The 3-step ICP evaluation I use before any outreach
Before I figured this out I was just DMing random founders on LinkedIn and X hoping something would stick. No system for tracking who I contacted, no idea which type of person was actually a good lead vs just someone who looked interesting. Complete mess.
I eventually realized the problem wasn't my outreach - it was that I had no clue WHO to target. So I built a simple scoring system and it changed everything.
- Step 1: List what your product actually solves
Sounds obvious but I skipped this for weeks. Write down your core features and
what type of person needs each one. This alone eliminates like half the people you think are leads but aren't.
- Step 2: Score each potential customer type on 4 things
For each group of people you think might buy:
- Pain (how bad do they need this? Nice-to-have or hair-on-fire?)
- Fit (can your product actually solve their problem well?)
- Business (can they pay? How fast do they decide? Can you even reach them?)
- Value (is the customer worth pursuing long-term?)
I weight pain the heaviest because if the problem doesn't hurt they won't pay no matter how good your product is.
- Step 3: Compare and pick ONE to test first
I evaluated 6 different customer types for my SaaS. Scores ranged from 56% to 82%. The gap between #1 and #6 was massive and I would've never seen it without scoring them.
The top one had high pain, was easy to reach, and made fast decisions. The bottom one had decent pain but was hard to find and slow to buy. Completely different outreach strategy needed for each.
Biggest lesson: I was wasting time on customer types that scored in the 50-60% range when there was an 82% sitting right there.
Anyone else been guilty of skipping ICP evaluation and just blasting messages to whoever? What made you stop and actually figure out who to target?
This is a really clean way to force focus and especially weighting pain the heaviest.
One thing I’ve found useful to layer on after scoring is a quick “urgency-in-motion” check:
Are people in this ICP already hacking together workarounds, escalating internally, or missing deadlines because of this?
Sometimes two segments score similarly on pain, but the one already acting (manual scripts, messy spreadsheets, Slack fire drills) converts dramatically faster than the one that just agrees the problem exists.
When you tested the 82% ICP, did you notice any clear behavioral signals like that before outreach?
I am still testing, haven't found any clear signals yet
I may have gotten the ICP right (maybe), but that doesn't mean the product is validated... A story for another time.
You "urgency-in-motion" check does sound useful, I need to add that to my process
Yeah, that distinction matters a lot because getting the ICP roughly right just means you’re asking the right people, not that you’ve earned demand yet.
On the urgency-in-motion side, I’ve found it often shows up quietly during testing rather than as an obvious signal. Little things like: people asking “can I just export this for now?”, sharing screenshots of hacks, or saying “we’re doing X manually until we figure this out.”
When those start appearing unprompted, it’s usually a stronger validation cue than stated interest alone.