SmartMarket · Data and integrations
The data behind every score, and how it plugs into your stack
A Smart Score is only as good as the data underneath it. SmartMarket runs on the Lepton Data Observatory, a unified India dataset stack, brings in Google Places Insights for live POI quality, and enriches it all with your own sales and store data. It then connects to the tools your team already works in.
Get a demo →Why the data matters
A score is only as good as its inputs
SmartMarket replaces guesswork with verifiable data science, and that promise lives or dies on the data behind it. Every Smart Score is built from layers that have to be granular, current and consistent across the whole country, so the same question gets the same quality of answer whether you are looking at a metro high street or a tier-three town.
Three sources feed every analysis: the Lepton Data Observatory, a unified India dataset stack; Google Places Insights, for live point-of-interest quality; and your own data, your sales history, channel performance and store formats. Together they enrich a 150m by 150m micro-grid into cells that are ready to score.
The Lepton Data Observatory
One unified India dataset stack
The Data Observatory is Lepton's unified India dataset stack feeding SmartMarket. It pulls demographics, mobility, infrastructure and built form into one catalog, so the model does not stitch together mismatched sources at analysis time. These are the layers it carries.
- 01
People · Demographics, affluence and footfall
Census demographics, an affluence index across the top 500 cities, and footfall measured at 150m grids, so each cell carries who lives there, what they earn and how busy it is.
- 02
Places · 30M+ POIs across 300+ categories
More than 30 million points of interest spanning over 300 categories, plus boundaries that nest locality hierarchies, pincodes, wards and villages for clean roll-ups.
- 03
Infrastructure · InfraNow upcoming-project feed
InfraNow, the Lepton feed of upcoming infrastructure projects with type, cost and completion, so a site can be read against where the city is heading, not only where it is today.
- 04
Mobility · Roads, transit and traffic
Road network with tolls and closures, transit timetables as GTFS across 16 cities, PCU traffic counts in time buckets, and drive-time and walk-time catchments for accessibility.
- 05
Built form · 3D buildings and house-level geocodes
Over 800 million 3D building footprints, land use and land cover, residential-versus-commercial split, and house-level geocodes for fine-grained built-form context.
Google Places Insights
Live POI quality, integrated
On top of the catalog, SmartMarket integrates Google Places Insights, delivered through BigQuery, for point-of-interest quality: ratings, review counts, sentiment and competitor benchmarking. It keeps competitor reads current, so the map reflects how locations are perceived today rather than a stale snapshot.
The coverage is the point. Places Insights spans 250M+ places across 300+ place types, carrying 70+ attributes per place and refreshed monthly, so the place quality feeding a Smart Score is first-party and current rather than scraped and stale.
Your own data
Tuned to how your business actually performs
Public data tells you about a place. Your data tells you how that place performs for you. SmartMarket brings client-side inputs into the same model, so the score is calibrated to your business rather than a generic average.
- 01
Sales history · What your existing sites earn
Your historical sales feed the revenue-prediction model, which learns from sales history plus location attributes to forecast what a candidate site could earn.
- 02
Channel performance · How your formats and channels behave
Channel performance and partner scores let the model weight the factors that actually drive results in your network, not assumptions.
- 03
Store formats · Your own network on the map
Store formats and locations let cannibalization checks model catchment overlap against your existing stores before any new site is recommended.
Many layers, plus your data, into one analysis-ready cell.
The Data Observatory layers stack over a single grid column, Google Places Insights joins as one accent layer, and your own sales and store data enters from the side. They converge into one enriched cell, ready to score.
The analysis toolset
From layers to an enriched, scorable grid
Once the data is unified, SmartMarket's toolset turns it into analysis. Spatial join and enrichment attach every layer to the grid; clustering, buffers and traffic-aware isochrones shape catchments; proportional allocation splits demand across overlapping zones. Heatmaps and weightage scoring then resolve the cells into a composite you control, and the revenue-prediction model and dashboards turn that into decisions.
Into your stack
Connect to the tools your team already uses
- 01
Google Sheets · Work where your team already lives
Google Sheets integration lets analysis flow into the spreadsheet your team already plans expansion in, with no export-and-reformat round trip.
- 02
BI tools · Feed your dashboards directly
BI-tool integration on the platform tier connects SmartMarket outputs into your existing business-intelligence dashboards, so expansion data sits beside the rest of your reporting.
- 03
The platform · Raw datasets for your own models
For GIS analysts and data scientists, the platform tier exposes the raw, granular datasets and custom expansion models, so your team can build and tune directly.
Proof
The same data stack, applied across industries
This data was not assembled for one product. The same Lepton location-intelligence stack is applied across real estate, retail and BFSI datasets, and the company serves 1000+ active customers across industries. SmartMarket is the expansion-decision layer on top of a catalog that is already maintained and in production elsewhere.
- Retail
- Real estate
- BFSI
- Quick commerce
See your own data scored on the map.
Bring a market and your store list. We'll show you the Data Observatory, the Places Insights layer and your own data resolving into one analysis-ready grid.
Get a demo →