In partnership with

While everyone's racing to build faster, flashier AI products, Liner is betting on something different… and they think they're the only ones getting this right.

In an era where Perplexity (valued at $14Bn) is solving search and Claude (valued at $61Bn) writes everything for everyone, Liner's CEO Luke Jinu Kim made a contrarian choice: Instead of trying to beat the giants at their own game, he built an AI search engine that prioritizes accuracy above all else - and now he’s putting up a fight to some of the AI giants. 

The Conventional Wisdom Everyone Follows:

Speak to any “AI for [INSERT INDUSTRY]” founder, and you'll hear the same thing:

1. Ship fast, fix later

2. Build horizontal assistants

3. Monetize through freemium or display ads

These beliefs have driven most AI search companies toward the same model: general-purpose tools that try to answer any question about anything, monetized through freemium subscriptions or display advertising.

Perplexity epitomizes this approach—fast, broad-ranging answers with a CPM advertising model that pays just for views, not performance.

Liner's Contrarian Bet

Luke saw the same opportunity everyone else did but drew a different conclusion. "People don't trust AI," he observed from user interviews. 

Instead of building another general-purpose AI assistant, Liner went as narrow as they could, with three central decisions defining their strategy.

#1) Choosing a niche customer 

While competitors tried to be everything to everyone, Liner consciously limited their scope:

While it may seem like Luke was restricting his market a bit — this was very ‘Zero to One’ as Peter Thiel also said that it's better to dominate a small market than compete in a large one. 

Liner chose to own the academic research niche completely rather than fight for scraps of the general AI market. We actually covered Luke’s approach to gathering customer feedback here. 

Here's why going narrow works:

1. Deeper Understanding = Better Product: When you serve everyone, you serve no one particularly well. By focusing only on researchers, Liner could build features like sentence-level citations and academic paper access that their broad competitors couldn't justify.

2. Community-Driven Growth Academic communities are tight-knit. Liner's 10% penetration at Berkeley created word-of-mouth growth within research networks. As Luke notes: "When they got to kind of love liner, they tend to introduce liner to other student groups."

The result: Massive growth with their core customer segment across top tier universities including but not limited to over 10% penetration at Berkeley (who some competitors cough Perplexity cough is targeting quite aggressively) 

#2) Building what people actually want 

Once Liner went narrow and understood their specific users, they discovered something crucial: researchers and students don't care about sub-second response times when the wrong answer costs them hours of work.

In a world obsessing over response time, Liner optimized for something harder to measure: being right. 

They built systems to provide "creative sources" and "fact-check line by line," with every answer traced back to specific parts of specific sources.

Understanding What Actually Matters: While general AI tools compete on speed and creativity, Liner's narrow focus revealed that their users had a different priority hierarchy. They found that researchers were burnt by inaccurate data from the AI giants and as a result, they were willing to sacrifice some time for accuracy. 

This insight led to product decisions competitors couldn't make:

• Sentence-level hallucination reporting (useless for casual users, essential for researchers)

• Click-through tracking to original sources

• Integration with academic databases and patent documents from day-1

The result: Liner put accuracy first and achieved 96% accuracy compared to GPT-4's 90% and Perplexity's 93.7%. In their target market, being right mattered infinitely more than being fast.

#3) Taking a different approach to monetization

Once Liner proved they could deliver accurate, trustworthy results, they made another contrarian choice: while Perplexity chose CPM advertising (paying for impressions), Liner went with CPC (pay-per-click).

This wasn't just a monetization decision—it was a confidence statement about their product quality.

Why This Model Choice Matters: While CPM pays just for showing ads (regardless of user action), CPC only pays when users actually click through. 

This choice reflects their core insight: when users trust your sources, they'll actually click through to verify and learn more. It's a business model that rewards quality over quantity.

Aligning Business Success with User Value: 

By choosing CPC, Liner aligned their economic incentives with user behavior. They make money only when users find their sources valuable enough to explore—not just when they show up.

The result: Liner's CPC model outperforms display advertising in conversion rates, with Luke backing himself & Liner to the point where he believes that they might even challenge Google Search ads.

What this means for you

1. Find your underserved niche: Look for markets where the conventional wisdom leaves specific users poorly served. Liner found that researchers needed accuracy while everyone else optimized for speed.

2. Let constraints drive excellence: Choosing a narrow market isn't limiting—it's liberating. It lets you build something exceptional rather than mediocre for everyone.

3. Align your business model with your value: Your monetization should reward you for providing what your customers actually want. If they value quality clicks over volume views, choose CPC over CPM.

4. Question industry best practices: Just because everyone else is doing it doesn't make it right. Sometimes the biggest opportunities lie in doing the exact opposite of conventional wisdom.

If you enjoyed this issue, please share it with fellow founders, operators, & investors and fill out our feedback form below:

See you next week!

Rahul & Aryaman