Audience simulation

Know what the algorithm does before you post

Most creators post and hope. Korviq simulates. In the time it takes to scroll your feed, it runs your video through a complete platform distribution model — the same mechanics that decide whether you get buried or go wide.

STAGE 1 · FOLLOWERS

50 follower agents watch first

50
ALGORITHM GATE

Pass to earn a cold push

STAGE 2 · COLD

50 cold-audience agents · 7 cascade rounds

50

Real simulation result

Tech video · 52 seconds · Instagram · Jun 2026

77%

Hook score
PASS · threshold 75%

100%

Gate completion
PASSED · 100% / 55%

52%

Loved it
26 of 50 agents

t=48s

Clip moment
Peak shareable second

555

7-day reach
Followers + cold + cascade

TRENDING

Verdict
64% positive response

What the simulation shows

It mirrors how platforms really distribute content

Real followers

Watch first — including ghost accounts that never engage.

Algorithm gate

Decides whether your content earns a cold push.

Cold audience

People who have never seen your content get served it.

Social cascade

7 rounds of shares, DMs, group chats and flips.

At the end, you have the numbers. Not guesses — simulated outcomes built from your actual brain engagement data.

Hook score

The first 5 seconds decide everything

Your hook score is the average brain activation across the opening 2 seconds (t=0–2s), measured by the Helix model. Low means followers scroll past before the algorithm even sees completion data. High means the gate opens.

Hook score range

0.0 1.0

Pass threshold: 0.75+ — aligns with brain report HOOK PASSED verdict

Stage 1 · Follower batch

50 agents representing your real base

Completion rate% of real followers who watched to the end
Ghost scroll rate% who scrolled past instantly (industry avg 60%)
Engagement rateLikes, saves, shares from your actual audience
Algorithm gatePass / Fail — did you hit the threshold?

Gate thresholds

Completion needed to pass

PlatformCompletion
Instagram55%
TikTok65%
YouTube Shorts50%

A gate fail means no cold-audience push. The simulation tells you exactly how close you came — and what to change to flip it.

Stage 2 · Cold push

If the gate passes, 50 cold agents receive it — people outside your niche with no prior exposure. The hardest crowd to earn.

  • Cold engagement rate
  • Cold vs follower gap
  • Save rate — strongest amplification signal

Viral coefficient K

How many new viewers each existing viewer generates through sharing.

< 0.8Contained — limited spread
0.8–1.0Slow burn — some organic growth
1.0–1.5Growing — each viewer pulls one more
> 1.5Viral — exponential potential

7-day projected reach

Computed from your follower count, gate outcome, K score and cold engagement rate.

Not an average benchmark — computed directly from your specific video's neural data, run through a real distribution model.

K × reach

Drop zones

Where viewers bail — to the second

Every second the brain entered a DRIFTING or DISENGAGED state is flagged. Drop zones are where scroll probability spikes — where real viewers leave before the algorithm registers a completion.

Fix the top 2 and re-run. Watch the gate flip.

NONE DETECTED

This video had no drop zones — brain remained ENGAGED or HOOKED across all 52 seconds. Gate passed at 100% completion with 77% hook score.

[00:09] · Example of a typical drop zone

Neural engagement drops sharply at this second. High scroll probability. Music or pacing change recommended to push completion past the gate threshold.

Improvement intelligence

It doesn't just show numbers. It tells you what to change.

Fix the hook · score < 0.65

  • Lead with motion — visual cortex peaks within 1–2s of movement or a face
  • Delay the setup — exposition in the first 3s kills hook scores; start with the payoff
  • Image over text — language-heavy openings suppress visual engagement

Close the completion gap · gate failed

  • Shorten the video — cutting a slow ending often flips the gate
  • Add a save prompt — verbal CTAs at peak seconds lift saves
  • Reorder the peak — move your highest-activation second earlier

Improve cold engagement · Stage 2 weak

  • Reduce assumed context — make every concept self-contained in 10s
  • Optimize for saves — saves signal value-to-stranger
  • Check your brain type — language-heavy cold pushes need a universal hook

Raise the viral coefficient · K < 1.0

  • Surface the clip moment — your max-activation second is the share point
  • Controversy or tension — content that splits the cascade propagates further
  • Emotional peaks — high limbic activation drives sharing

How the simulation is built

Not a generic model — a model of your video

Every agent receives a personalized version of your brain engagement data, with individual variation. Agents peak at different seconds. They quote different lines from your transcript. Their scroll decisions are driven by your content's specific activation pattern, not a population average.

Ghost followers

Scroll past within 2 seconds at a rate matching research (60% of follower lists are inactive). They depress completion honestly — the way they do on real platforms.

The algorithm gate

Uses published platform completion benchmarks. The numbers are approximations — no platform publishes its exact algorithm — but they are the best available signal.

Social cascade

Runs 7 rounds of stateful group conversation. High-influence agents can flip undecided peers. A live force graph shows the network forming in real time.

Full reference

Simulation metrics

MetricSourceMeaning
Hook scoreAvg Helix activation t=0–2sEarly-bail probability · PASS ≥ 75%
Completion rate% HOOKED+ENGAGED secondsAlgorithm gate pre-signal
Clip momentSecond of max activationBest share / save candidate
Drop zonesDRIFTING / DISENGAGED secondsWhere viewers bail
Stage 1 ERFollower engagement rateCore audience response
Stage 2 ERCold audience engagement rateReach-expansion potential
Viral coefficient KSocial cascade outputSpread multiplier
7-day reachK × base reach projectionTotal projected audience

Run your first simulation free

The first 50 users get 2 free brain reports, each with a full audience simulation included. No credit card. Just submit and see what the algorithm would do.