Hilead now runs inside Claude.
Tell Claude who you want to reach:
"Find me founders of US startups that are hiring right now."
One prompt. You get a list of qualified prospects with:
→ Founder, company, headcount, location
→ Open roles and the hiring signal, with its date
→ LinkedIn profile for each one
Claude finds them. It never sends anything. You stay in control.
To celebrate the launch, your first leads are free: up to 750 with the trial.
Comment "API" below and I'll send you the setup. 3 minutes, no API key needed.
HubSpot put its co-founder back in the CEO seat on November 12. He calls it a refounding.
Brian Halligan started @HubSpot with Dharmesh Shah in 2005 and ran it for 16 years. Under Yamini Rangan, who took over in 2021, HubSpot grew from 135,000 to more than 311,000 customers.
Halligan plans to reimagine the company from the ground up. He describes the next chapter as building a "customer relationship harness", and the release gives no detail yet on what that means for the product.
At @salesforce , co-founder @Benioff has been CEO since 2001.
When you pick a CRM, you also pick whoever decides what it becomes. At HubSpot and Salesforce, that person is a founder.
Sources: HubSpot press release (Nov 12, 2026), Salesforce 2026 proxy statement (Apr 16, 2026)
Which one is your team on? Or something else?
I asked Fable 5.2 for a demo video of Hilead. One sentence.
It found our brand files, wrote the animation in code and rendered this. I changed nothing.
The video shows the other thing it does for us, the one that matters more.
You ask Claude for 50 Heads of Sales in France with a buying signal.
Hilead finds them, each with a reason to write today. A hire, a raise, a new role.
Every DM is written from that signal. One fact, one question.
You read them. You launch.
The model writes. The signal is what makes it worth reading.
What would you ask it for first?
By the time you read the funding announcement, the sales roles were already open.
I went back to my October list, the 10 AI startups in GTM that raised last month, and checked their careers pages.
Melius raised a $20M Series A on Oct 6, earmarked for its go-to-market and sales team. The SDR and AE roles went up on July 2.
Azoma announced $11.2M on Oct 22. Its Enterprise AE role had been open since Aug 7.
Broccoli AI closed $30M on Oct 28 to hire engineers and salespeople. Head of Enterprise Sales, Head of RevOps and an Enterprise AE were posted days before the news.
Othello raised $8M on Oct 14. 4 of its 5 open roles are GTM: founding AE, senior founding account manager, founding business development, solutions engineer.
Numeracle ($28M) and Cuttable (A$20M): no sales roles I could find.
So the sequence is the opposite of what most outbound playbooks assume. The hiring comes first, and the announcement follows weeks later. If you wait for the press release, you are starting weeks after the people who watched the careers page.
4
5
148
Everyone emails the CEO the week of the raise.
I write to someone else, on a different day.
A company announces an $18M Series A on Monday. By Tuesday the CEO's inbox is full of "congrats on the raise." Skip it.
Look at the careers page instead, find the 4 SDR roles that just went up, and write to whoever runs that team about what breaks when four reps ramp at once.
That's one signal out of 50. I put all of them into one playbook, with what to send when each one shows up, and it's free.
If you run outbound, you know these:
→ You email a company the week of its raise, with everyone else, and hear nothing.
→ A new VP of Sales shows up in your feed and the only thing you can think of is "congrats."
→ Your list has 2,000 accounts and you can't tell which 50 have a reason to talk this month.
→ You find out a prospect switched CRMs three months after it happened.
50 signals in 5 groups: people moved, company is scaling, said it out loud, stack in motion, timing you already own. For each one, how long it stays warm and what to send.
How to get it is on the image.
6
116
Top 10 AI startups in GTM that raised in October 2026 (w/ round size) 💰
1. Whop · platform to sell and distribute digital products · $200M Series D (Oct 21)
2. broccoli AI · AI agents that answer calls and book jobs for home-service businesses · $30M Series B (Oct 29)
3. Numeracle · stops business calls from showing as "Spam Likely" · $28M (Oct 28)
4. Melius · AI platform for ad campaigns, images and video · $20M Series A (Oct 6)
5. live2ai · AI agents for social media marketing · ~$15M pre-Series B (Oct 27)
6. Cuttable · AI ad creation for performance marketing teams · A$20M Series A (Oct 21)
7. Azoma · gets brands recommended by AI shopping assistants · $11.2M Series A (Oct 22)
8. Omnichat · AI marketing and sales agents on WhatsApp, LINE and Messenger · $10M Series A (Oct 27)
9. Othello · real-time AI coaching on live sales calls · $8M seed (Oct 14)
10. Dapta · AI sales agent that answers and qualifies inbound leads for SMBs · $6M seed (Oct 13)
Sources: LinkedIn + Hilead MCP (company, HQ, funding)
What October tells me:
Whop is the outlier, a commerce platform rather than a GTM tool. Take it out and the month is $141M across nine rounds, none above Series B. The money went to AI that talks to customers (calls, chat, social), not to another data layer.
Numeracle is my favorite one. $28M so that your sales calls stop landing as "Spam Likely". Nobody writes posts about the phone anymore, and it just got funded.
A round in October usually means SDR and AE roles open by January. If you sell to GTM teams, that's your window, not the week the press release turns three months old.
If you sell to this list, which one would you reach out to first? #startup
1
3
87
Mathieu retweeted
How is this not viral
Asked Claude to make Hilead’s product video last night. @raylightapp was the editor.
Woke up to this.
No designer, no After Effects. 50 seconds 👇
43 startups raised money last week. Every one of them now has a budget they didn’t have two weeks ago.
