Analytics, heatmaps & feedback for shared docs
After you share an ilolink, you see who opened it, how far they read, where they clicked, and what they said. It runs cookieless — a rotating visitor hash, no fingerprint, no cross-site profile.
What gets measured, and how privately?
Every open is counted without a cookie. ilolink derives a rotating visitor hash to tell one reader apart from another for a short window — there is no fingerprint, no cross-site tracking, and no personal profile stored anywhere. Because the hash rotates, unique views are approximate by design. You get total views, approximate uniques, average time on page, scroll-depth, top referrers, countries, device class, and 30-day daily views.
Views vs unique views
Views is every time the page was opened — one person who reopens the link three times is three views. Unique views is the approximate number of people behind those opens. The gap between them tells you something: a large view-to-unique ratio means people are coming back, which is a good sign for a proposal or a spec. A ratio near 1:1 means one-and-done reads. Treat the unique number as a close estimate, not a headcount.
The scroll-depth funnel
Scroll-depth is bucketed at 0/25/50/75/100%. Read it like a funnel: if 90% of readers pass 25% but only 30% reach 75%, the drop-off is in the middle of the page — the part they abandon is where you lost them. Pair it with average time on page. High time plus a shallow funnel means people are stuck early; low time plus a deep funnel means they skimmed to the end. Both are signals about where to cut or tighten.
Heatmaps
Heatmaps come in two kinds — click and scroll — and you can view each by device, because a layout that works on desktop can bury your call to action on mobile. Hot zones show where attention and clicks land; cold zones show what got ignored. If the section you care about is cold, it is either too far down or not pulling the eye. Check the click map for taps on things that are not links — that is a sign readers expected something to be interactive.
Feedback and comments
Readers can react or leave a short note without making an account — the friction is deliberately near zero, so you actually hear back. Comments are threaded and anchored to the doc, so a reply sits next to what it responds to instead of floating in a separate inbox. You see reactions, notes, and comment threads in the same place as the analytics, so quantitative drop-off and qualitative why sit side by side.
Where it's going
Today the metrics are format-neutral — they work the same whether you shared Markdown, HTML, or a mockup. The direction is format-aware measurement: per-slide view time once slides ship, per-page read depth for PDFs, and watch- or listen-through once media hosting exists. None of that is live yet. When the formats and media land, the analytics grow to match them — until then, assume only what is described above.
Questions
- Do you use cookies?
- No. Analytics run on a rotating visitor hash — no cookies, no fingerprint, and no cross-site profile.
- Are unique view counts exact?
- No. Uniques are approximate by design. The visitor hash rotates for privacy, so the count is a close estimate rather than a precise headcount.
- Can readers comment or react without an account?
- Yes. Reactions, short notes, and threaded comments all work with no reader login — that low friction is the point.
- Where is the data processed?
- On Cloudflare's edge, globally over HTTPS, and cookielessly. There is no server-side account tied to a reader.
Try it on your own doc
See it on your own doc.