The system that turns customer feedback into faster product wins

This article summarizes an episode of Notion’s video series featuring Mohammed Shahin, founder and CEO of Twain.

Mohammed Shahin, founder Twain AI/ Photo credit: Mohammed Shahin

Most startups build products in a vacuum and hope customers will eventually care. Mohammed Shahin, founder and CEO of Twain, flips this process. He uses early customer outreach not just to sell, but to figure out exactly what to build in the first place.

This approach combines market feedback, product focus, and AI-driven messaging into a singular system that punishes slow-moving competitors.

Using early outreach to force market feedback

Founders often spend months building features without ever testing if buyers actually care. Finding the right product requires acting on missing facts.

Shahin explains, “You have to start with a hypothesis around a problem. Then you have to make an assumption. You can’t know for sure. You just have to make an assumption on who has that problem.”

Once you have an assumption, you need raw volume to test it. Shahin adds that this means sending a message to a few Ideal Customer Profiles who might have the problem. “Whoever shows interest, you’ll just talk to them, and if you find repeatable problems, that’s when it sounds like you’re on to something.”

Guessing and answering doubts builds trust
Buyers ignore cold outreach, so the first message must immediately address their skepticism.

Shahin notes, “Always remember what they don’t know about you, what’s their wall of defense… For [Twain], ‘I tried to generate personalized emails with AI and they sucked. AI can’t write.’ We preemptively talk about that.”

He treats rejection as a data point rather than a failure. Companies run into trouble when leaders treat communication like advertising, causing them to miss the specific complaints that actually block a sale.

Centralizing product control to avoid groupthink

When early teams try to turn market feedback into a group vote, progress immediately grinds to a halt. Group agreement slows early teams down. Shahin treats clear direction as a product owned by one person.

“I’m still the product owner, which I think is very important as a CEO,” Shahin says. “You have to be the product owner. It’s not a democracy… it’s very rough [in the] early stage, and you can’t ask 10 people for their opinion. It just doesn’t work.”

Rather than long wishlists, Shahin forces the team into rigid priority lists to stop distractions.

He notes, “In Notion, I describe features… every single feature gets its own page.”

Overcoming skepticism in automated messaging

Using AI to self-check and protect reputation