# Marcus Delgado, Founder/CEO at Stagehand (B2B SaaS, 6 people) — read of LexiRisk, June 17 2026

> 8 years building software products, 3 as a solo operator before hiring. Currently drowning in vendor agreements and a reseller contract that showed up last week that my lawyer quoted $800 to review.

## How I got here

Googled "how to read a vendor contract without a lawyer" after getting that $800 quote. LexiRisk showed up third or fourth in the results, maybe fifth. I clicked it because the title in the SERP wasn't trying to rank for "best contract review software 2026" like the others. It just said what it did.

## What I clicked first

"Lawyers charge $300+/hour. LexiRisk reads your contracts in seconds, flags the risky clauses, and explains them in plain English."

That first section is the whole pitch in two sentences and it's pretty clean. I didn't feel like I was being sold something. I kept reading, which is more than I can say for the four competitors I closed without finishing the hero.

## Where I paused

The four bullet points under "What LexiRisk Catches." Specifically: "IP ownership buried deep in appendices or side clauses, quietly transferring your rights."

That one stopped me because I literally had that happen. Signed a white-label reseller agreement in 2023 and didn't notice a clause in Exhibit C that gave the vendor co-ownership of any integrations I built on top of their platform. My lawyer caught it six months later. Cost me a full renegotiation. The page named the exact thing I already got burned by, and it didn't feel invented.

## What I distrusted

"Join 2,000+ founders and freelancers who've used LexiRisk to catch bad terms before they become headaches."

2,000 is the number every landing page uses. It's low enough to be plausible for an early product and high enough to sound legitimate, which is why it reads fake. No timeframe. No "since March." No quote from a single named person. Not even a first name and a city. I've seen products ship with 40 real users and show their faces. This has nothing.

The 94% accuracy claim also sits weird. "Benchmarked against human contract reviews" -- whose? What jurisdiction? NDAs? SaaS MSAs? Construction contracts? A 94% catch rate on California tech contracts and a 94% catch rate on employment agreements with non-competes are very different numbers. I don't actually know what they're measuring.

## What would convince me

One real testimonial from a named founder with a company I can look up, saying specifically what clause the tool caught and what it saved them. Not "LexiRisk saved me thousands" -- something like "caught an auto-renewal trap in a $4k/year SaaS contract I was about to sign, the clause was on page 9, I never would have seen it." That's the kind of thing I'd forward to my co-founder.

A screenshot of the actual report output would also close a lot of the gap. Right now I have no idea what "clause-by-clause summaries" looks like in practice. Is it a wall of paragraphs? A clean table? Does the risk color-coding make sense at a glance? The page describes the output but shows me nothing.

## What I'd ask in an email reply

1. What contract types does the model perform worst on? I'm specifically thinking about licensing agreements and reseller contracts with a lot of exhibits -- does accuracy hold up there?

2. The report says I can add notes and share with my lawyer. Does my lawyer see the original contract or just the LexiRisk summary? I want to know if this actually speeds up my lawyer's review or just gives me more confidence walking in.

3. Has anyone used this for investor SAFEs or term sheets? The page mentions "investor term sheets" once but doesn't say anything about how well it handles those specifically.

## Verdict: curious-enough-to-reply

The pain is real, the pricing is low enough that I'd try it on the reseller contract sitting in my inbox right now even if the product is mediocre. The free trial with no credit card removes the friction. I'm going to upload a contract before I reply to anything.

---
*Memo by skeptic persona, generated 2026-06-17. Studio breaks own self-grading loop.*