Get AI summaries of any video or article — Sign up free
How Twain's Founder Sells Without a Sales Team | Founder Fridays thumbnail

How Twain's Founder Sells Without a Sales Team | Founder Fridays

Notion·
5 min read

Based on Notion's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Twain is built around the belief that communication is a learnable skill and that AI should reduce rejection by improving outreach quality at scale.

Briefing

Twain’s founder, Muhammad Shaheen, built a deep-research platform aimed at one stubborn sales bottleneck: getting personalized outreach right without forcing founders and sales teams through “thousands of rejections.” The core belief driving the company is blunt—communication is a learnable skill, and AI should reduce the pain of repeated failure by helping people send outreach that’s more likely to land. The product matters because early-stage teams often can’t afford the time, volume, or emotional cost of trial-and-error messaging, especially when they’re still proving product-market fit.

Shaheen’s origin story traces back to multilingual life and the friction of communicating across cultures. While expanding his first venture, Eat Clever, he had to explain a novel restaurant concept to very different markets—then faced the reality that rejection is amplified when a product isn’t fully validated yet. That combination—difficulty conveying value quickly and the high rejection rate common in sales—became the practical problem behind Twain.

In terms of what works for early-stage founders trying to generate leads, Shaheen recommends starting with a hypothesis about a problem, making an assumption about who has it, and then testing quickly with targeted outreach. He points to LinkedIn as a practical starting point: send a small set of templates to a few ideal customer profiles (ICPs), talk to anyone who shows interest, and look for repeatable problems that suggest we’re onto something. But the messaging guidance is more psychological than technical. Founders should write from the recipient’s mindset—no one is eager to hear from a “no-name” stranger—and should anticipate surface-level objections. He gives a concrete example from his own experience: AI-generated personalized emails “sucked,” so Twain’s approach accounts for the skepticism people have toward AI and preempts the “why should I trust you?” barrier.

A standout product feature is Twain’s qualification agent. After generating outreach and conducting research, it also qualifies the data and the messages themselves—checking whether a generated message sounds good and whether it’s appropriate to reach out to a specific person. If the output doesn’t meet the bar, the system regenerates, turning personalization into a loop that filters for quality rather than just quantity.

Growth and go-to-market themes show up repeatedly. Twain’s rapid scaling is attributed primarily to product-market fit and, later, to usability improvements—especially making the workflow convenient for companies with a few hundred employees. Shaheen also emphasizes that the founder can be the first salesperson; he personally handles demos and top-of-funnel outreach while using Twain himself. Internally, he uses Notion to create clarity through priority batches, avoiding sprawling to-do lists and keeping engineering and data science aligned on what matters most next.

On fundraising, Shaheen credits a pre-seed introduction path tied to Sequoia Capital’s renewed interest in pre-seed investing, and he advises founders to stay authentic—rejection from top-tier firms doesn’t necessarily reflect the idea’s value, just timing and fit. Building Twain differently from his first company, he highlights the speed of solo decision-making, while admitting it’s “brutal and lonely.” The final takeaway is simple: trust yourself, because confidence is hard to outsource—execution still has to be done by the founder.

Cornell Notes

Muhammad Shaheen built Twain to reduce the rejection-heavy grind of personalized sales outreach. His guiding belief is that communication is a skill people can learn, and AI should help teams avoid “thousands of rejections” by generating outreach that better matches what recipients will actually accept. For early-stage founders, he recommends hypothesis-driven outreach: pick an ICP, send a few LinkedIn templates, talk to responders, and look for repeatable problems. Twain’s qualification agent adds a quality-control layer by evaluating research and generated messages and regenerating when outputs don’t sound good or aren’t worth sending. Usability and workflow fit—plus founder-led sales early on—are presented as key drivers of traction and team clarity.

What sales problem pushed Shaheen to build Twain, and why does it matter for early-stage teams?

