
@nifineti
iAccount based inBelgium
About this account
- Account based in
- Belgium
- Connected via
- Belgium App Store
Account-level information from X, not a live location or the device used for a specific post.
🔥🔥
Just did a cold call session on the warm intent list
80% connection rate (I dial twice every time).
15% meeting booked 😳
this is strooooong
getting a phone number usually means export, enrich somewhere else, import again.
here's the phone playbook in YourMax.ai, step by step, without leaving the Leads page:
step 1: pick the leads worth calling
- start from your best ICP fits
- the Phone column shows "Reveal · 5 credits" on each lead
step 2: one lead or a batch
- reveal a single number from its row
- or select several leads and click Find phones
step 3: check the cost before it runs
- the dialog tells you upfront: 5 credits per number found, nothing when none is
- and how many credits you have left
step 4: let it run in the background
- YourMax.ai reveals the numbers it already holds and runs its enrichment waterfall for the rest
- numbers appear in their rows as they're found
step 5: automate it for new leads
- turn on Auto-enrich: it finds phones, or email and phone, for new leads above an ICP-fit threshold you choose
- it's off by default and only applies to new leads sourced while it's on
one setup note: you only pay when a number comes back. no result, no charge.
bookmark this if you want to set it up later.
Damn, X just paid me 533$ for the airport laptop guy!
9.8m views... about 5 cents per 1K.
Saw someone get 2.6k$ for almost the same reach (9.6m).
Turns out X only pays for views from Premium accounts, and apparently my viewers are not on Premium haha
Back to shipping 🚢
(there's also a totally not viral, but super useful article I put together recently - link in replies)
Karpathy put numbers on the AI gap.
0.2% of people are watching software build itself.
The other 99.8% still think the hard part is the prompt.
That gap is the biggest go-to-market opportunity I've seen in 12 years.
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
getting a phone number usually means export, enrich somewhere else, import again.
here's the phone playbook in YourMax.ai, step by step, without leaving the Leads page:
step 1: pick the leads worth calling
- start from your best ICP fits
- the Phone column shows "Reveal · 5 credits" on each lead
step 2: one lead or a batch
- reveal a single number from its row
- or select several leads and click Find phones
step 3: check the cost before it runs
- the dialog tells you upfront: 5 credits per number found, nothing when none is
- and how many credits you have left
step 4: let it run in the background
- YourMax.ai reveals the numbers it already holds and runs its enrichment waterfall for the rest
- numbers appear in their rows as they're found
step 5: automate it for new leads
- turn on Auto-enrich: it finds phones, or email and phone, for new leads above an ICP-fit threshold you choose
- it's off by default and only applies to new leads sourced while it's on
one setup note: you only pay when a number comes back. no result, no charge.
bookmark this if you want to set it up later.
Took me 10 years as a founder to get this.
McGregor is the funding announcement.
Khabib is doing the boring work every day until it works.
This quoted post is unavailable.
3 kids here. Same lesson.
The hard stop didn't cost me anything.
It just killed the meetings that never mattered.
I stop working at 4:30 PM every day to be with my 14-month-old.
Could I fit in another 2-3 hours every day? Absolutely. Another call, another deliverable, another thing that could move the business forward.
But Palmer is 14 months old. She's changing by the week. New words, new expressions, new things she suddenly understands. None of that happens twice.
This isn't a productivity strategy. It's a trade I'm making on purpose.
The business might grow a little slower because of it. I can live with that. There will always be another way to make more money. There won't be a way to buy back a random Tuesday afternoon when she was 14 months old.
Last year when she was a newborn, I blocked 3-5 hours every day for her. That same month I hit $36,825 in revenue at 86% margin. This year, we are double that.
When your day has a hard stop, you get ruthless about what fills it. You stop sitting in meetings you don't need to be in and approving things your team can handle.
Agency owners protect client deadlines with everything they've got. Then let every personal boundary slide without a second thought.
