For two decades, link building was the main way to build off-site authority. Links still matter: they help search engines discover pages and judge which ones others vouch for. But AI assistants add a new layer. When someone asks "what's the best tool for X?", the answer is shaped by how often and how consistently your brand is mentioned across the web, linked or not.
Why mentions matter more now
Large language models learn from huge amounts of text. When they browse, they read the pages that rank. In both cases, they form a picture of your brand from what is written about it:
- How often you are mentioned in the context of a topic.
- Who mentions you: respected publications, practitioners, review sites, communities.
- What they say: the category you belong to, what you are good at, who you are for.
- How consistently these facts line up.
A brand that is mentioned in many credible places, described the same way, is far more likely to show up in AI answers than one with a strong link profile but little real conversation around it.
Links vs. mentions
- Main value: Backlinks: Discovery, ranking authority; Brand mentions: Entity understanding, recommendations
- Measured by: Backlinks: Link indexes; Brand mentions: Web, news, social, and AI answer monitoring
- Best earned by: Backlinks: Linkable assets, digital PR; Brand mentions: Expertise, community presence, PR, reviews
- AI impact: Backlinks: Indirect, through rankings; Brand mentions: Direct, through how models describe you
The best activities earn both at once.
Where mentions come from
- Industry publications and news: expert commentary, data stories, product launches.
- Comparison and "best of" articles: being included in list posts that rank for your category.
- Review platforms: G2, Capterra, Trustpilot, app stores, and niche review sites.
- Communities: Reddit, Stack Overflow, Hacker News, industry Slack groups, and forums. These are heavily represented in AI answers.
- Podcasts, webinars, and conference talks, especially when transcripts are published.
- Partner and integration pages: marketplaces and partner directories.
- Your own people: founders, engineers, and experts writing and speaking under their own names.
How to earn them
Publish things worth citing
Original research, benchmark data, free tools, and clear frameworks get referenced repeatedly. One strong data study can earn mentions for years.
Be useful where your buyers ask questions
Answer questions in communities honestly and without a hard sell. Practitioners notice, and these threads are often what AI assistants read.
Make journalists' and writers' jobs easy
Keep a press page with a clear company description, founder bios, logos, and key facts. Respond quickly to expert requests in your field.
Get into the lists
Find the "best [category] tools" articles that rank and that AI answers cite. Reach out with accurate information and, where appropriate, a demo or trial for the author.
Keep the story consistent
Describe your company the same way everywhere: your homepage, LinkedIn, directories, review profiles, and press materials. Inconsistency confuses both people and machines. Organization schema and llms.txt help reinforce the official version.
What to avoid
- Paid or manipulative link schemes. They still violate search engine policies and can lead to penalties.
- Fake reviews or astroturfed community posts. They damage trust when discovered and often break platform rules.
- Mass-submitting to low-quality directories. It adds noise, not authority.
Measuring progress
Track three things over time:
1. Mention volume in credible sources for your core topics. 2. Earned links from those mentions. 3. AI Share of Voice: how often AI assistants mention or cite you for the prompts that matter.
If mentions grow but AI visibility does not, check that the mentions describe you in the right category. If AI visibility grows, you will usually see more branded search follow.
SearchVoro tracks how AI assistants describe and cite your brand compared with competitors, so you can see whether your PR and community work is changing what people hear about you.
About this byline
“SearchVoro Team” identifies collaborative product and engineering material, not an invented individual author. Named contributor and reviewer profiles will be added only with the person's approval.