Graphite and Similarweb published research in late September 2026 on the early ChatGPT advertising market. Graphite’s practical playbook says its partner panel measured exposure and click behavior while its own team examined ad copy and landing pages. It reports that 43% of ad-eligible US users had encountered ads and that click-through rates had risen since launch. These are panel and agency observations, not audited conversion rates for every sector or a forecast for Saudi Arabia.

Why a conversational placement behaves differently

An ad appears alongside an active information-seeking conversation. The advertiser supplies context and creative; matching depends on the current discussion rather than a conventional list of search keywords. A person asking about a problem may be closer to evaluation than someone scrolling a feed, but curiosity is not purchase intent. The landing page must answer the specific need implied by the ad, and the offer must be clear enough to assess quickly.

Build an incrementality test

Choose one service with a reliable conversion path and an eligible market. Define a narrow context, write several distinct use-case ads, and send each to a relevant page with tracked parameters. Establish a baseline for qualified enquiries and revenue before the campaign. Run a capped pilot, inspect search-term equivalents and lead quality weekly, and compare incremental booked or paid outcomes with Google Search and Meta. Count view-through results cautiously and remove duplicates in the CRM. Graphite’s reported costs and average click rates are planning inputs, not local benchmarks.

Healthcare and GCC considerations

Healthcare advertisers need eligibility and policy review before launch. Avoid implying a diagnosis or targeting sensitive personal conditions. A Saudi advertiser should confirm local availability and Arabic creative performance in the actual account, then assess consultation quality and patient acquisition cost, not just clicks.

Karim’s strategic takeaway

Treat conversational ads as a small, instrumented experiment. The opportunity is to appear during problem research with a useful next step; the proof is additional qualified business after attribution and policy checks.