Real Screens, Real Journeys
Every card below is a live screenshot from the v2 User Analysis view — no mockups. The product UI ships in English and Italian, and the full visitor suite is live today.
The whole audience in eight numbers
The view opens with the ledger: 4,293 total users — 4,285 new, 8 returning — across 4,432 sessions. Depth follows volume: 1.03 sessions per user, 1.0 pages per session, 9m 48s average duration. And trust gets a card of its own — a bot ratio of 0.0%, with the single flagged cookie counted right there. Every chart below hangs off these numbers.
- Volume, new vs returning and session depth on one strip
- Bot ratio sits in the KPI row — trust is a first-class metric
- Same totals the detail views reconcile against — no sampling gaps
One cookie, one visitor, one line
Daily Active Users counts unique first-party cookies per day — not raw hits, not sampled estimates. A full month on one line: peaks near 240 when a launch lands, dips to ~60 on quiet weekends, nothing smoothed over. When this line moves, a real audience moved it.
- Unique cookies, not hits — one visitor counted once per day
- Peaks and dips stay honest — no averaging away the weekends
- A month of context on a single line
Replay a session, event by event
Pick any visitor and their session opens as a timeline — session start, engagement pings while they read, idle and active transitions, scroll depth, clicks — each stamped to the second and anchored to the exact page it happened on. This 58-minute session crossed five pages, scrolled 25%, clicked twice, and never touched a form. That's a story no aggregate chart tells.
- Engagement pings, idle/active switches and scroll depth at second-level fidelity
- Every event anchored to the page it happened on
- New vs returning badge on every visitor card
Bots get their own tab
This “visitor” browsed ten pages in 73 seconds — homepage, contact, support, gone. Classic automation signature, so it's typed Bot, first seen 17 Sept, and kept out of your human numbers. The same timeline you use for customers lets you audit the verdict: rapid-fire page exits, no engagement pings, no scrolls.
- Suspicious traffic is labelled, not deleted — audit it yourself
- Type, first seen, pages, duration and source on every visitor card
- Bot sessions stay out of conversion and engagement metrics
Detection you can read, not a black box
The criteria are printed on the page: session-to-navigation ratio above 5:1, sessions under 10 seconds. This workspace counts 4,214 users — 4,213 real, one flagged — with the offender's 10.0:1 ratio and 1m 13s average session on display. Your dashboards don't just exclude bots; they show you which ones and why.
- Published criteria: ratio > 5:1, sub-10-second sessions
- Flagged count and real-user share side by side — 4,213 of 4,214 here
- Every flagged cookie listed with its evidence
Campaigns graded by who's actually human
Bot filtering rolls up to campaign level — users versus real users, per campaign. ai-governance: 6 of 6 real. The listing campaign: 3 of 3. The AI Act quiz: 2 of 2. Before you renew a listing or scale spend, you know whether a channel brings people or scripts.
- Per-campaign real-user rate at a glance
- Catches bot-inflated channels before they eat budget
- Same detection engine — one source of truth, zero extra setup
New vs returning — and who your regulars are
Three widgets, one question answered from all sides. The split donut: this audience is 100% new. The returning rate: 0.2% — 8 of 4,294 — with its trend. And Power Users filters for genuine regulars (navigation/session ratio above 0.3): the top one here ran 204 sessions across 29 days. You don't guess who's loyal — the table ranks them.
- Frequency split plus a returning-rate trend, not a vanity percentage
- Power Users filter (nav/session > 0.3) surfaces real regulars
- Sessions, navigations and days active per returning cookie
Consent status as data
GDPR handling shows up as an auditable ledger, not a footnote: every cookie marked unknown, granted or declined; opt-in rates per category — Necessary, Analytics, Marketing, Profiling; and the trend over time. Here 100% reads Unknown because no consent banner has fired yet — and the point is you can see that and prove it.
- UNKNOWN / GRANTED / DECLINED counted per cookie, not estimated
- Analytics, marketing and profiling opt-ins tracked separately
- Consent trend over time — evidence for audits, not compliance theatre
Sources, UTM and referrer
Where users come from, at visitor level — UTM parameters plus referrer, counted per visitor rather than per hit. Direct website entries lead at six, the tg4g source brought three, resources two. Attribution starts from the cookie, so it survives the whole journey.
- UTM + referrer resolved per visitor
- Counted on unique users, not raw pageviews
- Feeds campaign quality and lead scoring downstream
Parameter hygiene in one donut
utm 52.2% versus ads 47.8% — how much of your traffic arrives properly tagged. In other tools untagged traffic becomes “direct” mystery; here the split is measured, so tagging gaps become a to-do list instead of a blind spot.
- utm vs ads vs organic measured, not assumed
- Exposes tagging gaps other tools file under “direct”
- Parameter coverage as a KPI, visible at a glance
The tab that pays for it all
Filter the visitor list to the ones who converted, and the sessions that got them there line up beneath — five pages, sixty minutes, one form submit. It's the same cookie from shot 2: the timeline you replayed ends on this list. From first anonymous click to conversion, one thread.
- Converted visitors isolated with their full session history
- Duration and page counts per converting session
- Journey and outcome in one place — no stitching tools