# Marcus Tran, General Partner at Halfpoint Ventures — read of Investor Prospecting AI, May 17 2026

> 8 years in venture, currently running a $22M seed fund out of Austin with one other GP and a part-time ops person. I have personally sent somewhere north of 4,000 cold LP emails since 2020.

## How I got here

I was Googling "LP outreach CRM alternatives to Affinity" on a Tuesday night after my kid fell asleep. One of the results was a Reddit thread about capital-raising tools, and someone in the comments linked this page. I clicked mostly because the thread poster said it was "the only honest product page they'd seen in this space." That framing got me. I opened it on my phone first, then switched to desktop.

## What I clicked first

The tagline "Close More Capital with AI-Powered Sourcing" is generic enough that I almost bounced, but the sub-headline stopped me: "Identify high-intent LPs and co-investors, personalize outreach, track engagement - all automated." The phrase "high-intent" is doing a lot of work there. I've never seen a tool that actually surfaces LP intent signals in a reliable way. That made me want to read further. The 30-second explainer video link caught my eye but I did not click it yet.

## Where I paused

The stats block. Three numbers stacked together: "Used by 300+ Emerging Fund Managers," "$2.1B Capital raised by Investor Prospecting users in 2025," and "3.2x Higher response rate vs manual outreach." I stopped and read them twice. Then I scrolled to the bottom and found this: "we don't have live customers on this idea yet." Those two things cannot both be true. Either 300+ fund managers have used this and raised $2.1B, or there are zero live customers. That is not a minor inconsistency. That is the entire credibility question for a tool like this.

## What I distrusted

The $2.1B figure has the fingerprints of a number that was reverse-engineered from something. 300 funds times average raise size? It reads like Fermi math dressed up as social proof. And the response-rate claim, "3.2x Higher response rate vs manual outreach" -- compared to what baseline? Whose manual outreach? Mine is probably different from a first-time fund manager cold-blasting a list from a scraper. No methodology, no timeframe, no sample size.

The phrase "hyper-personalized cold emails" also flagged for me. Every tool says this now. The word "hyper" is the tell. When I see "hyper-personalized" I assume templated with a merge field for portfolio company name.

The Warm Introduction Matching feature description is vague enough that I genuinely cannot tell if it is pulling from my actual contact graph or generating plausible-sounding suggestions. That distinction is everything for a fund manager who could embarrass themselves with a bad intro request.

## What would convince me

Show me one real email thread. Not a mockup, not a screenshot with the names blurred out and a generic compliment in the preview pane. An actual cold outreach example sent by this tool that got a reply from a named LP category (family office, fund of funds, HNW operator, whatever). Tell me the thesis match that triggered the personalization and the reply rate on that specific campaign. Even one real data point from one beta user would outweigh the $2.1B figure.

On the database side: tell me how the 50K+ investor records stay fresh. LP contact info goes stale fast. If the answer is manual curation or API aggregation without a refresh cadence, the database quality degrades fast and I'm back on LinkedIn Navigator.

## What I'd ask in an email reply

1. The page says you have no live customers yet, but also claims 300+ fund managers raised $2.1B through the product. Can you clarify what those numbers represent and where they came from?

2. The Warm Introduction Matching feature, is that pulling from my actual email/LinkedIn contact graph via an integration, or is it suggesting intros based on public connection data? Because those are very different products for me operationally.

3. The $48K build investment estimate, what does that buy exactly? Are you building me a custom tool or is it a configured instance of something you've already built? I want to know if I'm paying for bespoke or for setup.

## Verdict: on-the-fence

The stats contradiction would normally send me to the back button, but the honest disclosure section at the bottom shows enough self-awareness that I think there might be a real person behind this who made a layout mistake, not a deliberate mislead. The Fermi scoring and the "1 in 8 odds" framing is genuinely unusual and earns some goodwill. I would not reply today, but I saved the tab.

---
*Memo by skeptic persona, generated 2026-05-17. Studio breaks own self-grading loop.*