He linked sales outreach failure to two compounding forces: people struggle to convey a novel idea quickly, and rejection becomes harsher when a product isn’t yet product-market fit. During Eat Clever’s expansion across Europe, he saw how difficult it was to communicate across cultures and how rejection stacks up in sales. Twain is designed to make personalized outreach less rejection-prone by improving communication quality at scale, so founders can test and learn without burning time and emotional bandwidth.

How should early-stage founders structure outreach experiments before they have certainty?

Shaheen recommends starting with a hypothesis about a problem, then making an assumption about who has it (since certainty isn’t available). From there, he suggests testing with a small number of LinkedIn templates sent to a few ICPs. The goal isn’t immediate perfection; it’s to talk to anyone who shows interest and identify repeatable problems. If patterns emerge, the founder is “on to something.”

What mindset should founders adopt when writing messages to strangers?

He argues founders must write from the recipient’s perspective: no one is waiting for a message from an unknown brand. That means outreach must convey value, but also anticipate the recipient’s “wall of defense”—surface-level objections and skepticism. He gives a personal example: AI-generated personalized emails he tried “sucked,” so Twain’s communication strategy accounts for the objection that AI output may be low quality or untrustworthy.

What exactly does Twain’s qualification agent do, and how does it improve outreach quality?

After Twain generates outreach and performs research, the qualification agent evaluates the data and the messages. It checks whether Twain should reach out to a specific person and whether the generated message sounds good. If the message fails those checks, Twain regenerates it for the user, turning personalization into a quality-filtering loop rather than a one-shot generation.

Why does Shaheen emphasize usability after finding an ICP?

Once Twain identified its ICP and the core problem it solves, the bottleneck shifted from discovery to execution. Shaheen says usability is easy to underestimate, especially for scaleups with a few hundred employees, where teams need convenience and minimal disruption to existing processes. He also notes that native integrations improved usability and correlated with impact on adoption.

How does Shaheen use Notion to run product development without turning it into a democracy?

He uses Notion to create priority clarity rather than generic to-do lists. Features are documented with screenshots and videos, and each feature lives under a parent page. Work is organized into batches (e.g., batch one for engineering and data science), with the team moving to batch two only after batch one completes. After batch three, items go to a parking lot—keeping focus on what matters most next.

Review Questions

  1. What evidence does Shaheen use to justify that communication quality—not just outreach volume—is the core sales lever?
  2. Describe the step-by-step outreach experiment Shaheen recommends for early-stage founders using LinkedIn templates and ICP assumptions.
  3. How does Twain’s qualification agent change the workflow from “generate personalization” to “generate and validate personalization”?

Key Points

  1. 1

    Twain is built around the belief that communication is a learnable skill and that AI should reduce rejection by improving outreach quality at scale.

  2. 2

    Early-stage outreach should follow a hypothesis → ICP assumption → small template test → conversations → identification of repeatable problems loop.

  3. 3

    Effective messaging must account for the recipient’s mindset and preempt surface-level objections, including skepticism toward AI-generated personalization.

  4. 4

    Twain’s qualification agent evaluates both research and generated messages, regenerating outputs when they don’t sound good or shouldn’t be sent.

  5. 5

    Usability becomes the main growth lever after ICP and problem clarity, especially for teams with established processes.

  6. 6

    Founder-led sales early on can work: Shaheen personally runs demos and top-of-funnel outreach while using Twain.

  7. 7

    Notion is used for product clarity through priority batches and feature documentation, avoiding broad opinion-seeking in early-stage execution.

Highlights

Twain’s qualification agent doesn’t just generate outreach—it qualifies whether the message is worth sending and whether it sounds good, then regenerates when it fails.
Shaheen’s outreach advice starts with hypotheses and assumptions, then uses small LinkedIn template tests to find repeatable problems rather than chasing perfection immediately.
A central messaging principle: write from the recipient’s perspective—no one is eager to hear from a no-name stranger, so objections must be anticipated upfront.

Topics

  • Personalized Outreach
  • Founder-Led Sales
  • Qualification Agent
  • Notion Workflows
  • Usability
  • LinkedIn Templates

Mentioned

  • Muhammad Shaheen