Search. Discover. Earn. Repeat
If you're a model, you don't need a spreadsheet to know things have changed.
But the data confirms what the Reddit threads have been screaming: the cam boom is over.
Traffic isn't collapsing—it's thinning. Viewers didn't disappear overnight; they spread out, fragmented, and changed their habits faster than platforms adapted.
We analyzed twelve months of anonymized Modelfindr click data across major platforms.
Here's what it says—without PR filters.
That's not catastrophic, but it's a long fall for an industry that built itself on perpetual growth.
Chaturbate saw the steepest dip (−28%), BongaCams close behind (−21%).
Only smaller networks like Stripchat and CamSoda showed minor gains—mostly because they started smaller.
The real story: the distribution curve flattened.
No site dominates anymore, but none are growing fast enough to offset attrition.
Even returning users are spending less per session.
Average time-on-page dropped from 6m24s → 4m57s in a year.
Viewers are watching more models for less money.
Total click volume increased 11%, but token spend stayed flat.
That means lower average revenue per creator even if overall traffic holds steady.
It's not about attention; it's about intent.
More browsing, less tipping.
The average "tip intent" window (time between session start and first tip) increased from 2m43s to 4m12s, showing users hesitate longer before engaging.
High-value tippers—the "whales"—used to drive up to 60% of creator income.
Where did they go?
They didn't all quit—they diversified.
Some moved to subscription-style ecosystems (OF, Fansly). Others shifted to "fan club" micro-donations across multiple models instead of one dominant loyalty stream.
That breaks the old stability loop where one loyal whale sustained multiple performers for years.
You can see burnout in the data long before it shows up on Reddit.
We analyzed live session frequencies and downtime patterns across 8,000 anonymized profiles.
Result: the average "consistent streak" (number of active streaming days before a break >7 days) dropped from 41 → 23 days in twelve months.
Meanwhile, session duration per stream increased 18%.
That means models are streaming longer, resting less frequently, and lasting half as long before burning out.
They're chasing algorithmic exposure curves that no longer reward consistency the same way.
The top comment in that Reddit post nailed it: "Some of us are showing everything for pennies."
They're not exaggerating.
Average tokens per minute of explicit performance dropped 32% since 2023.
Why?
Because algorithmic preview loops now leak enough content that paywalls feel redundant to viewers.
In other words, creators are giving away the good part just to get noticed.
When discovery is based on free exposure, pricing power dies.
Platforms created this problem by optimizing for retention over reward.
Let's talk numbers no one posts on marketing decks:
| Platform | Claimed Payout Split | Realized Share (after fees) | Notes |
|---|---|---|---|
| Chaturbate | 60% | ~52% | Token conversion + payout delay |
| BongaCams | 50% | ~45% | Currency rounding + commission buffer |
| Streamate | 35% | ~30% | Highest platform retention |
| CamSoda | 55% | ~48% | App/plugin fees reduce margin |
| Stripchat | 60% | ~53% | Slightly higher in EU |
| MyFreeCams | 50% | ~44% | Strong club retention, low base pay |
Source: composite partner data, affiliate tracking logs, and creator disclosures (Q3 2025).
It's not that models suddenly became "lazy." It's that math stopped working.
The Modelfindr analytics feed shows session overlap rising 19%.
That means more users open multiple models simultaneously.
Engagement becomes shallow and fragmented; emotional investment disappears.
In 2019, the average user interacted with 1.4 creators per session.
In 2025, it's 3.2.
That triple-dipping shatters the old "regulars" model of predictable support.
The data supports what performers feel: there's no loyalty loop anymore.
Tag velocity—our measure of how quickly users click into a category—is flattening.
Five years ago, spikes defined the meta: "cosplay," "GFE," "interactive toys."
Now the top 20 tags rotate slower, meaning less discovery and fewer breakout performers.
Tags like "new model" used to generate huge click momentum (average +42% CTR vs baseline).
Now? +8%.
Audiences aren't exploring; they're grazing.
You can verify this yourself on the live leaderboard.
When a site changes its algorithm, creators have no way to carry their audience elsewhere.
No export tools, no cross-platform ID portability, no transparency about how rankings work.
That's why ObsidianSignal calls for Data Portability and Algorithmic Disclosure.
Because the numbers prove what policy hasn't yet caught up to:
without leverage, even the most creative model is just another unpaid metric.
Data doesn't solve burnout, but it arms you for the next move.
This isn't a one-off story.
We'll continue releasing:
Because sunlight still works.
And because the only thing worse than a dying economy is pretending it's healthy.