Jack Taubl
Director of Sales at Ocean.io
On Pipeline
-
Why Does AI Prospecting Get Worse the Longer You Use It?
AI prospecting gets worse in long chats for two reasons: models are built to compact context, and vague prompts drift. Jack Taubl, Director of Sales at Ocean.io, explains both on Pipeline, along with his fixes. He works inside Claude…
Key takeaways
- Do open-ended prospecting work inside a Claude project, where each new chat carries context from the others, instead of one long thread that gets compacted.
- When Claude gives an answer you don't trust, open a separate chat in the same project and consider a different reasoning model, as Jack Taubl does.
- Replace prompts like "What does a good XYZ look like?" with specific requests, because ambiguous questions get ambiguous results.
- Ground a proposal by adding the prospect's most recent 10-K to the Claude project's files.
- Schedule a daily briefing to run before your day starts. Jack Taubl runs his at around 4:00 a.m.
In Jack's words
"it's fundamentally designed to erode the longer you use it if you're staying in a single chat."
"if you ask ambiguous questions, you're gonna get ambiguous results."
"I will immediately switch to a separate chat and potentially switch the reasoning model that I'm using as well"