Two Years, a Handful of Clients – Then Dozens of Enterprises in One

01 /The Concept

Fibr built a platform for website optimization – specifically for improving conversions on landing pages and key site pages.

One use case is adapting pages to match paid traffic – ensuring that whatever impulse made a user click an ad doesn't evaporate the moment they arrive on the destination page.

Another is tailoring pages for traffic arriving from AI chat referrals – ChatGPT, Perplexity, and similar. When a user follows a link from an AI recommendation, they arrive expecting either a direct answer to their question or confirmation that the site genuinely delivers what the AI suggested. Generic landing pages fail this test.

A third dimension is personalization by visitor context: country, referring source, and visit number. A returning visitor who didn't convert before can be shown something different on the same URL – a fresh angle that might finally land.

Fibr's framing extends the concept further: every page on a site can be treated as a landing page, because every page is reached via a link carrying a prior context and user expectation.

The AI handles hypothesis generation and A/B testing, running the optimization loop "on autopilot" – though this is partly a simplification. In practice, humans approve the AI's hypotheses and set the strategic parameters. The better description is that humans handle strategy and the AI handles execution.

Results from existing deployments: 4× more leads, 30% lower customer acquisition cost, 28% improvement in return on ad spend.

Fibr was [covered previously](/review/chem-tochnee-sootvetstvie-tem-luchshe-prodazhi) when it raised its prior funding round in summer 2024. The company has now raised an additional $5.7M.

02 /Why It Matters

Fibr was founded in 2023 but spent its first two years with almost no paying customers – at one point hovering between one and two active accounts. Something shifted last year: by year-end the startup had 12 clients, including large US companies in financial services and healthcare.

That's meaningful validation that enterprise appetite for this category has finally materialized.

Kenobi ([related review](/review/glavnyj-sekret-vysokih-konversij)) provides further confirmation. Having completed Y Combinator back in 2022, it launched a fully rebuilt platform late last year focused on real-time site personalization. Their current approach differs in execution – a small overlay at the top of the page invites visitors to enter their company name or website URL, after which the AI analyzes what that company does and adapts the page content accordingly. Enter a startup intelligence publication's URL and the site starts framing its pitch around how it could serve that publication's audience.

Personalization is now a recognized trend across the full site experience. Several startups have gone deep on specific page types:

PrettyDamnQuick ([related review](/review/esli-ne-hochesh-terjat-uzhe-gotovyh-pokupatelej)) has raised $38M for a platform optimizing e-commerce checkout pages – converting cart additions into completed purchases.

Helium ([related review](/review/kak-lego-no-dlja-produktovikov-i-marketologov)) raised $7.9M to optimize paywall and subscription pages inside mobile apps. The company is now expanding scope while maintaining its focus on app developers – new areas include onboarding flow optimization and web-based subscription pages.

03 /Opportunities

The obvious direction is building AI-powered platforms that dynamically adapt websites and apps to individual visitors. If Fibr's trajectory is any indication, enterprise demand for this capability is just starting to build.

But there's a second direction emerging that points further out.

An Accel partner who led all of Fibr's investment rounds noted that the real long-term potential of the platform is in optimizing sites not just for human visitors – but for AI agents.

In the near future, users will increasingly delegate browsing and purchasing to AI agents, which will navigate the internet autonomously on their behalf. Those agents will be evaluating sites according to their own logic, not human psychology. A site that isn't legible to AI agents will be invisible to a growing share of traffic.

"This part is still early," the investor noted, "but we prefer to back startups that build for today's needs in a way that can flex to meet tomorrow's demands." That's a useful lens for any founder in any vertical – not just personalization.

So if you're building a site optimization platform today, design it from the start to work for AI agent visitors as well as human ones.