Semrush published "Digital PR for AI visibility: 5 tactics + how to measure" on July 30, 2026. The company is trying to pull digital PR out of the backlink era and into the answer-engine era. Earned media is now part of the source layer AI systems use to generate citations, mentions, and recommendations, which makes PR a visibility tool, not just a reputation play. For marketers, the shift is blunt: if your brand is not present in the external sources AI systems trust, it is easier to leave you out of the answer.
What Semrush is really signaling
The 12-minute read is by Cecilia Meis, with Faizan Ali listed as contributor.
The timing fits Semrush’s wider AI search push. The company’s June 1 AI Visibility Index study used 126 million real user prompts and set benchmarks across 22 industries, while Semrush Academy’s AI Visibility Essentials course focuses on prompts and brand mentions in AI responses instead of keywords.
Which digital PR tactics are most likely to move the needle
The strongest tactics are the ones that produce material AI systems can quote with confidence. That means original data, expert commentary, highly specific insights, and clear brand entity signals that survive distribution across credible third-party outlets. Answer engines are more likely to reuse what looks authoritative, specific, and widely corroborated than a generic brand pitch.
- Original data and reports carry the most weight because they create something worth citing. Semrush’s own AI Visibility Index is built on 126 million prompts, which shows how central data-led benchmarking has become to this conversation. A smaller brand can use the same logic with surveys, proprietary research, customer data, or category analysis that gives journalists and AI systems a concrete fact to reuse.
- Expert quotes matter when they are attached to a recognizable subject and repeated consistently across placements. Answer engines do not need a polished campaign statement as much as they need a clear, attributable point of view that matches the topic. The more often a named expert shows up in context-rich coverage, the easier it is for AI systems to connect that person and brand to a topic cluster.
- Third-party coverage is the real distribution layer. A Muck Rack study found that 84% of AI citations come from earned media and other third-party sources such as editorial coverage, reviews, and forums.
- Brand entity consistency matters. If the company name, product names, category language, and geography are inconsistent across the web, AI systems have more work to do before they can confidently surface the brand. Clear naming across authoritative coverage helps turn scattered mentions into a stronger entity profile.
How to measure visibility when attribution is weak
Semrush is pushing measurement toward prompts and responses for a reason: traditional PR reporting was never built for AI answers. AI visibility should be tracked by prompts instead of keywords and measured through brand mentions in AI responses. Keyword rank does not tell you whether an AI assistant cited your brand, skipped it, or recommended a competitor.
A practical measurement model needs a baseline and a repeatable prompt set. Track a defined set of buyer-intent and category prompts, log whether the brand is mentioned, cited, or recommended, and compare those results before and after major earned-media pushes. Use Semrush’s 22-industry benchmarks as directional context where they fit, but keep the real focus on your own prompt set and response patterns.
Do not expect referral data to tell the full story. AI visibility often shows up in weak or inconsistent analytics because answer engines can summarize sources without sending a clean click, and some visits will be buried inside blended traffic. That means the measurement stack has to include direct brand-mention tracking inside AI responses, not just backlinks, impressions, or sessions.
Why this is bigger than link building
Earned media is becoming a discovery signal inside AI search. In a LinkedIn post, Geertrui Laleman said Semrush found Patagonia was mentioned in 87,000+ AI answers in the United States between January and April.