The biggest rounds (US + Europe, Oct 19-25):
Tenstorrent · AI chips · $1.5B Series E
Flapping Airplanes · one-year-old AI lab · $1B Series A at a $6B valuation
Legora · legal AI · $550M Series E at an $8.8B valuation
ASI · AI for critical infrastructure · $325M Series C
Xona · GPS alternative via satellites · $250M Series D
Pigment · enterprise planning · $250M
Pathos AI · AI drug development · $237M Series E
Warp Sciences · AI + robotics for drug labs · $215M Series A
Whop · platform for online businesses · $200M Series D
Star Therapeutics · biotech · $200M
Warhound · engineered military dogs · $200M seed
Over $5B in a single week, and these are the ones already getting coverage.
The other 32 range from a €540K pre-seed to a $100M Series C. I put the full list in a doc with round type, lead investors and location.
Why this matters if you sell B2B: a fresh round usually means a hiring plan and a VP who needs pipeline by next quarter. The best time to reach out is now, not when the press release is three months old.
And if you’d rather not wait for next Monday: funding is one of the 26 signals Hilead tracks, with hiring and new exec signals on top, and the right person to contact at each company.
Comment “startup” and I’ll send you the full list 👇
Raised $18M on Monday.
Hiring 6 SDRs on Tuesday.
New VP Sales on Wednesday.
3 companies just told you they're ready to buy.
Usually someone else gets there first.
So I built a radar for it:
→ it watches 26 kinds of buying signals (hiring, funding, new role, product launch…)
→ it finds the right person inside the company
→ it writes the first message from the signal itself
You end up with a list of people who have a real reason to talk to you.
To run it on your own market 👇
I'm Mathieu 👋
Building Hilead, born in France, based in Kuala Lumpur 🇲🇾
Looking to meet founders, early users and marketers!
Just put myself on the map 👇
makermap.lol/mathlleu?v=whl8…
teapawt.org/verbove/status/2108611…
3
9
185
Your 1,000-contact list replies at 2.1%. Cut it to 50 and it replies at 5.8%.
Nearly three times the replies. Woodpecker's compilation, built on Belkins' 16.5 million emails and their own platform data. Every bucket in between slopes down: 50 to 200 contacts, 4 to 5%. 200 to 500, about 3%.
Nobody writes a better email because the list is short. The short list is the result of something else: you picked those 40 accounts for a reason, and the reason is in the first line.
The 1,000-contact list exists because the tool made it easy. Titles, company size, export. The reason to write today is missing, so the first line is about you.
Run the math on 300 replies. At 2.1% that is 14,000 emails and a sending setup to survive it. At 5.8%, 5,200. Same replies, a third of the list, no warm-up farm.
7
120
Belkins tracked 7,530,489 cold emails in 2025. 34,393 got a reply. That is 0.45%.
Replies divided by emails sent, auto-replies and bounces excluded. Their own words: dramatically lower than the figures they published in previous years.
The softer counts fall too. Woodpecker's platform average went from 5.1% in 2024 to 3.43% last year, which is also Instantly's figure. Same direction, different denominators.
Three things changed in the inbox at once. AI made a sequence cost nothing to send, so everybody sent one. Gmail and Outlook started enforcing the bulk-sender rules. And the people reading stopped answering anything that looks like it went to 1,000 others.
What still gets replies in that inbox is the email that could only have been sent to one person. A hire they announced. A post they wrote. A signal.
Which number does your team report, replies per email sent or per opener? The gap between the two is the whole story.
5
56
58% of cold email replies come from the first email. The other three share the remaining 42%.
Instantly's 2026 report, billions of sends between January and December 2025. Woodpecker lands on the same split.
Here is the part people get backwards. Follow-ups still double the total: 4.1% replies with no follow-up, 8.3% with three to five. And 48% of reps never send one.
So both are true at the same time. The first email does most of the work. The follow-ups are the difference between 4 and 8.
Where the hours go, then:
Write the first email like it is the only one they will read. One reason you are writing today, in the first line. A hire, a post they wrote, a signal.
Make step 2 read like a reply to your own email. The formal "just bumping this" gets about 30% fewer replies, by Instantly's count.
Stop between 4 and 7 touches. After four follow-ups, complaints and unsubscribes triple.
On the reply-rate ladder I published this week, the whole gap between 4.4% and 10% sat in that first line.
How many steps in your current sequence, and which one actually gets the replies?
1
5
60
I read 22 cold email reply-rate benchmarks this week. All 22 were written by vendors.
Including this one.
The number everyone repeats is 15 to 25% for signal-based outreach. I chased it. One tool puts it in the headline and writes 2 to 5% in the body of the same article. Another credits a study whose headline finding is 10%. Nobody has a dataset.
So here is the ladder with only the numbers I could trace to a sample:
3.43%. Average cold email, all senders. Instantly, 2026 Cold Email Benchmark Report. Woodpecker, cold email statistics 2026, 20M+ emails.
4.4%. Opener about the problem you solve. The Digital Bloom, 2025 reply-rate benchmarks.
10%. Opener about something that just happened at the account. Same study, same buyers.
12%. Our signal-based campaigns at Hilead. Smaller sample than theirs, and our list-based ones sit at 1 to 3 (same as everyone).
The gap I care about is 4.4 to 10. Same study. Same buyers. Only the first line changed.
What's your actual reply rate right now? Not the one in the deck.
Instantly.ai, Woodpecker.co, TheDigitalBloom












