Samwise

A founder at an early-stage B2B AI startup spends weeks cold-messaging people on LinkedIn who seem like good fits

13 views

🔍 Click to enlarge

A founder at an early-stage B2B AI startup spends weeks cold-messaging people on LinkedIn who seem like good fits. Half ignore the messages. The other half respond but turn out to want something completely different from what the product does. Three months in, there's still no clear picture of who actually needs this thing. Samwise walks through ICP definition using a tested framework, then searches LinkedIn and Reddit for people matching that profile. It suggests conversation approaches and tracks what gets learned from each interaction.

Samwise analyzes call transcripts and email threads to pull out patterns. A founder talks to twelve potential customers across two weeks. Samwise identifies that three mentioned the same workflow problem, two asked about a feature that doesn't exist yet, and four weren't actually decision-makers. It updates documentation automatically and recommends which direction to explore next. The memory system connects dots across conversations that happened days apart.

A technical co-founder who hates sales can follow the message templates and conversation tips. Samwise doesn't write messages automatically—it provides structure for what to ask and how to position things. For pre-revenue teams, milestone tracking shows progress when revenue metrics don't exist yet. The brainstorming feature helps refine who to target when the initial hypothesis falls flat.

The Reddit search finds people complaining about problems in specific subreddits. Someone posts about struggling with data pipeline reliability in a developer community. that is a lead for an observability startup. Traditional outbound tools find job titles. This finds pain points.

It breaks down when conversations happen off-platform. A founder meets someone at a conference and has a detailed discussion about their needs. They'd need to manually input that conversation for Samwise to incorporate it. The LinkedIn search caps at 50 for the Pre-Revenue plan, which runs out fast if the ICP definition is too broad. Call transcript analysis requires someone to upload or paste the transcript—there's no automatic recording integration.

Teams that already have product-market fit and a defined sales process won't get much value. The iteration recommendations matter when everything's still uncertain. Once a startup knows exactly who buys and why, traditional CRM tools make more sense. Companies selling to enterprises with long sales cycles and multiple stakeholders will find the approach too simplified.

The 7-day trial includes unlimited brainstorming and ICP definition work but limits Reddit searches to 10 maximum. Good for testing the framework. Not enough to run a real outreach campaign. The Pre-Revenue plan costs $49 monthly and allows 50 LinkedIn searches with unlimited Reddit searches, memory, and milestone tracking. The Product-Market Fit tier runs $499 monthly and removes the LinkedIn search limit.

A founder running customer discovery interviews every week gets immediate value. Someone who already validated their market and needs help scaling outreach should look elsewhere. Samwise assumes conversations happen on LinkedIn and Reddit. Teams that find customers primarily through referrals, partnerships, or industry events won't have much to feed into the system. It's built for the messy early stage when nobody knows what works yet.