I ran a $2M raise through Claude. Here is what happened.
Here is what fundraising work looks like in 2026.
Read time: 6 min
Last Saturday, I told you a series was coming. The one where I show you the AI fundraising stack I have been using for the last 4 months.
This is edition one.
I am going to walk you through one workflow, end to end, with one tool.
Not because I think you should buy it. Because I want you to see what fundraising looks like now.
If you are still doing this manually, you are not slow. You are a generation behind.
The tool is VC Boom. I wrote about it briefly two weeks ago. Today I am going deeper.
This weekend, I sat down with a friend prepping a $2M climate hardware seed raise.
We ran her actual deck through the tool. What came out was so much sharper than what most founders ship that I have been thinking about it ever since.
Here is what it looked like.
Step 1. Score the deck
Most founders never get an honest read on their deck before they send it.
You know who does? The investor reading it for the first time on a Saturday morning. By the time they form an opinion, you have already burned the introduction.
VC Boom starts here. Upload a PDF. Get a score across 7 dimensions in 90 seconds.
Market. Team. Traction. Ask. Problem. Moat. Storytelling.
Her deck scored 67.
Market and Ask came back at 82. The easy ones. Most founders nail them because the data is external.
Traction came back at 42. The flag was specific.
“Numbers shown but no momentum narrative. Investors cannot feel acceleration from a single bar chart. Add MoM growth and retention.”
That is the slide. That is always the slide.
I have looked at hundreds of decks in 8 years of investing. The slide that kills 7 out of 10 raises is the traction slide. The failure mode is exactly what the tool flagged. Numbers without narrative.
The score is not the magic. The diagnosis is.
If you do nothing else, score your deck before you send it. You will see what investors see in the first 90 seconds.
Step 2. Match the investors
This is where most founders waste the most time.
Standard playbook. Open Crunchbase. Search “climate seed”. Export 400 names. Spam.
Reply rates: 4 to 6%. Time to first meeting: 6 to 8 weeks. Half the list is wrong fit. The right ones are buried.
The matching engine cross-references your deck against 8,600+ investor profiles. Stage. Vertical. Check size. Recent deal activity in the last 18 months.
For my friend’s company, it returned 23 strong matches.
Not 400. Twenty-three.
She recognized 14. The other 9 were funds she had either never heard of or had not realized were active in her exact thesis.
One was a family office that had quietly written 4 climate hardware checks in the last 12 months. Not on any list she would have found manually. The kind of LP that surfaces only when something is grinding through the data underneath.
This is the part that is actually new.
The old fundraising stack was about volume. The new one is about precision. You do not need 400 investors. You need the 20 who are predisposed to say yes.
Speed is not the point. Coverage is.
Step 3. Write the cold email
This is the part I was most skeptical about.
Every AI email tool I have tried in the last 2 years failed the same way. The output sounds like AI. Investors smell it. Reply rate craters.
I tested VC Boom on the toughest case I could think of. Cold email to a Tier 1 climate partner my friend has never met. Opening with a reference to one of their recent bets.
What it produced was not a template. It was a 4-line opener that quoted a specific essay that partner had published 3 weeks ago, mapped it to the company’s thesis in one sentence, then pivoted to a single traction stat.
Not “I noticed your interest in climate tech.”
Specific. Recent. Provable.
78 words. No “synergy”. No “leverage”. No closing pleasantries.
Thesis. Hook. Ask.
VC Boom publishes a 3.8x reply rate vs cold templates. I cannot verify their dataset. The output quality is consistent with that order of magnitude. I have seen worse cold emails from human founders who spent 2 hours writing them.
The thing nobody tells you about cold email in 2026. It is now beating warm intros at pre-seed and seed. Reply rates flipped 6 to 9 months ago.
The reason is brutal. Warm intros got commoditized when everyone started using their network the same way. And good cold emails started looking nothing like cold emails.
Personalization at scale used to be an oxymoron. It is now a workflow.
Step 4. Coach the raise
This is the feature I did not expect to use, and now use most.
It is a chat interface. It already has your deck loaded. Your matched investor list memorized. 500+ scored decks as reference material.
You ask it questions about your specific raise. It answers in the context of your specific raise.
I tested it with the question every founder is afraid to ask out loud.
“My traction is weaker than my market story. How do I structure the slide so the partner does not lose interest before page 8?”
It did not give me a generic answer.
It pulled the specific traction numbers we had uploaded. Compared them against the median traction profile of seed-stage climate hardware companies in the dataset. Walked through a slide restructure that led with the contract pipeline rather than current MRR. Because pipeline was where the story was strongest.
A consultant would have charged $400 an hour for the same exercise. And the consultant would not have had the dataset.
If you have ever been mid-deck-revision at midnight, alone, wondering whether your moat slide is good enough, this is the answer.
What this actually is
Pull back. Look at the four pieces.
A deck score. A targeted investor list. A personalized cold email. A coach that knows your raise.
That is the entire workflow of a fundraising consultant. Compressed from 6 weeks to 2 hours. Priced at $297 lifetime instead of $30,000 retainer.
I am not being dramatic. The math is what is forcing the change.
Founders raising in 2026 against founders using this stack are not in the same fight.
The hyperscalers spending $725B on AI infrastructure this year (as I wrote about Friday) made their bet. Founders treating AI as fundraising leverage are making theirs.
The middle option. The “I’ll just use ChatGPT a bit.” The one most founders are sitting in right now.
That is the trap.
ChatGPT is a productivity hack. VC Boom is a workflow. The difference is whether you are saving 20 minutes on an email or compressing 6 weeks of work into 2 hours.
I am partnering with the team to bring this to my audience because the math is too good to keep quiet.
The reason I am writing this is simpler. I have watched too many great climate founders lose to bad fundraising processes. Real traction. Real teams. Real markets. Killed by 6 weeks of unfocused outreach.
What to do this week
If you are raising or about to raise.
Score your deck. Free. 90 seconds. You will learn something. Start here.
If the score reveals a real gap (most do), fix it before sending another email.
Promo code CLOSE20 takes 20% off for Climate Insiders readers. VC Boom.com
If you are not raising right now, save this email. You will need it within 6 months.
—Yoann
Quick note: I run a small number of brand collaborations each quarter with companies, tools, and VCs aligned with the climate audience here. VC Boom is one of them. Always disclosed. Always tools I use myself. If your team wants to reach 20,000 climate decision makers, you can see past campaigns and results here.
Super cool approach. Thanks for sharing
Yeah ChatGPT is kinda crap for cold email outreach. Never really gets specific about investors and there's no differentiation. But this tool must pool from ChatGPT (or Claude?) no? How do you ensure it's much better at it