By Ether DataEvery building in America.
Every hour.
Every building in America has a different population every hour of the week. 168 is the only product that maps it. Drag the slider — see Manhattan at 9am, then at 6pm.
168 runs both ways.
Backward, for measurement. Forward, for planning.
Most measurement data tells you what already happened. Most planning data is a guess. 168 is built from the structural rhythm of a city — public records, regular patterns, the way an office or a hospital or a hotel actually works — so the same data answers both questions. As far as we know, no other national dataset does both.
On held-out hours of NYC subway ridership, the model matched 81% of the variation — using only structural workforce data, without ever seeing the live counts.
On Boston subway ridership — a different city, a different transit system — the model matched 96% of the variation, using the same structural approach.
At city blocks the model had never seen before, prediction error fell 78% versus the city-wide hourly average.
Two products. Two sides of every building.
Who's working here.
Every building in your metro, hour by hour. The dynamic picture.
- 20 industries — finance, tech, hospitality…
- 168 hours per week, per industry, per building
- Refreshed every quarter
- Paid per metro
Who lives here.
Every neighborhood in the country, year by year. The static picture.
- 250+ attributes from the U.S. Census
- Incomes, ages, languages, household types, commute habits
- Refreshed every year
- Free in New York State · paid nationally
Foundation is where. 168 is when.
Three things every building tells you, every hour.
For every building in America, 168 answers three plain questions, every hour of the week: who's working there, when they're present, and where they're coming from. Any address snaps to the dataset — no joins or lookups required.
Who's working here.
Every building, labeled with its mix of industries — finance, tech, hospitality, retail, healthcare, government. Built from public workforce records. No device tracking, no purchased panels.
When they're here.
Every hour of every day, week after week. Monday at 9am is not Saturday at 9am — the data shouldn't pretend it is. 168 hours, mapped building by building.
Where they're from.
The home neighborhoods of everyone working in a given building. Connects where people work to where they live — by industry, hour by hour. Trace any building back to the people behind it.
One building. One query. Three answers.
-- Who is in this Manhattan block, by sector, on Tuesday at 14:00? SELECT h3_r8, sector, workforce_share AS composition, -- A: Workforce activation[38] AS present_at_tue_14, -- B: Hours (Tue 14:00 = h38) workforce_share * activation[38] AS live_share FROM `etherdata.168.hex_hour_v1` WHERE h3_r8 = `carto-os`.carto.H3_FROMGEOGPOINT(ST_GEOGPOINT(-73.9857, 40.7589), 8) ORDER BY live_share DESC;
The Census, neighborhood by neighborhood.
Over 250 attributes from the U.S. Census — incomes, ages, languages, household types, commute habits, vehicle access, tenure — mapped to every neighborhood in the country.
Free across New York State. National coverage on subscription. Snaps to the same grid as 168, so you can put who lives there next to who's working there in a single query.
· ACS R8 panelWhat teams build on the Census layer.
Live demos showing 168 Foundation joined to operational data — bike-share, 311, public health. The Census layer becomes a stable equity baseline for any place-based decision.
Bike-share equity & need.
Snap Manhattan bike-share onto Foundation: car ownership, transit access, rent burden, work-from-home. Surfaces neighborhoods where supply doesn't match the structural need — Service Gap Score in one number.
See the demo02Equitable bike access.
Compare low-income exposure to bike infrastructure on a single grid. Foundation gives planners a stable equity baseline — one that doesn't move with usage data or seasonal patterns.
See the demo03311, per capita risk.
Anchor NYC 311 health and sanitation data to Census denominators — rent burden, density, response time. The risk surface re-ranks per resident, not per population mass.
See the demo04Member vs casual riders.
Membership penetration overlaid with income and education at every block. Surfaces neighborhood pockets with structurally lower conversion — equity-aware segmentation, no causal claim.
See the demoFour industries. The same buildings.
168 is built for teams that already think in places. The data doesn't change between use cases — only the question does. These are the questions we hear most often.
OOH · DOOH"Who's actually in front of this board at 4pm Thursday?"
A board in Midtown is not selling the same audience at 9am, noon, and 9pm. 168 plans and prices for that fact, hour by hour, instead of averaging it away into a single "daytime population" number.
Commercial Real Estate"Who's in the trade area when this site is open?"
Site selection with a clock attached. Foot-traffic vendors give you a number. 168 tells you who those people actually are — by industry, by hour, and by where they came from. The kind of detail a site committee can defend.
Healthcare Networks"Where is the demand we're not serving?"
Resident-only models miss the workers — the 9-to-5 population that has injuries and symptoms during the workday, miles from home. 168 shows you the gap between where people live and where they actually are when they need care.
DSP · SSP"Who's actually around this impression?"
Bid context, not an audience ID. For every geo-stamped impression, 168 tells you who's working in that building right now and where they're coming from — without device data, panels, or identity stitching. Privacy-preserving by construction.
See 168 in action.
Short demos showing what changes when a building has a different population every hour.
See all examples
What we've published.
Public releases and research drops about 168 and the methodology behind it. Working journalists are welcome to email us directly.
Press inquiries: press@etherdata.ai. Logos, screenshots, and founder bios available on request.
Bring your model.
We bring the where and the when.
National coverage and partner integrations: info@etherdata.ai
Press: press@etherdata.ai
Built without device data. Not because we couldn't get it. Because we didn't need it.