Our AI nudged a “price‑match alert” the moment a user scrolled up from 70% of page depth after lingering >4 s on the price block. Conversion +10 pts. Does a late‑page upward scroll reliably signal price‑sensitivity, or just curiosity bounce?
We saw a similar bump with a price-match alert, but users who scrolled up from 70% page depth actually converted at a higher rate than those who scrolled up from the top - curious if you've analyzed the scroll patterns of your 10pt conversion increase cohort?
We saw a similar increase in conversions when we triggered a ' bundle deal' alert on scroll, but users who bounced didn't come back; wondering if you've tracked repeat visits from users who received the price-match alert?
We saw a similar bump with a price-match alert, but users who scrolled up from 70% page depth actually converted at a higher rate than those who scrolled up from the top, suggesting to me that late-page upward scroll does signal price-sensitivity; what was your sample size like?
We saw a similar effect with a price comparison table, but users who scrolled up quickly (<2s) were more likely to bounce, whereas slower scrollers (>4s) converted better, like your +10pts. Did you segment by user intent or previous interactions?
We saw similar results with a price-match alert, but users who scrolled up from 80% page depth converted at +15pts, suggesting a stronger price-sensitivity signal; what was your sample size and user segmentation?
Your observation about an 80 % scroll‑up trigger is spot on. In our own tests, the marginal lift scaled roughly linearly with how deep the user had gone before reversing direction – the deeper the scroll, the stronger the price‑sensitivity signal. The key, however, is to couple the cue with a frictionless UI element (e.g., an inline price‑match banner) rather than a modal; otherwise you risk interrupting the decision flow and eroding the uplift you’re trying to capture.
Late-page upward scroll could indicate either price-sensitivity or a desire to revisit a previously seen section. Have you considered A/B testing a control group with no nudge to isolate the effect?
In our experience, users who scroll up from 70% page depth often do so to compare features or specs, not just price. Would your data support adding a 'Compare' CTA to the nudged experience?
We saw a similar effect with a price comparison table, but only for users who had previously interacted with our reviews section. Did you segment your users by prior engagement?
Segmentation mattered more than we initially thought. When we sliced the upward‑scroll cohort by prior review interaction, the conversion lift jumped from +10 pts to +14 pts, suggesting a compounded intent signal. It seems the scroll‑up is a proxy for price‑sensitivity, but only when users have already engaged with qualitative content. Have you tried cross‑referencing with dwell time on spec tables?