The three eras of prospecting: manual, AI SDR, and AI-native GTM

Every so often a story comes along that makes the shape of a whole market obvious. This one did for us.

In The Signal, Brendan Short published an inside look at “ChatGTM,” the sales AI that Cursor’s Head of Enterprise Growth, George Hou, built for their 400+ person sales org. The numbers are hard to argue with: SDRs booking 3x qualified meetings, top reps at 4x, and AE ramp time cut by more than 50%. The rep quotes are even more telling:

SDRs describe not being able to live without ChatGTM at this point. Some half-joke that if ChatGTM has an outage, they might as well go home because they can’t effectively outbound at the same pace. - Brendan Short, The Signal

That is not what “another AI tool” adoption looks like. That is what a new era looks like. And the clearest way to see why is to line up the three eras of prospecting side by side.

Era 1: Manual prospecting

The rep opens seven tabs. A dashboard, Salesforce, Gong, LinkedIn, and a couple of others. They spend 45 minutes aggregating data, then sit and stare at a blank canvas trying to figure out how to reach out. They send 10 emails. They get 0 replies.

This is the era most teams still live in. The work is real, but almost all of it is assembly: gathering context that lives in ten places, reconciling it in your head, and only then getting to the part that actually matters, the message. The research tax is paid per account, by hand, every single time.

Era 2: The AI SDR

The obvious fix was to automate the sending. Point an AI SDR at your market and let it run: 1,000 emails, no context, no champion, no judgment. Spray and pray, at scale.

It burns the list. With no real context behind them, the messages all read the same, and buyers learned to tune them out. As we wrote in The AI SDR flooded your inbox, when the cost of sending drops to zero, everyone sends more, the channel fills up, and reply rates go down, not up. Era 2 didn’t beat the noise floor. It raised it.

Era 3: AI-native GTM

ChatGTM is what the third era looks like. The champion is identified overnight. Context is pre-loaded from the systems that already hold it. The outbound is drafted before 8am, each message tied to a real signal, and the rep just reviews, tweaks, and sends. The result is 3x meetings, and a leaderboard that keeps score.

The difference from Era 2 is the whole point: AI does the research, the human keeps the relationship. As Hou put it, an agent without the right context won’t magically solve anything, most of the work is figuring out what context it actually needs. Era 3 isn’t more automation bolted onto Era 2. It’s AI aimed at the part of the job that was always the bottleneck, and a person left in charge of the part that closes.

The three eras at a glance

	Era 1: Manual	Era 2: AI SDR	Era 3: AI-native GTM
The research	45 min, seven tabs, by hand	Skipped	Pre-loaded overnight
The champion	Guessed at	Ignored	Identified before you log in
The outbound	Blank canvas at 9am	1,000 generic sends	Drafted before 8am, tied to a signal
The human	Does everything	Watches a dashboard	Reviews, tweaks, sends, closes
The result	10 emails, 0 replies	Burns the list	3x meetings

You don’t have to be Cursor to get there

Here’s the honest part. George would be the first to tell you that most GTM teams shouldn’t build their own ChatGTM. Cursor had a data warehouse, a data team, a build-over-buy culture, a product research engineer who’s built search engines and browsers, and hard-won context-engineering muscle from building Cursor itself. Even then, the core took two months on top of a pile of earlier failures.

Almost no one has that on the shelf. But the era it points to isn’t reserved for the labs. The pattern underneath ChatGTM is portable: go to the systems of record on demand, pull the right context, surface the champion and the reason to act, and draft the outreach for a human to send. That’s a platform decision, not a two-month internal build.

How Alfa fits

Alfa is Era 3, without the internal build. You describe what you sell, and Alfa turns market movement into a live stream of accounts, likely champions, and reasons to act, doing the research, enrichment, and context-building overnight so the work is waiting for you in the morning. It pulls from where your data already lives instead of asking you to maintain yet another repo of context, which is the same collapse of the tooling layer we described in what comes after Clay.

What Alfa deliberately leaves to you is the message and the relationship, because that is exactly the part Era 3 protects. Cursor proved the ceiling: 3x meetings, ramp cut in half, a team that can’t imagine working without it. You don’t have to be Cursor to work like this. You just have to start in Era 3.