Skip to content
SaaS Yacht Club

Topics / AI SDRs

AI SDRs: where agents help outbound, and where they don't

AI SDR products promise to research, write and send outbound on their own. Members compare notes on which parts of the job agents handle well today, and which still need a person in the loop.

In the club

Counted across 255,431 messages from 1,844 members. Nobody is named or quoted.

Rising threads of conversation

  • Context and memory for agents

    20.4/1k now vs 12.8/1k before

  • Browser agents for GTM work

    27.8/1k now vs 18.5/1k before

What people ask for help with

  • Messaging, copy & personalization1100%
  • AI & agent reliability / setup700%

Share of 497 sampled questions

Why it matters

Buyers notice generic, machine-written outreach. Teams that get results use agents for research and drafting at scale while keeping targeting, offers and final judgment human.

What practitioners should know

  1. 01

    Research scales well

    Agents are good at gathering account context and drafting first versions quickly.

  2. 02

    Judgment still matters

    Choosing who to contact, what to offer and when to stop are where fully autonomous setups tend to struggle.

  3. 03

    Watch the brand risk

    High-volume automated sends can burn domains and reputations faster than a human team.

Worth reading

Research gathered Sep 29.

Questions to be able to answer

  • Which part of your outbound would you trust an agent to run alone?

Public threads on this

Club numbers as of Oct 24. Words are reviewed before they go live. Links go to their publishers.