Featured Work
SEO turns AI referral traffic into Softeq’s highest-converting demand channel
Overview
Softeq is a custom software and hardware development company serving enterprise clients across North America and Europe, with engineering work spanning embedded firmware, IoT, mobile, cloud, and AI applications.
By the middle of 2025, a fresh audience segment surfaced on the Softeq site, users arriving via ChatGPT citations after the large language model (LLM) recommended the company in its responses. That’s a different kind of intent than a typical search click, and it showed up right as Softeq set aggressive targets for 2026.
That traffic was attributable to our targeted SEO efforts. Our SEO work throughout 2025 had been built around how LLMs read and cite content, using clear page structure, plain-language technical writing, topic clusters, and consistent terminology. By August, AI referral sessions had reached 7% of Softeq’s total site traffic, and ChatGPT was driving 86% of it.
Most companies were treating this kind of traffic as a curiosity. We treated it as a channel to prove out, tracked it through the full funnel in HubSpot, and asked the only question that mattered.
Was this generating real demand or just traffic?
Recognition
AMA Houston Crystal Awards Winner
Digital Marketing: Artificial Intelligence
Solution
We set four benchmarks against Softeq’s best-performing existing channels, then built on the SEO work already underway to see if it could clear them.
On-page optimization: We optimized 26 to 30 high-priority pages across 2025, backed by more than 20 blog posts built to establish topical authority. The first half of the year focused on broad coverage, but by Q3 we narrowed to the six pages most likely to earn AI visibility.
Content built from real buyer language: Analysis of sales call recordings surfaced the ten services buyers asked about most, which shaped new pillar content. FAQ sections on priority pages came directly from that same buyer language, not from assumed search terms.
By mapping AI referral sessions through the HubSpot funnel (from initial contact to final deal), we measured performance against every established demand channel to validate the quality of the new traffic.
Results
%
Session-to-contact rate
%
Contact-to-lead rate
This is awesome!
I’d like to talk about getting results like this on my next project.