DATA CAPTURE
How We Capture
Baseerah is the Arabic-first mobile app that powers IntellX's panelist network. It turns 10,000 Saudi households and 30,000 individuals into a distributed data collection system that captures every purchase, every channel, every occasion. This is the supply side of the intelligence layer.
HOW IT WORKS
From onboarding to daily capture in under five minutes.
The app is designed for a single outcome: sustained, high-quality data capture with minimal panelist friction. Every design decision optimizes for capture frequency and accuracy, not feature count.
Onboarding
Eight-screen flow. Language selection, demographic collection (the four weighting dimensions plus household profile), camera permission, and first capture tutorial. A panelist is capture-ready within 3 minutes. Demographic fields map directly to the IPF weighting dimensions: region (GPS auto-detected), socioeconomic class, nationality, and household size. No field is collected unless it serves the analytical output.
Daily Capture
Four capture methods, one unified camera interface. Barcode scan for packaged goods (instant SKU lookup against a 50,000-product dictionary). Receipt photo for grocery trips (AI extracts every line item in one shot). Product photo for unpackaged or out-of-home items (AI identifies brand and category). Manual quick-log as fallback. Average capture time: under 10 seconds per item. The app uses fire-and-forget submission: data saves locally, displays a reward animation instantly, and syncs in the background. Panelists never wait.
Context Tagging
Every capture is tagged with context that legacy panels never collect. Channel (modern trade, traditional trade, delivery app, gas station, cafe, vending). Occasion (commute, work break, social, exercise). Companion (alone, family, friends, colleagues). Consumption timing (immediate, within 1 hour, later). This context data is what transforms a purchase record into a consumption occasion record, and it is what makes out-of-home analytics possible.
Gamification
Five-layer reward architecture designed to sustain daily engagement over months, not days. Fixed base rewards provide predictability. A mystery multiplier after every capture triggers variable reward psychology. Streaks and milestones create loss aversion (7-day streak completion reduces churn by 30-40%). Tier progression (Bronze, Silver, Gold) builds switching costs. The engagement model is built around utility and habit formation, ensuring sustained participation through intrinsic value.
Character System
Baseerah assigns each panelist a consumption archetype based on their purchase behavior. This is not cosmetic. It drives personalized engagement: different archetypes receive different nudges, different milestone celebrations, and different streak messaging. The character system is the behavioral engine that keeps capture frequency high after the initial novelty fades.
THE MOAT
Out-of-home capture at the item level.
This is the capability that no other panel in MENA possesses. The ability to capture what consumers buy and consume outside their homes, at the individual item level, with channel and occasion context.
The Problem
Legacy purchase panels are built on a diary model: panelists log items they bring home. The methodology assumes that 'purchase' and 'bring home' are the same event. In Saudi Arabia, they are not. 25-35% of FMCG consumption happens at gas stations, cafes, delivery app orders consumed at the office, convenience store grabs during a commute, and vending machine purchases. These occasions are structurally invisible to diary-based panels.
How Baseerah Solves It
Baseerah solves this with individual-mode capture. When a panelist is logging for themselves (not the household), the app captures the item, the location (GPS city-level), the channel (gas station, cafe, delivery app, convenience store), and the occasion context. The AI identifies the product from a photo or barcode. The panelist confirms with one tap. The entire capture takes under 10 seconds. This data flows into the same pipeline as household captures: parsed, matched, weighted, analyzed. The result is a complete consumption picture that includes every channel, every occasion, and every product.
Defensibility
The moat is not the technology alone. It is the combination of a trained panelist network that captures out-of-home behavior, AI that identifies Arabic-English products in non-standard retail environments, and analytical methodology that knows how to weight and project OOH consumption to national estimates. Each component is necessary. None is sufficient alone.
PANEL QUALITY
The metrics that determine data reliability.
A purchase panel is only as good as its panelists' compliance. IntellX monitors four quality dimensions continuously and intervenes when any metric falls below threshold.
Response Rate
Percentage of panelists actively capturing at least 5 items per week. The primary indicator of panel health. Below 70%, recruitment triggers to backfill underperforming strata.
Panel Stability
Month-over-month retention rate. Target: 75%+ annual retention (industry benchmark for MENA is 60-65%). The five-layer gamification system is specifically engineered to exceed this benchmark.
Attrition Management
When a panelist drops out, the system identifies which demographic stratum is affected and triggers targeted replacement recruitment. The goal: maintain demographic proportionality without relying on heavy weighting corrections.
Diary Fatigue Detection
Capture frequency is monitored per panelist over time. Declining capture rates trigger behavioral interventions (streak recovery nudges, tier demotion warnings, targeted reward boosts) before the panelist is lost entirely.
Engagement model.
Baseerah is designed as a utility app with proprietary FMCG methodology that makes receipt capture a natural extension of how users manage their household spending. The engagement model sustains daily participation through intrinsic value and behavioral mechanics rather than compensation escalation. This architecture enables scaling to larger universes without proportional cost increases.
PANEL HEALTH
Real-time monitoring of data quality.
Panel health is not a quarterly report. It is a continuous monitoring system that catches problems before they affect client-facing data.
Compliance Scoring
Every panelist carries a compliance score based on capture frequency, capture completeness (did they log all purchases, or just the big ones?), and data accuracy (do their captures match expected patterns for their demographic profile?). Scores below threshold trigger automated interventions.
Demographic Representativeness
The panel's demographic composition is continuously compared against GASTAT population estimates. When any stratum drifts below its target proportion by more than 5 percentage points, the system flags it and triggers recruitment in that stratum.
Capture Pattern Anomalies
Fraud detection identifies panelists whose capture patterns deviate from expected behavior: identical purchases logged repeatedly, captures concentrated in unrealistic time windows, or purchase volumes that exceed plausible household consumption.
Data Freshness
The pipeline monitors data lag from capture to availability. Target: captured purchases are processed, matched, and available for analysis within 72 hours. Delays beyond this threshold are escalated.
Understand the supply side.
Request a briefing on how IntellX captures, validates, and processes consumer data. We will walk through the panelist experience, the quality metrics, and the infrastructure that turns raw captures into decision-grade intelligence.
Request a Capture Briefing