Demo data

Every place, brand, demographic value and visit in MapDove’s hosted product is synthetic. The dataset is a demo modeled from neighborhood-level parameters. It is not collected from mobile devices, location panels, payment cards, cameras, Wi-Fi sensors or any other measurement of real people, and it contains no personal data.

We built it this way so that the full product — analytics, market intelligence and the SDSS — can be explored end to end without privacy risk and without licensing a data feed. This page explains exactly how the data is produced so you can judge what conclusions it supports.

At a glance

Item	Value
Cities	Austin, Houston, Phoenix, Miami (USD, imperial); Dubai (AED, metric)
Neighborhoods	16–17 per city, 82 in total
Hex cells	~0.8 km² each, covering each city’s study area
Places	247 per city, 1,235 in total, in 14 categories
Brands	31 fictional chains; ~27% of places are independent
Lifestyle segments	8
Daily visits	2025-11-01 to 2026-10-31 (365 days) for every open place
Determinism	Fixed seed — every installation produces identical numbers

All brand names are invented. Street names are real streets in each city used only to make addresses look plausible; the places on them do not exist.

Deterministic generation

All randomness comes from a seeded hash function of a text key (for example the place key and the date). Running the generator twice produces exactly the same database, so screenshots, tutorials, tests and support conversations all refer to the same numbers.

1. Neighborhoods and study area

Each city is described by 16–17 hand-specified neighborhoods with a center, a radius and five parameters: an income index and density index relative to the city baseline, a commercial index, a median age and an annual growth rate. Cities add a baseline income and density, an average driving speed and a seasonality profile.

The study area is the union of the neighborhood discs. Neighborhood boundaries are Voronoi cells around the neighborhood centers, clipped to the study area — so boundaries are geometric, not official.

2. Hex grid and demographics

A hexagonal grid (≈560 m side, ≈0.8 km² per cell) is laid over the study area. For each cell, neighborhood parameters are interpolated with inverse-square distance weights w = 1 / (d² + 0.6), then perturbed by seeded noise:

• Population follows base density × neighborhood density, decays toward neighborhood edges and away from the city center, with ±30% noise.
• Median income = city base income × interpolated income index × (0.85–1.15).
• Median age = interpolated age ± 5 years.
• Family share rises with age and falls with density; college share rises with income.
• Households = population ÷ an average household size of 2.0–2.6 depending on family share.
• Daytime population scales with the square of the commercial index — business districts have far more daytime than resident population.
• Rent and traffic indices combine income, commercial intensity and proximity to the center, then are rescaled to 0–100 within each city.
• Growth rate = neighborhood growth ± 1 percentage point.
• Segment is assigned from age, income, density and family share by fixed rules (Lifestyle segments).

3. Places

Each city receives a fixed quota per category (for example 40 quick-service, 34 coffee, 4 shopping centers). For each slot:

• A neighborhood is drawn with probability proportional to its commercial index (squared for big-box categories, which concentrates them in commercial districts).
• A location is drawn uniformly within 80% of the neighborhood radius.
• Brand: 27% of slots are independent; the rest pick one of the category’s fictional chains. Shopping centers are always unbranded.
• Size = category base size × (0.6–1.4); attractiveness = (sqft / base_sqft)^0.7 × brand strength (independents get a random strength of 0.8–1.15).
• Dwell time = category typical dwell × (0.85–1.15).
• Growth — an annual trend between about −12% and +20%.
• Opening date — 6% of places open during the data year (and ramp up); the rest opened between 2012 and 2025.

4. Base visit level

Each place’s base daily visits:

base = category_base_visits · A · (local_demand / city_avg_demand)^0.5 · (0.85–1.15)

where local_demand = Σ_h (population_h + 0.4 · daytime_pop_h) · exp(−d / 3 km) over hexes within 8 km. Well-located, larger, stronger-brand places get more visits — the same drivers the SDSS models look for, which is why the models calibrate well on this data.

5. Daily visits

For every day a place is open:

visits = base
       · weekday_weight(category, dow)
       · (1 + amp · cos(2π · (doy − peak_doy) / 365))
       · (1 + growth · (day − 2026-05-01) / 365)
       · holiday and event factors
       · opening ramp-up
       · (0.88–1.12 noise)

• Weekday weights per category (e.g. coffee strong Mon–Fri, apparel and malls strong on weekends).
• Seasonality — amplitude and peak day per city: Austin 6% peaking late April, Houston 5% peaking in April, Phoenix 12% peaking mid-February, Miami 12% peaking early February, Dubai 15% peaking mid-January.
• Trend — linear in the place’s growth rate, centered on 1 May 2026.
• Ramp-up — new openings start at 55% of potential and reach 100% after 90 days.

Visitors = visits × category visitor ratio × (0.95–1.05). Dwell = place dwell × (0.92–1.08).

Holidays and events

• Holiday season — a bell-shaped bump centered on 20 December, mostly within ±12 days, scaled by each category’s holiday sensitivity (strong for apparel, electronics, department stores and malls; negative for fitness).
• Black Friday (28 Nov 2025) — a one-day spike for holiday-sensitive categories.
• Thanksgiving and Christmas Day in US cities — most categories at 30% of normal (hotels, fuel and entertainment stay open).
• City events — Austin spring festival week (13–21 Mar 2026, +30% dining, hotels, coffee, quick service); a Houston storm (8–10 Jul 2026, −65% everywhere); Dubai’s winter shopping festival (15 Dec 2025–25 Jan 2026, +25% for malls, apparel, electronics, department stores and dining).

Hourly profiles

Hourly heatmaps combine the generated average visits per weekday with a fixed hourly curve per category. Hour-of-day patterns are therefore category-typical, not place-specific.

Limitations

• No real behavior. Patterns are plausible by construction, not observed. Do not use the demo data to make real investment, leasing or staffing decisions.
• Models are partly self-consistent with the data. Visit levels are generated from gravity-like demand, so the Huff model fits better here than it would on real data.
• No visitor origins or individual paths exist. Trade areas and cross-shopping are modeled from demographics, not measured.
• Geography is simplified — circular neighborhoods, Voronoi boundaries, no water or road network.
• One year only, so “YoY trend” is the generated growth parameter rather than an observed year-over-year comparison.
• Demographics are not census values, although they are tuned to look realistic for each city.

What changes with real data feeds

On the Enterprise plan, MapDove can be deployed with customer-supplied or licensed data in place of, or alongside, the demo:

• Your places — store lists, sizes, opening dates and formats replace synthetic places, so calibration uses your actual network.
• Your traffic — store-level counts (door counters, transactions, aggregated visit estimates from a provider you license) replace synthetic visits. MapDove only ingests data that is already aggregated to place and day (or hour) level.
• Demographics — official statistics or licensed estimates at block or tract level, re-aggregated to the hex grid.
• Custom markets and segments beyond the five demo cities.
• Network drive times can be supplied as precomputed isochrones.

The models, API and app stay the same; only the inputs change. Real feeds are governed by the Data Processing Addendum, and the Acceptable Use Policy prohibits loading or deriving data that identifies individuals.