Avito Follow-Up Campaigns

Scored by TRIBE v2 Neural Engine — Cognitive & Engagement Correlation Analysis

N = 72,426 Follow-Ups Sent • 101 Definitions • 29 Campaigns
Follow-Up Analysis Avito Campaigns All Campaigns Other Orgs
Total Follow-Ups Sent
72,426
Cumulative outreach
Template Sends
54,546
Static standardized
AI-Generated Sends
17,880
Dynamic personalized
Overall Response Rate
10.8%
Re-engagement rate
Qualification Rate
24.8%
Of responding leads
Definitions Scored
96/101
48 templates • 53 AI-gen
Performance Score Distribution (All Definitions)
Cognitive Load vs. Response Rate (Combined)
Template vs. AI-Generated — Avg Rates Comparison
Response Rate by Follow-Up Delay Bucket
TRIBE Score Profile — Template vs AI-Generated (Radar)
Perf Score vs. Response Rate — AI-Generated Definitions

Follow-Up Definitions (101)

Aggregated send/response/qualification rates per follow-up definition across all Avito campaigns

Campaign / DefinitionTypeDelaySent Resp RateQual (Responded)Disqual RatePerf ScoreAction

AI-Generated Messages Sample (100)

Individual messages sampled for TRIBE v2 scoring — outcome-labeled for correlation

Conv IDCampaignMessage Cog. LoadEngagementPerf ScoreOutcomeAction

🔬 Key Findings

  • TRIBE metrics strongly predict response rate across all follow-up definitions (r ≈ 0.68, p < 10⁻¹³). Higher cognitive load, engagement, and attention correlate with more re-engagement.
  • The correlation is driven by the contrast between template vs AI-generated types: templates score lower and have different response rates vs AI-generated. Within each type alone, no significant signal emerges.
  • attention_capture is the only metric that also significantly predicts qualification rate of responded (r = 0.20, p = 0.048).
  • At the individual message level (n=90 AI-generated), no TRIBE metric significantly predicts response or qualification — the individual message text variation has limited effect vs. campaign-level factors.
  • Template follow-up definitions all received the same TRIBE score (zero variance) since they share a single text. This means the template-vs-AI structure is the dominant variable.

Definition-Level Correlations (Combined, n=96)

TRIBE Metrics → Response Rate (Correlation Heatmap)
TRIBE Metrics → Qual Rate of Contacted (Correlation Heatmap)

Definition-Level Scatter Plots

Performance Score vs. Response Rate (All Defs)
Attention Capture vs. Qualification Rate of Responded

Message-Level Correlations (n=90 Individual Messages)

TRIBE Metrics → Response Outcome (Binary: responded or not)
Performance Score Distribution — Responded vs. Not Responded

Full Correlation Table — Definition Level (Combined)

AI-Generated Definitions Only (n=52) — No Significant Correlations

AI-Generated: Perf Score vs. Response Rate
Score Variance by Type (Template vs AI-Generated)