Case study · Business plan and market analysis
A conversation with a passenger became several years of work, and that work produced a filed patent application, a federal seed funding package, a venture-fund introduction, three faculty advisors across two universities, and a defined service for a market of tens of millions of people a year.
A passenger — a research director at a federal transportation center — described a problem he could not get solved. People relocate for work constantly, and there is no serious intelligence available to help them decide where. Rankings, yes. Advice columns, yes. Nothing that answers should I move here, given who I am.
That conversation became several years of research and a set of assets that outlived the conversation entirely.
Most plans open with market size. This one opened with a harder question: what does a relocation decision actually consist of? Six categories — and the sixth is where every existing product stops.
The six-part structure the whole design rests on
Structural analysis rather than survey data. The categories were derived from what relocation decisions actually turn on, then tested against what existing products already answer for free.
City rankings are a solved problem and a free one. Nobody will pay for another. The unsolved part is personalization — taking the same public data and answering it for one person’s circumstances, constraints and stage of life. That reframing determined the data architecture, the pricing model and the patent claim, and it came from defining the problem properly rather than reaching for a market-size figure first.
Scale of the opportunity: US Census figures put annual household relocations between roughly 26 and 41 million depending on year and measure, with survey work consistently finding a far larger population actively considering a move in any given year.
The architecture called for ingesting well over a hundred public and licensable feeds — housing, labor, transit, education, health, climate, tax, crime, amenity and community data — and resolving them to a common geography so they could be scored together for one person.
| The hard part | Why it is hard |
|---|---|
| Geography resolution | Every source uses different boundaries. Counties, tracts, ZIP codes, metro areas and school districts do not nest cleanly |
| Update cadence | Some feeds are annual, some monthly, some real time. A composite built on mixed vintages misleads unless it says so |
| Licensing | Public feeds are free and often coarse. The precise ones are licensed, expensive, and carry redistribution limits that shape the product |
| Weighting | The weights are the product. Getting them wrong produces a confident answer that is wrong, which is worse than no answer |
Data licensing, not software, was the dominant cost and therefore the number that determined how much money the venture needed. That only surfaced because the architecture was worked out in detail before the financial model was written. Plans built the other way around routinely understate the raise by an order of magnitude.
Research that stays in a folder has produced nothing. This produced five things, and every one of them required somebody outside the project to look at the work and agree it was worth something.
Five assets, none of them bought
Each represents an independent external judgment: a filed utility patent application, a federal small-business research seed application, a warm introduction to a venture fund made by a faculty member using his own credibility, three faculty advisors across two universities, and a defensible market at national scale.
| Asset | What it is worth |
|---|---|
| Utility patent application | A filed priority date and a defensible claim over the personalization method, independent of whether this venture ships |
| Federal seed funding application | A complete, submittable technical and commercial package — reusable, and the hardest document in the set to write |
| Investor introduction | A faculty member willing to spend his own credibility on a warm intro to a venture fund. That is earned, not requested |
| Academic collaborators | Three advisors across two universities, including one whose own research on remote-worker migration bears directly on the thesis |
| A defined service | A product concept addressing a market of tens of millions of household moves a year, with the unmet need named specifically rather than asserted |
The venture is currently paused while the founder weighs the scale of a solo launch against other commitments. The patent application, the funding package, the data architecture and the relationships all remain. Work of this kind does not expire when a launch date moves.
| The usual business plan | What was done here |
|---|---|
| Opens with market size | Opens with what the decision actually consists of |
| Architecture sketched after the financials | Architecture first, because it set the cost base and therefore the raise |
| Circulated to people who want it to succeed | Put in front of a patent examiner, a federal panel, a venture fund and three faculty |
| Produces a document | Produced filings, applications, introductions and advisors |
The method is not specific to relocation software. Define the decision precisely. Work out what it would genuinely take to answer it. Cost that honestly, before writing a projection. Then put the whole thing in front of people whose job is to find the hole in it — because the ones who agree to spend their own reputation on it are the only real validation there is.
Marketing Analytics Consultants
We define the decision, cost the architecture before the projections, size the market against real sources, and help you get it in front of the people whose agreement actually means something.