This post was originally published on LinkedIn.

The Human Middleware Problem

This is a timeless problem, where people are spending way too much time on: copy/pasting copy from tool to tool, moving data around in spreadsheets, armies of SDRs doing account research, etc

But, now more than ever this problem is creating a divide between companies.

The Leaders have AI and automation handle this type of busy work so their humans can do the type of work only humans can do. A virtuous cycle of AI handling more, humans being more effective and handing off more, becoming more effective, handing off more, etc etc

Third-party signals from unstructured data

AI agents can find buying signals that don't live in any database:

• Job postings mentioning they need to implement new systems
• Earnings calls where leadership talks about operational pain
• Negative reviews
• Press releases on new hires and their mandates
• Partnership announcements
• Expansion into new regions
• Hiring/restructuring
• etc etc

Only AI can find these at scale. The best signals are:

• The hardest to find
• The less likely you can source from a pre-existing database

1. AI prioritization models vs deterministic signal workflows

(is it either or? Or is it both?)

Some AI tools feed all your signals/context into a model that decides which accounts to surface to reps. You don't have much control.

Some tools enable you to setup deterministic: if THIS signal fires, guarantee THIS action happens, report on THIS signal outcome.

Should we do both at the same time OR just trust the AI Account prioritization OR only do signal plays?

We'll see.