Our Methodology
Every number on Scout.ng is derived from this process. No black boxes, no hand-curated estimates — just a repeatable pipeline from raw listings to statistical medians. Here is exactly how it works.
Data Sources
We scrape publicly listed properties from Nigeria's major property portals: NigeriaPropertyCentre, PropertyPro, and Jiji. Only residential listings within Lagos State are collected.
Scraping runs weekly. Each run collects all available listings, deduplicates them against our existing database, and flags any that appear to have been removed from the source.
Geographic Coverage
We track 24 Lagos neighborhoods across the Island (Lekki Phase 1, Ikoyi, Victoria Island, Ajah, Sangotedo, etc.) and Mainland (Yaba, Surulere, Ikeja GRA, Gbagada, Magodo, etc.).
Each listing is matched to a neighborhood using our location normalizer — a database of hundreds of spelling variations, abbreviations, and sub-areas mapped to canonical neighborhood names. Listings that cannot be confidently matched are excluded.
Price Normalization
Raw prices from listings come in various formats: "₦85M", "$150,000", "NGN 2.5 million per annum", etc. Our price parser handles all common Nigerian price formats and converts everything to Naira (NGN).
Rental prices are annualized — monthly rents multiplied by 12, daily rates by 365. Foreign currency prices (USD, GBP, EUR) are converted at current parallel market rates.
Median Calculation
We use medians rather than averages for all headline price metrics. The median is robust to outliers — a single ₦5 billion mansion listing won't distort the numbers for an entire neighborhood.
Medians are calculated per neighborhood, per listing type (sale/rent), and optionally filtered by bedroom count and property type using PostgreSQL's percentile_cont function.
Outlier Detection
Listings with prices outside 2× the interquartile range (IQR) of their peer group (same neighborhood + listing type + bedroom count) are flagged and excluded from metric calculations.
We also apply absolute sanity bounds: sale prices below ₦5M or above ₦50B, and annual rents below ₦100K or above ₦500M, are automatically flagged as outliers.
Duplicate Detection
The same property often appears across multiple listing sites, or is relisted by the same agent. We detect duplicates using a fingerprint based on neighborhood + listing type + property type + bedroom count, then checking if the price is within 5% of a match.
Duplicates are linked in our database; only the most recent version is used in metric calculations.
Implied Gross Yield
This is an impliedyield calculated from asking prices — not actual transaction yields, which would require sale completion data we don't have access to. Treat it as an upper bound.
Lagos Reality Factor
The net yield subtracts four real costs that every Lagos landlord faces:
Confidence Levels
Each neighborhood snapshot is assigned a confidence level based on active listing volume.
Robust statistical basis. Medians are reliable and unlikely to shift significantly with minor listing changes.
Reasonable confidence. The direction is correct but the median can shift with a few new listings.
Treat as directional only. A few listings can move the median significantly — useful for context, not precision.
Important Disclaimers
All prices shown are asking prices from active listings — not transaction prices. Actual transaction prices are typically 10–15% below asking in Lagos.
We do not verify listing accuracy. Some listed prices may be incorrect, outdated, or deliberately misleading.
Scout.ng is a data intelligence tool, not financial advice. Always conduct proper due diligence and engage a qualified property professional before making any investment decision.
Scraping runs weekly. There is always a lag between when a listing is removed from a source site and when it disappears from our calculations.