Don't let your business take more from you than it should.
been running a marketplace for 12 years that's literally this
humans pick the prettiest profile on page 1
agents gonna read all 400 case studies at 2am and pick the one that actually fits
kinda terrifying. also kinda fair ngl
The minimum viable company is changing.
Starting a company used to mean forming the entity and opening a bank account. The Internet added a website and online payments.
As more discovery and purchasing moves to agents, the company itself becomes something software has to understand and act on.
An agent should be able to understand the business well enough to decide when it is relevant and complete the transaction.
Founders are going to have to design for this from the beginning. I think that changes what it means to start a company.
the wispr flow team is just cooking right now
first the paid ugc playbook
now a full 0 to 100m views/month guide
imagine what they're not sharing yet
taking notes on every single one
the next buying signal isn't a human on your pricing page.
it's their agent comparing you to 4 competitors at 2am.
whoever learns to read agent visits first eats everyone's outbound lunch.
Some interesting machine & agent Vercel traffic stats.
• Across the entire Vercel network, 58.18% traffic is bot-originated (last 30d). This was 32% in Jan 2024.
• 60%+ of deployments on Vercel are now agentic, up from ~3% in Jan 2026.
• Up to 83% of pageviews are now coming from agents to Vercel's own documentation sites. The % reliably increases the more we optimize content for them.
I expect 'direct' human internet traffic to basically become a rounding error in coming years. The web will thrive, but it will be built for and by agents.
the scariest competitor for a saas in 2026 is your own customer opening claude code on a saturday.
if your price isn't tied to the value, someone will vibe code the gap.
Bye bye Beehiiv 👋
I vibe-coded my own newsletter platform:
📧 Campaign management w/ editor
🖥️ Web archives (newsletter.marclou.com)
🔗 API support
🫧 List cleanup, deliverability reports
It took a few prompts with Opus 5.5.
I left Beehiiv because they raised prices for the 3rd time in 2 years.
- My first yearly bill was $1,000.
- My future bill would have been $4,000.
Now I'm using AWS SES + MongoDB, and I will pay less than $300/year.
Newsletter platforms charge per subscriber, when in fact the real cost is the number of emails sent. Subscribers will naturally become inactive, so you end up paying more for something that hurts your deliverability (sending to dormant subscribers = spam trap).
So I've set a cleanup cron job for my newsletter alternative. It marks inactive subs as dormant, sends one last re-engaging email, and stops emailing them. Money saved. Deliverability increased.
I also gain a priceless feature: Freedom. I can change the UI, add features, connect all my apps - anything is possible with just a prompt.
I'm replacing Resend next. I love it, but it's very expensive ($400/mo)
Imagine tu as le choix entre:
- try hard sur vendre et build du produit pour faire du MRR
- try hard d’économiser 99€/mois (3€/jours)
quel est le meilleur ROI?
Replying to @EddyLeBough
alors non l'idée c'est pas d'envoyer des mails de prospection avec le mail de grok bot "@mail.grokbot.com" mais de créer des comptes Apollo et Hunter en illimité pour profiter à chaque fois des 14 jours d'essai
Most bad AI outbound comes from one thing: the model can't see what you know.
I turned Claude into a prospecting analyst with five prompts (in the article below). Here's what we learned to tell it, so the emails come out right.
If it makes things up:
1. Give it the actual page, the job post, the funding announcement. Don't let it work from a search snippet. That's where the congratulations email for a round nobody raised comes from.
2. Ask it to put the source next to every fact. If it can't point to one, delete the fact.
3. Keep two dates on every fact: when it happened, and when you found it. A page you found this morning can be about last spring.
The job post in the article had no date at all, so part of the hold was checking the role was still open.
4. Tell it what it's not allowed to guess. Ours can't guess budget, urgency or whether someone is ready to buy from a single event. A funding round that says nothing about sales is a weak reason to pitch sales.
If every email could go to anyone
5. Give it the reason you'd matter to them this week. Twenty cold emails a day can name the city I posted from. Not one of them knows if I have a problem.
6. Be exact. Which role they reposted, how much they raised, what the tender actually says. "They fit our ICP" gives it nothing to write with.
