How AI referral traffic drove 118% more free trials
Phrase It's search-led content and AI visibility work helped a real estate technology company turn growing visibility in AI assistants into measurable business actions. AI referral sessions increased 21.4%, but the bigger change came further down the funnel: key events rose 90.5% and free trials increased 117.6% compared with the preceding dashboard period.
AI visibility needed to become more than a traffic report
The client provides software for real estate professionals. Its website includes high-intent product pages alongside educational resources that help agents improve listing, buyer, seller, and marketing presentations.
As tools such as ChatGPT, Perplexity, Gemini, Claude, and Microsoft Copilot became part of the research journey, the opportunity was not simply to appear in AI-generated answers. The work needed to help the right prospects discover useful content, move into commercial pages, and take a measurable next step.
The performance review covered the exact 365-day period from July 22, 2025 to July 21, 2026. Results were evaluated against the preceding comparison period shown in the client’s AI performance dashboard.
- Real estate technology company with a product-led website
- Educational and commercial pages working across the buyer journey
- Exact 365-day measurement window with a preceding comparison period
More AI visits would not matter if they did not convert
AI referral traffic is still a relatively small and unevenly attributed channel. Source names and mediums vary between platforms, and even the same assistant can appear under several labels. That makes headline traffic totals easy to misread and creates a risk of treating visibility as business impact.
The real test was whether AI-referred visitors engaged with the site and completed meaningful actions. That meant looking beyond sessions to engaged sessions, engagement rate, form submissions, free-trial clicks, and completed free trials.
There was also a concentration challenge. ChatGPT already accounted for most measurable AI traffic, while other assistants contributed much smaller volumes. The strategy therefore needed to build on what was working without treating one platform as the entire AI search market.
- Fragmented source and medium labels in analytics
- Commercial value mattered more than raw referral volume
- Heavy reliance on one AI platform created concentration risk
Connect useful answers to the next commercial step
Phrase It focused the content and AI visibility work on the questions real estate professionals ask while researching presentation workflows, marketing tactics, and client communication. Educational resources were built to answer those questions clearly while creating a relevant path towards the product.
The content mix supported more than one stage of the journey. Informational pages explained topics such as pre-listing presentations, virtual staging, seller reports, REO listings, and buyer offers. High-intent destinations—including pricing, free access, and embedded signup pages—gave interested visitors a direct route to evaluate or try the software.
Performance was then reviewed at both source and landing-page level. AI sessions were separated from the site’s broader channel totals, related source variants were assessed together, and conversions were examined alongside traffic. This made it possible to see which assistants and pages were contributing meaningful actions rather than visibility alone.
- Created search-led resources around practical real estate questions
- Connected educational discovery with pricing, free access, and signup paths
- Measured AI sources, landing pages, engagement, and conversion actions together
Conversions grew more than four times faster than traffic
During the 365-day reporting period, AI platforms generated 1,458 sessions, up 21.4% from approximately 1,201 in the preceding comparison period. Those visits also became more engaged: engaged sessions increased 27.6% to 939, while the engagement rate reached 64.40%.
The strongest change happened at conversion level. AI-attributed key events increased 90.5% to 80, and free trials rose 117.6% from approximately 17 to 37. The reported session key-event rate improved from approximately 3.50% to 5.49%, indicating that AI traffic became materially more likely to produce a tracked action.
Commercial and educational pages both contributed. Pricing, free access, embedded signup, seller-report, virtual-staging, and pre-listing-presentation content all recorded key events from AI referrals. The pre-listing-presentation resource was especially notable because it generated visibility and conversions across ChatGPT and Perplexity source variants, rather than through only one assistant.
- 1,458 AI referral sessions, up 21.4%
- 80 AI-attributed key events, up 90.5%
- 37 AI-attributed free trials, up 117.6%
What the numbers include—and what they do not
ChatGPT generated 1,246 of the 1,458 AI sessions and 69 of the 80 AI-attributed key events. It also accounted for 35 of the 37 free trials. Perplexity and Gemini produced smaller but measurable conversion activity, while Claude and Microsoft Copilot contributed referral traffic during the period.
The dashboard splits some assistants across multiple source and medium labels. It also contains landing-page rows with reported key-event rates above 100%, so those row-level rate labels require validation before external use. All results describe this engagement and should not be treated as guaranteed outcomes.
- AI-specific conversion table used for all conversion totals
- Source variants assessed together where appropriate
- No revenue claim because purchase attribution was not available
Questions about the work and results
What counts as AI referral traffic in this case study?
It means sessions attributed in the client's dashboard to AI assistants including ChatGPT, Perplexity, Gemini, Claude, Microsoft Copilot, and a small number of other AI-related sources. Some platforms appear under more than one source or medium label.
Which AI platform produced the most measurable activity?
ChatGPT was the largest source by a wide margin. It generated 1,246 sessions, 69 key events, and 35 of the 37 AI-attributed free trials during the reporting period.
Did informational content contribute to conversions?
Yes. Resources about presentations, virtual staging, seller reports, REO listings, and buyer offers recorded key events from AI referrals. The pre-listing-presentation resource produced activity through both ChatGPT and Perplexity source variants.
Why did conversions grow faster than sessions?
The dashboard shows that the AI-referred audience became more engaged and more likely to complete tracked actions. Sessions increased 21.4%, while key events increased 90.5% and free trials increased 117.6%. The data supports stronger conversion efficiency, although it does not isolate one page or change as the sole cause.
Do these results prove revenue from AI search?
No. The dashboard recorded free trials, clicks, form actions, and other key events, but it did not record purchases. The results demonstrate measurable funnel activity from AI referrals, not closed revenue.
Why is the client not identified?
The company has been anonymized to protect client confidentiality. The dates, metric definitions, values, and reporting limitations have been retained so the case study remains specific and auditable without naming the business.
Tell us what good looks like for you.
Share your site, goals, and what is getting in the way. We'll reply with the clearest next step, whether that is working together or fixing something first.
Prefer email? info@phraseit.agency