The Funnel Mutated
Why ICP alone is no longer enough
April 18, 2026

Most B2B funnel problems are not really funnel problems.
They are focus problems wearing a funnel costume.
Teams publish constantly. They run campaigns across five channels. They have dashboards. They have nurture flows. They have “air cover.” Everyone is busy enough to feel productive. And yet pipeline stays weirdly uneven, sales says the leads are “fine but not really,” and every quarter somehow becomes a fresh round of “maybe we just need more top of funnel.”
Probably not.
Usually the real issue is simpler: the company has not defined who it reliably wins with and what makes those buyers ready to move.
That is where the old ICP model starts to break.
The old ICP was never enough. Now it is really not enough.
A traditional ICP is mostly a profile.
Industry. Employee count. Geography. Maybe a little tech stack detail if someone had a good week and access to enrichment.
That can help you build a list. It does not help you build timing.
An ICP tells you who could buy.
It does not tell you who is ready.
And in B2B, readiness is the whole game.
A perfectly matched account with no urgency is still a slow, political, committee-heavy sale. A good-fit account with a visible trigger can move quickly because the organization already has permission to change.
That is the mutation.
The funnel is no longer just about fit. It is about fit, timing, and motion.
The better model: fit, triggers, signals, plays
If you want a modern funnel that behaves like a real operating system, you need four things.
1. Fit
Fit is still the starting point.
Does this account look like customers who actually succeed with you?
That means more than firmographics. It also means operating reality:
- Do they have a real owner for the outcome?
- Do they have the systems or infrastructure your solution depends on?
- Do they have the kind of constraints that make your approach useful?
- Do they look like the kinds of customers you can win and keep?
Fit matters. It just is not enough.
2. Triggers
Triggers tell you why now.
A trigger is an event that makes change more likely this quarter than “sometime eventually.”
Common ones:
- a new leader joins
- a reorg breaks process
- a compliance or audit deadline appears
- a stack change creates disruption
- a hiring surge exposes bottlenecks
- a growth stall makes efficiency suddenly fashionable again
- a merger forces integration work nobody wanted but everyone now has to do
Triggers matter because B2B buying is rarely spontaneous. It usually needs a reason.
3. Signals
Signals are the observable clues that suggest a trigger may be active or that a committee is starting to form.
Some are external:
- job postings
- leadership changes
- funding announcements
- compliance statements
- public stack changes
Some are behavioral:
- repeated visits to comparison pages
- engagement with implementation content
- Trust Center views
- ROI calculator activity
- multiple stakeholders from the same account showing up at once
Signals are not truth. They are clues.
Treating them like truth is how teams build very expensive nonsense.
4. Plays
A signal only matters if it changes what you do.
That is where plays come in.
A play says: when we see this trigger and these signals in this ICP slice, we run this message with this proof and this offer.
That is what makes the system operational.
Without plays, teams collect signals like magpies and call it strategy.
What most teams still get wrong
A lot of companies are still using a static target-account mentality in a dynamic buying environment.
They build lists based on logos they want, not accounts that are actually in motion.
That is how you end up with a funnel full of technically “good-fit” accounts that never seem to do anything.
They were fit.
They just were not ready.
Another common mistake is confusing engagement with intent.
One person from an account reading a blog post is not a buying signal. It might be curiosity. It might be research. It might be a consultant. It might be someone procrastinating in a highly structured way.
What matters more is account momentum.
When several people from the same organization start engaging with higher-intent assets, especially when that lines up with a believable trigger, something different is happening. The committee may be forming. Or the champion may be quietly building a case.
That is when marketing gets interesting.
Old funnel logic vs. modern funnel logic
| Old model | Better model |
|---|---|
| Static ICP | Operational ICP |
| Firmographics first | Business situation first |
| More leads | Better readiness |
| Person-level activity | Account-level motion |
| Generic nurture | Trigger-led plays |
| Volume as proxy for progress | Movement as proxy for progress |
That is not just a tactics change. It is a worldview change.
The goal is not more activity
AI made browsing easier.
That sounds good until you realize it also made false positives easier.
Buyers can research faster, compare faster, and poke around your category with almost no commitment. That creates more top-of-funnel behavior, but not necessarily more buying motion.
So if your system still rewards raw activity, it will get dumber over time.
What you want instead is a readiness model.
A simple one asks:
- Which triggers actually correlate with urgency?
- Which signals actually correlate with committee formation?
- Which plays actually turn motion into sales-qualified opportunities?
- Where does friction still kill deals even when interest is real?
That is a much more useful set of questions than “how many leads did we generate?”
What a real operational ICP looks like
A useful ICP is not a slide. It is a working model of a business situation.
It should tell your team:
- what kind of organization this is
- what pressure they are under
- what tends to break in their world
- what triggers create urgency
- what proof closes deals
- what offer makes the first step feel safe
- where you are not a fit
If your ICP does not change what the team does next, it is not finished.
It is marketing décor.
The operator takeaway
The funnel did not disappear.
It mutated.
Static segmentation is too blunt. Generic nurturing is too noisy. High activity is too easy to fake.
The better model is:
- fit tells you who
- triggers tell you when
- signals tell you whether motion is real
- plays tell you what to do next
That is how you stop chasing curiosity and start building a real readiness engine.
And yes, that is less glamorous than saying “full-funnel.” It is also much more useful.
Want the fuller framework?
Get the e-book or the paperback.
Photo by Ricardo Gomez Angel on Unsplash