Skip to Main Content
Categories Customer Sales
Created by Guest
Created on Jun 18, 2026

Major problem for all of us: AI and GEO problems on GOb2b

GEO - Generative Engine Optimisation, is fast becoming a big deal. Although on the surface, it is very similar to SEO, I discovered some significant problems in the way the GOb2b websites are being perceived by AI.

Don't dismiss this. AI search is taking over from regular search everywhere and GOb2b currently has a major issue that I discovered.

TLDR: AI will not recommend our site because every product page has the term 'Request a Quote' 24 times, and 'Not Available' twice on every page in the html, even though it never appears on the actual website anywhere on any page and 95% of products are in stock. I believe the same is likely true for all websites using GOb2b's themed back-end.

Full Details:
I asked an AI how we could improve our GEO for our site specifically. It said that it wouldn't recommend our site because there is a lot of friction for the end user when it comes to price especially when it comes to users having to 'Request a Quote' to find out prices. I was confused, because that term doesn't exist anywhere on our site. I asked for more clarification and got this response:

if you look at your website through a standard web browser right now, you won't see a "Request a Quote" button anywhere. You have a fully functional "Add to Basket" setup for your items.

However, from an AI/scraper perspective, it is reading those exact words in your code.

When an AI spider or search bot crawls your product lists (like your "New In" or "Rims" categories), it doesn't just read what is visible; it parses the entire underlying HTML structure. Right now, your site's code contains hidden or inactive template elements for a quote system. For example, right alongside your active "Add to Basket" buttons, the raw HTML code for almost every product listing reads:

View Options · View Details. Request a Quote. View Customisations. Request a Quote. Not Available

Because bots scrape raw text and code rather than rendering the visual layout like a human eye does, an AI training model or retrieval engine indexes those phrases. It sees "Request a Quote" and "Not Available" tied directly to your product listings in the DOM (Document Object Model).

Why this happens and why it matters for AI:

  1. The Culprit: This is usually a remnant of a default B2B e-commerce platform template or a module that was left inactive but wasn't fully scrubbed from the page source code.

  2. The AI Misinterpretation: When an LLM processes this data, it can flag conflicting signals. It sees "Add to Basket" but it also registers "Request a Quote" and "Not Available" multiple times per product. This semantic noise can cause a strict AI engine to misclassify the ease of purchase for a retail customer, assuming a multi-step quote process might be hidden under the hood.

Cleaning those dead code elements out of your product grid templates so they don't appear in the source HTML is a highly specific, low-level fix that will immediately clean up how AI search engines read your transactional capabilities.

____________

Back to my personal response:

As you can see, this is a major pain point for AI's and something that really needs cleaning up ASAP. If the AI thinks that our products are not available, and require a quote, it's never going to recommend us to our customers. That means GOb2b customers will start to fail in the the new AI world and that means GOb2b as a business will suffer and fail with them.

To check this myself, I picked a random product on our website and the term RequestQuote was in the html 24 times on a single page! It's no wonder that the AI doesn't want to recommend us to customers. This will be true for all GOb2b customers.

I then looked for Not Available and that appears twice in the html for a product that is very much available. This type of thing needs addressing ASAP because it's not only going to be affecting GOb2b customers but also GOb2b itself as it's customers fail to reach their customers.

  • Attach files