Real Screens, Real Answers
Every widget below is a live screenshot from the v2 Lead Quality dashboard — no mockups. The product UI ships in English and Italian, and every number traces back to the scoring cards on the leads themselves.
Four numbers that start every meeting
Leads scored, average score, qualified rate, hot leads — each with its delta against the previous period. 16 scored (+10), average 35.8 (−19.0 pts), qualified rate 43.8% with its definition printed underneath, hot leads 7 with 1 nurturing and 8 discarded. Pick any date range; the comparison is automatic.
- Period-over-period deltas on every card — direction, not just level
- Qualified rate defines itself: leads in Qualified + High Priority
- Hot leads split into nurturing and discarded — nothing hides
The trend line that answers "are leads getting worse?"
Average score and qualified % as daily lines, lead volume as bars behind them — dual axes, one timeline. Here both lines peak around 22–24 September, dip on the 26th, and recover into October. Visible, dated, discussable — the day quality moved is the day you ask what changed.
- Average score and qualified % on one timeline
- Lead volume behind the lines — context for every swing
- Daily grain: trends you can correlate with campaigns
Humans, bots, and the honest unknown
Every conversion session is classified: human, automated, UA-bot, unclassifiable, or no form telemetry. This period: 12 human sessions (75%), 4 unclassifiable (25%), zero automated, zero bots. The unknowns are reported as unknown — never silently counted as people to make the funnel look better.
- Automated and bot traffic separated from human before you celebrate
- "Unclassifiable" reported honestly instead of guessed
- A clean denominator for every conversion metric downstream
Where your leads actually land
The six bands with counts and percentages: Disqualified 8 (50%), Cold 1 (6%), Qualified 7 (44%) — and zero in Really Cold, High Priority and Critical. Half the period disqualified means the gates are working, not that the business is dying. Zero High Priority is a fact about your market this month, not a failure of the model.
- Six bands, counts and percentages, no aggregation games
- Disqualified share = gate activity, visible at a glance
- Empty bands stay on the chart — absence is information
The shape of your lead pool
Leads bucketed by raw score: 2 in the single digits, 4 in the teens, 2 in the twenties — then a gap — and 6 at 50–59, 2 at 60–69. That bimodal shape is the system working: junk scores low and dies at the gates, real buyers cluster high, and the empty middle is exactly your nurture conversation.
- Raw scores, not bands — see the spread inside each band
- The gap in the middle is the nurture question, visualised
- Buckets align with the bands printed on the scoring cards
What drives your scores — and what your site can improve
Every signal with its hit rate, hits, misses, n/a and average points: multi_session 100%, demo_contact_page 85.7%, pages_viewed 85.7%, engaged_time 85.7%, valid_email 75%, scroll_depth 64.3%, valid_phone 50%, message_substantial 50%, semantic_alignment 50% — and form_struggle at 12.5% with a negative average, friction punished by design. A signal that never fires is a site fix, not a model bug.
- Hit rate per signal — your site-improvement backlog, ranked
- Misses and n/a separated: a miss is fixable, n/a is not scored
- Average points per signal — see what actually moves the score
Five axes, one honest bar chart
Engagement 51%, Quality 50%, Fit 50%, Intent 20%, Urgency 5% — computed over available signals only. The B2C masking is printed on the card: business_domain and commercial_language are excluded, not zeroed. Low urgency in a consumer market is a fact, not a bug — and the pill next to it says exactly which signals didn't apply.
- Which axes your leads satisfy — and which starve
- B2C masking declared on the card: excluded, never zero
- Available-signals-only denominator — no phantom penalties