Bear Is Building the Analytics Layer for a World Where ChatGPT Sends You Traffic

The Macro: AI Search Is Creating a Marketing Channel Nobody Knows How to Measure

I talk to a lot of marketing teams. The conversation has shifted dramatically in the past six months. A year ago, everyone was worried about AI eating their SEO traffic. Now the conversation is different. They are seeing referral traffic from ChatGPT and Perplexity in their analytics, and they have no idea where it came from or how to get more of it.

This is a legitimate new marketing channel. When someone asks ChatGPT “what is the best project management tool for small teams,” the model recommends specific products. Those recommendations drive clicks. Those clicks convert. And unlike traditional SEO, where you can see exactly which keywords brought someone to your site, AI referral traffic is a black box. You know the visit came from ChatGPT. You do not know what prompt triggered it, what other products were recommended alongside yours, or whether you are showing up more or less over time.

The traditional SEO stack does not solve this. Ahrefs tracks backlinks and keyword rankings on search engines. Semrush does the same. Neither of them can tell you how often Perplexity recommends your product when someone asks about your category. Google Analytics shows you the referral source but nothing about the context. The measurement infrastructure for AI-driven discovery simply does not exist yet, and the companies that build it first will own a very valuable piece of the marketing stack.

This matters because AI search is not replacing traditional search. It is creating an additional channel with different rules. The content that ranks well on Google is not necessarily the content that gets cited by LLMs. The sources that LLMs trust are not the same as the sources that Google trusts. If you are optimizing only for traditional SEO, you are ignoring a growing chunk of how people discover products.

The Micro: A Coinbase Alum and a Teenage Founder Who Spoke at the UN

Bear tracks how often your brand appears in AI agent recommendations across ChatGPT, Claude, Gemini, and Perplexity. It identifies which prompts trigger mentions of your product, monitors how competitors appear in the same conversations, and helps you create content that is optimized for LLM citation rather than traditional search ranking.

Janak Sunil is the CEO. He worked at Coinbase and founded UCLA’s largest apartment listing platform, which was acquired in 2024. Siddhant Paliwal is the CTO. He was an early engineer at Third Chair (YC X25) and Intel, started his first company at 15, and received a United Nations invitation. They are a two-person team in San Francisco, Y Combinator Fall 2025 batch.

The product has four main components. First, visibility tracking that monitors brand mentions across major AI platforms. Second, trending prompt analysis that identifies high-volume searches relevant to your category. Third, a blog agent that generates content structured for easy ingestion by language models, which is a different optimization target than traditional SEO content. Fourth, automated outreach that finds third-party sources cited by LLMs and helps you get mentioned on those pages.

That last feature is the most interesting and potentially the most valuable. LLMs do not just make up recommendations. They cite sources. If Perplexity recommends your competitor because a specific blog post mentions them favorably, Bear identifies that blog post and helps you get mentioned on similar pages. It is essentially a new form of digital PR, optimized for AI citation rather than traditional link building.

Pricing starts at $100 per month for the basic plan, which tracks GPT-5 mentions and includes up to 30 prompt monitors and two blog posts per month. Enterprise pricing is custom. They already have 60 or more companies using the platform, including Browserbase, Wispr Flow, and Cal.com. For a company at this stage, that is solid early traction.

The competitive landscape is still forming. Otterly.AI and Profound are working on similar AI visibility tracking problems. The space is early enough that nobody has locked it down, but it is maturing fast. Bear’s advantage is the combination of tracking and action. Monitoring where you show up is useful. Actually doing something about it, through content generation and outreach, is where the money is.

The Verdict

I think this is one of the more important emerging categories in marketing technology. The shift from traditional search to AI-assisted discovery is real, measurable, and accelerating. The companies that figure out how to track and optimize for AI recommendations will have a genuine competitive advantage, and the tools that help them do it will be valuable.

Bear is early but directionally right. The product covers the full loop from monitoring to action, which is better than tools that only show you data without helping you act on it. The founding team is young but has real building experience, and the early customer list suggests the product is already delivering value.

At 30 days, I want to see correlation data. When Bear identifies a prompt where a company is not showing up and they take action, does the company start appearing in responses? That feedback loop is the entire value proposition. At 60 days, I want to understand how stable AI recommendations are. If ChatGPT changes what it recommends every week, optimization becomes a treadmill. If recommendations are sticky, the ROI compounds. At 90 days, the question is whether this becomes a must-have line item in marketing budgets or a nice-to-have that gets cut when budgets tighten. The answer depends entirely on whether they can prove measurable revenue impact from AI-driven traffic.