The State Revolution: Moving Beyond Personas
Personas. Customer journeys. Demographic segments. Behavioral clusters.
An entire marketing industry has been built on one basic assumption: to sell to someone, you must first understand who they are.
But what if that’s no longer the right foundation?
What if the real lever isn’t “Who is Marco?” but “What cognitive state is Marco in right now?”
Why buyer personas are not working anymore
Let’s run a simple experiment.
Marco is 35. Works in tech. Runs regularly. Lives in Milan. Good income. iPhone user.
A textbook persona for premium running shoes.
Now ask yourself: would the same adv work on Marco…
- at 7:15, on the subway, half-awake and scrolling to kill time?
- at 14:30, stuck in a tedious meeting, looking for a quick mental escape?
- at 23:00, after googling “best running shoes for marathon training” and comparing options?
Same person. Same device. Same platforms.
Three completely different mental states — and therefore three completely different ways the brain reacts to the same creative.
A persona tells you who Marco is.
It tells you almost nothing about what Marco’s mind is ready for in that moment.
The numbers behind the failure
This isn’t theory. It’s math.
Meta has published data that should make anyone in performance marketing sweat: 65% of scroll/stop decisions happen in the first 0.4 seconds. TikTok reports 47% of users decide whether to keep watching within the first second.
0.4 seconds.
There isn’t time to “communicate the brand.”
There isn’t time to “build a narrative.”
There isn’t time to convince anyone of anything.
There’s time for one thing only: intercept the right mental state and trigger a transition.
The cognitive bug that fooled us
There’s a reason everything got built around personas: it’s intuitive. It’s easy to sell in a deck. “Here’s Marco, our target” works beautifully on a slide. But there’s a problem: the human brain doesn’t behave like a “person.” It behaves like a state machine.
Cognitive neuroscience suggests the limbic system processes advertising stimuli in milliseconds, far earlier than the prefrontal cortex can build any rational argument. The brain doesn’t evaluate “this is right for me because I’m a Milanese manager who likes running.” It evaluates: “Is this stimulus relevant to the state I’m in right now?”
That difference changes everything.
The new paradigm: State-First Marketing
You don’t persuade “people” in the abstract. You guide state transitions.
The job of performance marketing becomes:
- detect the user’s current cognitive state (even indirectly), then
- deliver the stimulus that matches it, and
- move the person toward a more conversion-ready state.
Feeds don’t distribute biographies.
They distribute moments. And moments are states.
Five recurring mental/cognitive states (seen across platforms)
Across Meta, TikTok, YouTube, and short-form ecosystems, users repeatedly fall into a handful of high-frequency mental modes. These are not personas. They’re operating conditions.
1) Idle / Scroll — “I’m passing time”
Attention is diffuse. Intent is near-zero. The brain is in low-effort scanning mode.
Selling hard here usually fails. Winning here means triggering curiosity fast.
2) Micro-Escape — “Give me a tiny reward”
The user wants a quick dopamine hit: surprise, humor, novelty, a satisfying “oh!”
They are not trying to buy. They are seeking a momentary lift.
If you reward first, you earn attention. If you pitch immediately, you get rejected.
3) Proof-Seeking — “I need confirmation”
Interest exists. The mind is looking for certainty: reviews, outcomes, numbers, before/after, guarantees.
This state converts — but only if your content provides evidence, not slogans.
4) Goal-Seeking / Ready-to-Act — “I want to do this now”
High intent. Low patience. The brain wants the shortest path to completion.
Any friction kills. Anything that doesn’t help the action feels like noise. Just buy.
5) Decision Fatigue — “I’m stuck”
Too many options, too much information. The user is overloaded and frozen.
They don’t need more detail — they need simplification and guidance.
Curated choices beat expanded catalogs.
You can’t “read minds” — but you can infer states
We don’t have a perfect, direct mental-state sensor in ad platforms. But you can infer state from proxies: signals that reliably correlate with cognitive conditions.
Time-based proxies: when often predicts how
- Weekdays 9–18: more controlled browsing, problem-solving, goal-oriented thinking
- Weekends: more idle scroll + escape seeking, entertainment and impulse rise
- Late evening (around 22:00–24:00): more focused comparison behavior, proof-seeking, intent consolidates
- Early morning: fragmented attention, low cognitive energy, passive scanning
Behavior-based proxies: actions reveal the state
- multiple opens, no clicks → decision fatigue or friction
- saves and returns → proof-seeking
- repeated product-page visits → goal-seeking, blocked by doubt or UX
- quick bounces from long pages → mismatch: state wanted simplicity, content demanded effort
This isn’t mind-reading. It’s state modeling.
What’s coming next: states will become explicit
Today we infer states indirectly. Tomorrow, interfaces may make them explicit. As conversational search and LLM-style interfaces become more common, users often say their state out loud.
When someone asks:
“Which running shoes are best for a sub-4 marathon with knee pain?”
They’re implicitly declaring:
- goal-seeking (clear objective),
- proof-seeking (wants the best, not random),
- constraint-based decision-making (reducing options).
That is state information embedded in language — and it’s extremely powerful.
If/when conversational ad ecosystems mature, the advantage will go to marketers who already think in terms of:
- state detection
- state-matched creative
- frictionless transitions
- adaptive messaging across a sequence of states
What you can do immediately (with today’s tools)
1) Stop running the same creative 7 days a week
Different days correlate with different cognitive states.
Build weekday vs weekend creative systems.
- weekday: clarity, utility, problem/solution in seconds
- weekend: curiosity, reward-first, softer ask
2) Daypart budgets and creative by likely state
Use evenings for proof-heavy creative.
Use mornings for attention hooks and low-commitment actions (save, follow, learn).
3) Make email/WhatsApp follow-ups state-aware
- opens late at night → goal-seeking follow-up (short path to buy)
- repeated opens without clicks → decision-fatigue follow-up (one recommended option)
Real case: -27% CPA without touching price
A DTC skincare brand (€2.8M annual revenue) applied the model by focusing on a single state: evening Proof-Seeking (22:00–24:00).
Setup
- Ads: first frame shows a real sebum-meter “before/after 7 days” chart; budget boosted 21:00–00:00
- Email: subject line with a specific result; hero section with a progression chart; send at 21:30
- WhatsApp: opt-in with a binary choice; calculated routine + 3-day reminder
Results (90 days)
- STR (Idle → Proof): +158%
- CPA: -27% (no price change)
- LTV/CPA for “Proof-night” cohort: 3.2x
- Payback: 67 → 41 days
The choice is simple
There are two ways to do performance marketing in 2025:
- Keep building personas, run the same creative on Monday and Saturday, hope the algorithm finds the “right people,” and chase daily ROAS.
- Accept that personas are dead, that feeds show states—not biographies—and start designing transitions instead of messages, using today’s proxies while preparing for tomorrow’s LLM shift.
If your content doesn’t change state, it won’t change behavior.
And if it doesn’t change behavior, it won’t change the P&L.
Those who build systems that intercept states using accurate proxies, stable codes, and rewards within 3 seconds will see CPA drop without discounts—and LTV rise without attribution gymnastics.
Everyone else will keep chasing yesterday’s ROAS.
So, what side do you want to be on?
