Most teams think they know their ICP. They usually don't.
What it looks like on paper
The ICP is defined. It's documented. It sounds reasonable: industry, company size, geography. Sometimes even job titles, use cases.
So outbound starts. Campaigns go live. Data gets pulled. Messaging gets written.
And then performance doesn't hold.
Some replies come in. A few meetings get booked. But it's inconsistent. Hard to scale. Hard to predict.
So the assumption becomes: "we need better messaging," "we need more volume," "we need to test more angles."
The issue isn't execution.
It's targeting.
Most ICP definitions are built on assumptions, partial data, what should work. Not on what actually converts.
A typical example
In one case, the target market included restaurants, catering companies, mobile food operators, hospitality venues. All of them could use the product. All of them made sense.
But not all of them convert.
Once campaigns ran, patterns started to appear. Some segments engaged but didn't move forward. Others had long, slow decision cycles. Some had interest but no budget. At the same time, a few segments consistently responded and converted.
That's where the real ICP is — not in the initial definition, but in validated behaviour.
What changed
Instead of expanding outreach, the system narrowed. Segments were reduced to those with clear use cases, operational need, ability to buy. In this case: catering companies and country clubs. Deprioritising first-time buyers and operators with limited expansion capacity.
The mistake most teams make
They define ICP once. Then they execute around it. And when performance is inconsistent, they adjust messaging, change channels, increase effort. Instead of questioning the ICP itself.
What ICP actually is
Not a static definition. Not a slide in a deck. Not a one-time exercise. A moving boundary based on real conversion.
The real process
- Start broad, but structured
- Run targeted outreach
- Observe response and conversion patterns
- Narrow aggressively
- Repeat until consistency appears
The takeaway
Outbound doesn't fail because it's hard. It fails because it's pointed at the wrong target.
Outbound doesn't fail because it's hard. It fails because it's pointed at the wrong target.