The usual way a small agency handles a Title VI equity analysis is to hand it to a consultant, wait, and get back a PDF you did not build and cannot re-run. It works, but you are renting the answer, and you cannot ask it a second question without paying again.
There is another way to spend the same calendar time. Here is what 90 days looks like when the analysis runs on your own data, in your own hands.
Weeks 1 to 2: get your real network in
Upload your GTFS feed. Run it through the validator, fix what it flags, and publish a clean version. Now the tool is looking at your actual routes, stops, and spans, not a sample.
The point of starting here is that everything downstream inherits this. Get the network right once and every cost figure and every equity result after it is grounded in the real thing.
Weeks 3 to 4: calibrate to your own numbers
Point the cost model at your published operating cost, your dollars per revenue-hour from your own budget. This is the step that turns "a national estimate" into "our number." When a board member asks where a figure came from, the answer is your CAFR, not a vendor's assumption.
Weeks 5 to 8: sketch the change you are actually considering
Not a demo. The real one. The frequency cut you are weighing for fall, or the crosstown you have been asked about for two years. Draw it, and watch buses required, revenue-hours, operating cost, and the demographics of who is affected update as you go.
This is where renting the answer and owning it diverge. You are not waiting to find out if the change is a problem. You are seeing it, adjusting the alignment, and seeing it again, in an afternoon instead of a billing cycle.
Weeks 9 to 12: generate the draft and read it
Run the four-fifths disparate-impact test across the scenario and export the draft Title VI report as a Word document. Then do the thing you could never do with a black-box study: read it critically, hand it to counsel, and re-run it when they ask a question.
The deadline is the forcing function
Ninety days is not arbitrary. It is about the runway between "we should look at this change" and "this is on the board agenda." Spend it renting an answer you cannot reopen, or spend it building the capability to run the analysis yourself, on your own data, as many times as the plan changes.
At the end of a trial the reports you generated stay with you. That is the difference that matters: not a faster PDF, but an instrument you keep.
