Search went to zero for five weeks. Conversion went up 2.5x.
What to look at, and what to ignore, in a month where traffic collapses — from one site's own numbers.
5 weeks
Search traffic at zero
29.6
Avg. position (down from 24.1)
2.5x
Social conversion rate
9 slides
2
One morning, search traffic stopped
2 / 9
On July 19, 2026, search traffic effectively stopped.
Nothing deleted. Nothing moved. Every article still live at the same URL. Trailing-seven-day search sessions fell from about ten to essentially zero and stayed there.
💡 A week went into hunting for a cause. Nothing was broken.
3
Five weeks later, still nothing
3 / 9
Metric
Before
Aug 22, 2026
Search sessions (7d)
~10
Near zero
Average position
24.1
29.6
Sessions (28d)
~241
173 (-28.2%)
Indexed pages
—
475, no manual actions
⚠️ Not invisible — outranked. Two different problems, two different fixes
4
The number that went the other way
4 / 9
Social conversion: 3.75% → 9.38%. Roughly 2.5x.
Fewer people arriving. A higher share of them reaching the signup button. Both at once, over the same 28 days.
✅ The rate metric improved while the volume metric sat on the floor
5
The problem swapped shape
5 / 9
📉
Before: people arrive and don't convert
📊
Five weeks in: people convert, but almost nobody arrives
💡 Same complaint on the surface. Completely different next move.
6
Volume and rate are separate dials
6 / 9
🔍
Rankings (volume): not something you can reach in and turn
🛠
The path to signup (rate): entirely yours to change
⏸
Shutting everything down is the worst move: the leaks are still there when volume returns
⚠️ A quiet month is the cheapest time to trace what each visitor does
7
Reading a rate on a small sample
7 / 9
4 of 100 becomes 8 of 100. Rate doubles. Real change: four events.
The smaller the denominator, the more a single event swings the percentage. Judge on the rate alone and you book noise as a win.
💡 When a rate moves, read the numerator with it
8
A checklist for the month traffic falls
8 / 9
1️⃣
Indexed pages and manual actions first. Healthy means not broken
2️⃣
Average position over time. Bottomed out, or still sliding?
3️⃣
List volume and rate separately. Opposite responses
4️⃣
Keep one route that isn't search. You choose routes, not rankings
Don't switch everything off — something survives the bad month
1
Search traffic flat at zero for five weeks; average position slid from 24.1 to 29.6
2
Over the same window, social conversion rose from 3.75% to 9.38%
3
Volume and rate are separate dials — one can break while you turn the other
4
Write down whether what's falling on your site is volume or rate