GTM Prospect
Finder.
An AI-assisted B2B outreach tool that brings prospect ranking, likely pain points, and opening lines into one workflow.
The story in brief
I built GTM Prospect Finder end to end in Lovable as a lightweight MVP for B2B outreach preparation.
A user enters a target company and what they are selling. The tool returns ranked prospect profiles, likely pain points, suggested opening lines, and LinkedIn search links in one screen.
An MVP built in two days. Time savings, prospect accuracy, and outreach performance have not been established through measured testing here.
The problem I focused on
Before sending a useful message, a sales rep needs to answer two questions: who should I contact at this company, and why would my offer matter to them?
That preparation can involve moving between prospecting tools, LinkedIn, and an AI assistant. Generic lists leave the rep to work out relevance and write the opener separately.
Help me choose a relevant person and prepare a useful first message in the same place.
The workflow
1. Add context
Enter the target company and describe the offering.
2. Compare prospects
Review ranked profiles with role context, likely pain points, and suggested opening lines.
3. Verify & act
Use a LinkedIn search link to check the person’s identity and current role before adapting the message.
The five output fields are name, title and seniority, likely pain points, an opening line, and a LinkedIn search link. A search link helps with verification; it is not a verified profile URL.
Product decisions & tradeoffs
| Decision | Reasoning | Tradeoff / limit |
|---|---|---|
| One-screen interface | Inputs on the left and results on the right keep the account context visible while comparing prospects. | A focused screen needs clear result structure as the list grows. |
| Rank by several signals | Combine seniority, function fit, and inferred buying influence rather than relying on title alone. | Buying influence is inferred. A high rank does not confirm authority or interest. |
| Tie pain points to the offer | Make the result relevant to what the user is actually selling. | Likely pain points are hypotheses, not facts about the person or company. |
| Draft an opening line | Connect research to a concrete next step in outreach preparation. | Generated personalisation needs review to avoid unsupported claims. |
| Link to LinkedIn search | Make it easier to move from an AI suggestion to checking a real person. | Search may return multiple or incorrect matches; identity and current role still need confirmation. |
Build & scope
I built the MVP end to end in Lovable over two days. The ranking logic blends seniority, functional relevance, and inferred influence on the buying decision.
The scope is outreach preparation: identifying potential contacts, forming relevance hypotheses, and drafting a starting point for a message. The user still decides whom to contact and what to send.
- Lovable
- AI-assisted ranking
- B2B outreach
- Sales enablement
What I would validate next
The intended benefit is less time spent assembling context before outreach. I would test that claim against the existing workflow rather than treating a fast demo as proof.
- Prospect quality
- Check suggested names and roles against current sources. Compare the ranking with a GTM operator’s assessment of account relevance.
- Time to a usable shortlist
- Measure preparation time with and without the tool, including verification and corrections.
- Message usefulness
- Assess how much editing each opening line needs and whether its claims are supported.
- Outreach outcomes
- In a controlled pilot, compare response quality and reply rates while accounting for differences in target accounts and offers.
These are proposed validation steps. No measured time savings or reply-rate improvement is claimed.