Quarantine Report
When everyone has AI SDRs, no one has AI SDRs.
The idea that anyone ever bought stock from a stranger who called them is hilarious to me. I’m sure it worked once. By the time I was paying attention, it was already a joke.
I suspect the generation going out on their own now will feel the same way about buying software from a stranger who emailed them.
In 1989, Jordan Belfort sat in a rented office on Long Island and dialed through a list of dentists. The pitch was penny stocks. The tool was a telephone. The math was simple: call enough people, and some of them will say yes.
By 1992, Stratton Oakmont had 1,000 brokers making 100,000 calls a day. The channel worked because it was novel. Most people had never received an unsolicited stock pitch before. The phone felt personal: a real human voice, saying your name, asking for a few minutes.
Then the dialers arrived. Then caller ID shipped. Then the Do Not Call Registry passed. Then your phone started labeling unknown numbers Spam Likely. The channel didn’t decline. It died. Defense caught up to offense, and now the only people still cold calling about stocks are the ones you see in documentaries.
I kept thinking about this while reviewing my fund’s email quarantine. Microsoft Defender caught 756 messages in the last 30 days. Sales tools. Lead lists. Outbound agencies. AI SDR platforms. Founders pitching their companies. Consultants pitching their services.
The irony: half of them were selling tools to send more emails.
Here’s one from last week, lightly edited:
“Joseph, not sure if it’s a technical issue on my end. Are you seeing my emails?”
This was the third message in nine days. There was no technical issue. The sender was testing whether fake confusion would generate a reply. The previous two emails offered to sell me a list of 5,000 pre-seed founders. The product was a list of people to spam. The delivery method was spam.
Here’s another:
“Would you like it if I bring some lunch to your office? What’s the best taco joint around your neighborhood?”
The sender had never met the recipient. The taco question existed to generate engagement. The actual pitch, buried in the thread, was for a lead generation service. The sender used two different domains across four emails in three days: one to send from, one to link to. Burn the outbound domain, protect the real one.
And another:
“Still happy to send the $100 Amazon gift card for the 15 minutes.”
This one was from an agency promising 100 qualified meetings per month for portfolio companies. The bribe was the tell. If the email had value, you wouldn’t need to pay people to read it.
The playbook hasn’t changed since Stratton Oakmont: volume, scripts, lists, metrics that reward activity over outcomes. The only difference is the medium.
In 1995, the tools were dialers and phone banks. In 2025, they’re Apollo and Instantly and Clay and a dozen AI SDR platforms that promise to “personalize at scale.” The pitch is always the same: more throughput, less effort, better targeting. The vendors show you the math. If 1% of 10,000 emails convert to meetings, you get 100 meetings.
They don’t show you what happens when everyone runs the same math.
When everyone has a dialer, no one has a dialer. When everyone has AI SDRs, no one has AI SDRs. The channel is a commons. The tools are the overgrazing. Every sender makes the channel worse for every other sender until the defense catches up and the window closes.
The phone window closed in the late 90s. The email window already closed. You just don’t see it until your sender score catches up.
My quarantine folder is the evidence.
Meanwhile, we keep meeting founders who never sent a cold email in their lives.
They spent years inside broken workflows. Fielding discovery calls. Sitting through demos. Explaining their problems to product managers who didn’t get it. Then the tools to build got cheap enough that explaining became harder than fixing. They jury-rigged something over a weekend because they couldn’t stop thinking about it.
When they went to sell, they didn’t buy a list. They posted in the group chat they’d been in for a decade. They texted the people who used to call them when something broke. They showed up at the same conference they’d attended for years, except this time they had something to show.
Their go-to-market looks unfair because it is. They have proximity. They were already in the room.
This is the flip.
The same forces that let everyone send 10,000 emails also let anyone build software. Outreach supply exploded at the same moment product supply exploded. More noise. More options. Less reason to open a cold email from a stranger when someone you actually know just built the thing you need.
The founders in my spam folder are playing a game where the scoreboard is the filter.
The winners aren’t better at outbound. They just don’t need it.