7. A test I like: cut the line about what you found and read the email again. If it still sounds like a decent email, send it. If it falls apart, that one line was doing all the work.
8. Keep a short list of things it may never write, next to the instructions. A model can follow every tone instruction and still write the exact email buyers delete on sight.
Ours started as four lines: no buzzwords, no em dashes, no fake urgency or "circling back", no "Hi {{firstName}}". It's longer now, and it does more than the instructions do.
If it pitches the wrong person
9. The person in the news isn't always the person who buys. A company posts its first SDR role. The person they hire is nobody to you.
Whoever runs the sales team is probably who you want, and even that stays a guess until someone has checked. Never let it make up a name.
10. Ask it to make the case against the lead in the same answer. A company hiring its first outbound rep looks like the perfect lead for outbound consulting.
It's also a company that just decided to build it in-house. My call on that one was hold.
If it opens with what it found
11. "Saw you raised your Series B" tells them a tool found it. Counting their pricing-page visits out loud is worse. What you found sets the timing and the topic. The first line is about something they did in public.
If it says yes to everyone
12. Let it say "not yet". Ours says hold when the fit looks right but something still needs checking, and nothing goes out until a person has checked it. A reason to stop is a useful answer.
13. Keep the first email small. 70 words at most: one thing you can back up, one thing you can actually send, one simple question. If what you're sending is a blank template, call it a template.
If it never gets better
14. Log the reason behind every no, in one line. "Wrong quarter" is a date to come back. "Already with a competitor" is a different angle later. "Not the buyer" is a contact to fix. Same word, three different next moves.
15. When the answer is wrong, change what it can see before you rewrite the instruction. Anthropic's own prompting video gets Claude from a wrong insurance ruling to the right one without really touching the instruction. Most bad answers are missing a fact.
All five prompts are in the article below. Copy them as they are.
(bookmark for later)
You can turn Claude into a prospecting analyst with five copy-paste prompts, a browser and a spreadsheet.
> Start with your offer and one buying signal.
> You get a shortlist with sources and a reason behind every account.
> The ones ready for outreach also get a first email.
In the article I walk through a real job post. It's a company hiring its first outbound rep to work on account selection and the prospecting playbook.
If you sell outbound consulting, that looks like the perfect reason to reach out. It may also be the best reason they won't need you. They're hiring someone to build it in-house.
So my call was HOLD, until I'd checked who owns that decision and whether the job post is still current.
The prompts ask for that kind of objection up front. Run them in order, each one feeds the next:
1. The signal brief
Give it your offer, who it's for, who usually buys and who's a bad fit.
It comes back with:
- three changes you can verify
- the decision the company may now face
- the strongest reason your offer might not be needed
It isn't allowed to guess budget or urgency from an event.
2. The search plan
It turns the brief into search phrases and the original pages to open.
For every candidate, save the URL, the exact excerpt, when it happened and when you checked.
A search snippet never counts as proof.
3. The account brief
One account at a time. It gives you:
- facts, each tied to a source
- the decision they might be facing
- the evidence against it
- the buyer, kept as a guess until you verify them
Then ADVANCE, HOLD or PASS. No message for a HOLD or a PASS.
4. The first email
Only for ADVANCE accounts.
Up to 70 words: one observation you can back up, a resource you can actually send, one simple question.
Then it maps every factual claim back to its source.
5. The weekly review
Log why you passed on each account: wrong company type, stale source, wrong person.
If your search keeps returning agencies and you sell to software companies, fix the search.
Claude hands you a revised signal brief for the next round.
For your first run, pick one offer, one signal and ten accounts.
Once it's a weekly habit, max can take over the collecting.
It's the signal side of Overloop AI, which we're building.
For job posts, max watches careers pages and job boards.
I put all five prompts in the article below, ready to copy.
(bookmark for later)
SAP buys TechWolf in record Belgian VC-backed deal tech.eu/2026/10/06/sap-buys-…





















