Hero CTA moved from position 1 to position 4 in the layout — visits held steady, but the button that used to convert stopped getting clicked.
/Features / Connections
Most analytics tell you a number changed. Connections tells you what changed on your site — or in the world around it — that lines up with the shift, how strong the correlation is, and what to do about it. Every Connection ships with its receipts, not just a claim.
Every Connection is one of four shapes. The card looks the same either way — a summary, the before/after numbers, and the window they happened in — but what anchors it is different.
Wiley already tracks every structural change it detects — a layout diff, a content edit, an accessibility finding. When one of those lines up with a real move in traffic, conversions, or session depth within the right window, that's a Connection.
Hero CTA moved from position 1 to position 4 in the layout — visits held steady, but the button that used to convert stopped getting clicked.
/Algorithm updates, backlink losses, and mention spikes all get tracked as external events. When one lines up with a rank or traffic move — and there's no site change to explain it — Connections surfaces the outside cause instead of leaving you guessing.
A regional news write-up sent a referrer spike three days before this traffic jump — that domain alone accounted for 41% of the increase.
wtsp-news.comWiley builds a seasonal profile for your industry — school calendars, retail peaks, quiet summers. When a dip or spike falls inside that expected range, the Connection reads as explanatory rather than a warning, and it doesn't push you toward an experiment that wouldn't tell you anything real.
Your traffic dip aligns with the typical summer-break pattern for K-12 sites. No site changes detected in this window.
Some signals aren't a single event — they're a habit. When Wiley notices the same kind of issue surfacing three or more times within a month, it promotes that pattern into its own Connection, worth fixing at the root instead of one nudge at a time.
Three separate pages have lost meaningful scroll depth after a content edit in the last 30 days — same pattern, different pages.
Every Connection gets one of two confidence labels — nothing in between, and no invented percentage dressed up as precision.
The change is structural, the shift is meaningful, the timing is tight, and nothing else in that window could plausibly explain it. Strong Connections get a solid accent border and show up first.
The change is cosmetic, the shift is smaller, the window is looser, or something else — an algorithm update, a seasonal pattern — happened at the same time. Still worth knowing, just not worth overreacting to.
When two explanations could apply at once — say, a site change during a known seasonal trough — Wiley downgrades confidence and says so, rather than picking one story and hiding the other.
Every open Connection (except the seasonal, explanatory kind) has a "Run an experiment" button. Wiley carries over its context and the metric it detected movement in — you describe what you're testing and pick a duration, then it starts immediately.
See how Experiments works →Locked Connection types show a locked pill with an upgrade prompt instead of disappearing — you'll always know what you're missing.
Connections is included starting on Wiley Solo.