🪄Methodology
Echo’s Insights transforms raw mobility signals into clean, structured, and privacy-compliant datasets for commercial analysis. Every metric, from visits to cross-brand interactions is accurate.
1. Data sourcing and signal validation
Our datasets originate from anonymized mobile location signals collected through third-party SDKs. These signals are filtered using multiple quality controls:
Spatial accuracy thresholds to exclude low-precision pings
Temporal validation to ensure consistent dwell and movement sequences
Device behavior modeling to detect real-world activity versus noise or spoofed signals
All signals are fully GDPR and CCPA compliant, and no personally identifiable information (PII) is ever included.
2. Visit attribution
Echo identifies real-world visits using a combination of location signals and POI geometries. Visit detection requires:
Minimum dwell thresholds within a POI polygon
Temporal continuity (e.g., repeated pings over a sustained time)
Spatial alignment with POI centroids and boundaries
Echo strives to remove any false positives, such as passerbys with low dwell times or inaccurate lat-longs.
3. Visit metrics and time series normalization
The Visits dataset captures the number of unique visit events per POI, per day. All visit counts are normalized across the platform to allow for relative comparisons across time and geography. Normalization includes:
Sample bias correction using panel calibration models
Device penetration rate estimation per country or region
Rolling average smoothing for trend analysis
Echo does not recommend using raw visit counts for absolute foot traffic volumes. Instead, our strength lies in directional, comparative, and time-normalized trends.
4. Dwell time modeling
Dwell Time measures how long visitors remain at a location. We compute:
Median and average dwell duration
Distribution across dwell time buckets (e.g., 5–10 mins, 30–60 mins, etc.)
Visit segmentation by duration bands to support use cases like engagement analysis or queue detection
Dwell is calculated using high-frequency ping sessions within POI boundaries, validated across multiple days.
5. Catchment area
Catchment Areas define the home and work origins of visitors to a POI. Echo uses long-term signal clustering to determine:
Inferred home locations (based on night-time behavior)
Inferred work locations (based on weekday daytime patterns)
We aggregate and spatialize these locations using privacy-preserving H3 grids (Level 8), then generate trade area polygons reflecting the actual geographic pull of each POI.
6. Cross-visitation analysis
Cross Visits measure the overlap between visitor sets for any two POIs, brands, or categories. Metrics include:
Shared visitor index: The percentage of visitors who go to both entities
Exclusive visitation rates: Measuring loyalty and brand differentiation
Category-level affinities: To detect latent brand connections
Cross-visitation is computed using de-duplicated, device-level visit logs and supports filters by time range, location, or visit frequency.
7. Heatmap generation
Visits Heatmaps spatially aggregate visit density into H3 Level 9 hexes. These are normalized by:
Device sampling rates
Total population baselines
Urban vs. rural weighting
Heatmaps help analysts understand neighborhood-level mobility patterns, detect points of interest without needing individual POIs, and support area-based demand modeling.
8. Data refresh
Visits and Dwell Time: Updated weekly
Catchment and Cross-visits: Refreshed monthly or quarterly, depending on country
Heatmaps: Recalculated monthly, with 3-month rolling averages available
All datasets are timestamped, versioned, and traceable to their processing pipeline version.
9. Privacy and compliance
Echo Analytics uses only privacy-compliant data sources. All datasets are anonymized, aggregated, and subjected to minimum threshold filters to ensure no individual behavior is ever exposed.
Last updated
Was this helpful?