is allowed

A page about yourself is your own claim. A fashion blog that features a collection, a local press piece, a stylist who links to your site, and reviews on your Google profile are other people confirming it, on pages the engines can read. Do not buy fake reviews or placed articles. Google's guide to its AI features says seeking inauthentic mentions across the web "isn't as helpful as it might seem".

When this step is done, a search for your brand name finds more than your own site.

Do I have to leave Instagram?

No. Instagram is where your customers see the clothes, and many of them will keep buying there. Keep posting. What moves is where the facts live. The price, the sizes, the delivery terms and the answers to common questions go on pages you own, and the Instagram bio links to them. Instagram stays your shop window, and the website is the catalogue an engine is allowed to read.

How long before ChatGPT names my brand?

Nobody can give you a date, and anyone who does is guessing. OpenAI, Google and Perplexity publish no figure for how fast a new page is picked up and used in answers. The one documented delay is OpenAI's, about a day for a robots.txt change to take effect. Run the question from step one on the same day each month and write down who is named. That record is your measurement.

Buyers are already asking these engines:

46%
Nigerian shoppers who use AI chat tools when shopping online, fourth of 29 markets
DHL, E-Commerce Trends Report 2026
49%
US adults who use AI chatbots, up from 33% in 2024
Pew Research Center, Americans and AI 2026
45%
US consumers who used AI for local business recommendations in the past year, up from 6%
BrightLocal, Local Consumer Review Survey 2026

A buyer who asks ChatGPT for a dress is asking a machine that reads text. Give it some to read, on a page it is allowed to open.

Want to see what the machines tell your customers?

Our free AI Visibility checklist covers the 12 checks behind the steps above, starting with what ChatGPT, Gemini, and Google actually say when someone asks about a business like yours. Enter your email and we will send it over.

✓ Check your inbox, your checklist is on the way!

We respect your privacy. Unsubscribe at any time.

Key Takeaway

ChatGPT's search is not allowed to read Instagram, so the facts that sell your clothes have to live on a page it may read. Keep posting on Instagram, and give every best seller a page on your own site with the item, the price, the sizes and the delivery terms written out.

One forward before you go: open instagram.com/robots.txt in your browser and scroll to the last two lines. Then ask ChatGPT the question your best customer would ask, without your brand name, and write down who it names. That list is who you are competing with.

Frequently asked questions

Can ChatGPT read my Instagram captions?

Not through its search. As of 8 November 2026, Instagram's robots.txt file closes the whole site to every crawler it does not name, and OpenAI's search crawler, OAI-SearchBot, is not named. OpenAI's crawler page says sites opted out of that crawler will not be shown in ChatGPT search answers, though they can still appear as navigational links. A well-known brand may still come up because other sites write about it.

Is a marketplace listing on Jumia enough?

Keep it if it sells, and do not rely on it alone. A marketplace page describes the product inside the marketplace's layout and rules, and it rarely says who your brand is, where you are based, or how long made-to-order takes. One page on your own domain carries those facts under your own name, and you control what it says.

Do I need a big website?

No. Start with one page that says who you are, what you make, where you are based and how to order, then add a page for each best seller. A few honest pages with the facts written out serve a buyer and a crawler better than a large site full of photos and no words.

Is this the same as SEO?

The work overlaps with good SEO, and a developer who says so is right. The difference is the measurement: SEO tracks rankings and clicks on Google, while AI visibility tracks whether your brand is named when a buyer asks ChatGPT, Gemini or Perplexity. Neither is possible for an Instagram-only brand until there is a page the crawlers are allowed to read.

Do I need product markup or schema on the pages?

Write the words on the page first. Schema.org defines a Product type with properties such as name, image, brand, offers, colour, material and size, and a developer can add it later so the markup matches what the page shows. Google's guide to its AI features says structured data isn't required for generative AI search, so the visible text is the part that cannot be skipped.

Should I block AI crawlers on my own website?

Never block the search crawlers if you want to be named. OpenAI's crawler page says each setting is independent, so a site can allow OAI-SearchBot for ChatGPT search and disallow GPTBot, which collects content for training. Make the training decision separately, and check your robots.txt file allows OAI-SearchBot, PerplexityBot and Googlebot.

References & sources

  1. Instagram (Meta), instagram.com robots.txt (read 8 November 2026: 'User-agent: *' with 'Disallow: /'; GPTBot and PerplexityBot disallowed by name; OAI-SearchBot not named; Googlebot allowed with some paths closed)
  2. OpenAI, Overview of OpenAI crawlers (OAI-SearchBot, OAI-AdsBot, GPTBot, ChatGPT-User) (read 8 November 2026)
  3. Perplexity, Perplexity crawlers (PerplexityBot, Perplexity-User) (read 8 November 2026)
  4. Google Search Central, AI features and your website (indexed and eligible for a snippet, no special optimisation)
  5. Google Search Central, Guide to optimizing for generative AI features on Google Search (quotes re-checked 8 November 2026)
  6. Schema.org, Product (read 8 November 2026)
  7. DHL, E-Commerce Trends Report 2026 (29,000 shoppers across 29 markets)
  8. Pew Research Center, Americans and AI 2026 (5,119 US adults, February 2026)
  9. BrightLocal, Local Consumer Review Survey 2026 (1,002 US adults)
A
Abdulmuiz Adebowale
AI Product Manager & Builder, Nexum-sol

Abdulmuiz helps Nigerian businesses build websites and digital systems that get found, by people and by AI. He writes about practical SEO, AI search, and growth systems for SMEs at Nexum-sol.