#dreamdatarecipes
Track previously hidden organic LLM-traffic in Dreamdata
Large language models (LLMs) are now a standard for researching brands, services, or products. Instead of following a link from Google search, people ask these AI chatbots their questions directly and click on the suggested pages.
This shift means that a big portion of your potential audience is now engaging with your content through an entirely new lens.
So, naturally, we want to know how to measure the impact of this 'LLM-influenced' traffic with Dreamdata. In this recipe, we will look at website traffic coming from LLMs and how this influences pipeline and revenue.
Here’s how to do it:
Step 1: Categorize LLM touches
To analyze the influence of LLMs, we need to categorize LLM touches into the right channels and sources.
Let’s go to the Data Hub, where you can control your data to ensure your reporting is accurate and reliable.
Click on UTM Mappings. We automatically map your UTM parameters directly into standardized channels and sources. But you can always edit these mappings.
As you can see, we have already mapped three UTM Sources: ChatGPT, Gemini, and Perplexity.
We’ve mapped them to the channel Organic LLM and into their respective sources to categorize any traffic with these specific UTM source tags as coming from those LLMs.
We’ve also done this for Referrer Hosts. If a link suggested by the LLMs lacks a UTM Source, any incoming traffic would be falsely categorized as “direct”.
To avoid this, you map the LLMs as Referrer Hosts so you won’t miss out on any incoming traffic from ChatGPT, Gemini, and Perplexity.
Step 2: Analyze page traffic on your website coming from LLMs
8. Now, let’s head to Engagement > Pages to identify and analyze website activity coming from LLMs over the past 12 months, and measure the MQLs this traffic yielded.
Time period: last 12 months
Stage model: MQL
Channel: Organic LLM
Secondary Group By: Source
9. Now we can see that 1,855 page views came from LLMs in the last 12 months and these have influenced 30 MQLs with a value of €820,0000.
10. In the graph ‘Views Group By URL Over Time’, we can see that the incoming LLM-sourced traffic has significantly increased since the beginning of 2025. Showing a rising trend in the number of people using AI chatbots for search.
11. Let’s sort the table ‘URL Performance Group by Source’ by ‘Influenced Value’ to see which LLM-surfaced URLs have the highest influence on MQL generation.
Step 3: Reflect on your SEO strategy
Seeing Rank, Traffic, Number of Deals, and Revenue of each URL in the same table makes it effortless to evaluate your SEO strategy.
For instance, you can assess whether going after high volume keywords which bring in tons of traffic actually pays its dues. Or whether niche low volume keywords are the source of value.
Similarly you can juxtapose rank with pipeline and revenue.
Once you find the patterns you can focus your SEO work on the pages that actually do bring in revenue for your company.
How it works:
Dreamdata offers end-to-end customer journey tracking: enabling the platform to see who’s coming onto your site and from where. Channel, campaign, content, if it’s tracked, it’s measured.
And by connecting to your CRM, the measurement benchmark is actual pipeline and revenue generated.
For the SEO, this means linking URL with pipeline and revenue. URLs (and keywords) whose SERP can then be cross-referenced, to reveal whether the keyword is actually delivering where it matters.