For years the deal was simple: rank on Google, get the click, earn the commission. AI answers have changed that. Google shows AI Overviews and AI Mode for many questions, ChatGPT and Perplexity answer them in a chat, and a growing share of searchers get what they need without visiting any site. For trading and crypto affiliates, who rely heavily on informational traffic, this is the biggest shift in search since mobile.
The good news: AI answers still need sources, and they link to them. This article explains which traffic is most at risk, how AI systems choose what to cite, and what you can change on your site now.
Which traffic is most at risk
Not every query is affected the same way. As a rule of thumb:
- Hit hardest: simple definitions and facts. “What is a pip”, “what is leverage”, “when does the London session open”. An AI answer covers these in two sentences, so few people click.
- Partly affected: general comparisons and how-to questions. Users get a summary first and click only if they want detail or proof.
- Still strong: anything that needs first-hand testing, current data, a tool or a decision. Hands-on broker reviews, fee comparisons with screenshots, calculators, live spreads and “which account should I open” questions still send visitors, and those visitors are closer to signing up.
If most of your traffic comes from glossary-style articles, expect it to shrink. If it comes from pages that show your own work, you are in a better position.
How AI systems choose their sources
The exact methods are not public and they change often, but some patterns are consistent across Google, ChatGPT and Perplexity:
- They can only cite what they can crawl. AI search features draw on search indexes. If a page is not indexed, or is blocked to the relevant crawler, it cannot be cited.
- They prefer passages that answer directly. A paragraph that states the answer in its first sentence, then explains it, is easy to quote. An answer buried under an intro about the history of forex is not.
- They favour sources others trust. Sites that are mentioned and linked by other respected sites in the niche get cited more often. Classic authority still matters.
- They like specifics. Numbers with a date, a clear method and a named author are more useful to an AI answer than vague claims.
Seven practical changes for trading affiliate sites
1. Lead with the answer
Start each section with one or two sentences that answer the heading’s question. Keep the explanation, examples and caveats underneath. This helps human readers who skim, and it gives AI systems a clean passage to quote.
2. Publish data, not just opinions
Add tables with real values: spreads you measured, fees from the broker’s own price list, deposit minimums, withdrawal times you tested. Put the date of the check next to them and update them on a schedule. “Updated monthly, last checked on…” is a strong signal for both readers and machines.
3. Show who wrote it and how
Every review and guide should have a real author with a short bio, plus a methodology section: how you test brokers, what you pay attention to, how you handle affiliate relationships. In finance this builds trust with readers, regulators and search engines alike.
4. Check your robots.txt
Some sites blocked every AI crawler in 2023 and forgot about it. If you want to be cited in ChatGPT search or Perplexity, their search crawlers (for example OAI-SearchBot and PerplexityBot) need access. Note that Google’s Google-Extended token controls use of your content for Gemini model training; according to Google it does not affect whether you appear in Search or AI Overviews. Decide deliberately which bots you allow instead of copying a list from a forum.
5. Use structured data where it is still useful
Article, Organization, Person and Review markup help search engines understand who you are and what a page is about. Do not expect special rich results from every schema type, but clean, accurate markup makes your content easier to interpret.
6. Become a brand people search for
AI answers often mention brands by name. When people then search for that brand, you want it to be you. Newsletters, YouTube, Telegram channels and appearances on podcasts build that recognition, and they give you traffic that no algorithm update can take away.
7. Earn mentions on other trusted sites
Links and mentions from relevant sites are still among the strongest trust signals, and AI systems lean on the same signals. A focused link building strategy for forex and crypto sites supports both classic rankings and AI visibility.
How to measure AI traffic
AI visibility is harder to measure than rankings, but you can start with what you have:
- In your analytics, look for referrals from
chatgpt.com,perplexity.ai,copilot.microsoft.comandgemini.google.com, and track them as their own channel. - Ask the same twenty questions your audience asks in ChatGPT, Perplexity and Google once a month and note whether your site is cited.
- Watch conversions, not just sessions. AI referrals are usually fewer, but the visitors have already read a summary and often arrive ready to act.
- Make sure your affiliate tracking survives the visit. Server-to-server postbacks are more reliable than pixels here; we explain why in Pixel vs S2S postback tracking.
What this means for your business model
AI search rewards sites that do real work: testing, measuring, comparing and explaining with evidence. Thin pages written only to catch a keyword will keep losing ground. For most trading affiliates the answer is not to fight the change but to move up: fewer, better pages, your own data, tools people return to, and an audience you can reach directly.
If you want a second opinion on how exposed your site is, or help restructuring content for AI search, take a look at our services or get in touch.
