We could move into the Empire State Building before the final network was ready. That did not mean pretending day one would look like the finished design.
The office was expected to have roughly 100 megabits down and 40 up by December 18. Another circuit planned for mid-January would bring the total to about 200 down and 140 up. Our old office at 60 Broad had 600 megabits in both directions. The target for the finished Empire State Building setup was one gigabit up and down, but that could take another 90 to 120 days.
The delays were not caused by a missing switch or a cable somebody forgot to order. Building riser work had been held up by legal disputes, and new fiber paths and service turn-ups take time even when everybody wants them faster.
That left a real operational question. Most office work would be fine on the initial capacity, especially with the right quality-of-service rules on the firewall. Video review was different. Moving large footage files changes what “enough bandwidth” means.
I did not want people walking into a famous new address and discovering that their daily work had become slower. The office could be beautiful and still fail if the network made the teams less effective.
So the answer was staged. Move what the early network could support. Add the next circuit in January. Use traffic controls to protect important work. Keep pushing the building and providers on the higher-capacity design. Do not promise that the final state will exist on opening day.
This was a familiar kind of growth problem. The company’s timeline and the infrastructure timeline were both real, and they did not line up perfectly. Waiting for ideal conditions could stall the move. Ignoring the constraint could create a miserable first month.
The job was to make the gap explicit and design around it.
We would have something workable on day one. It would not be the final design. That sentence was not an excuse. It was the operating fact the rest of the move had to respect.