You cannot say with a straight face that predicting 20 million users, getting only 1 million, is "a good prediction".

> cannot say with a straight face that predicting 20 million users, getting only 1 million, is "a good prediction"

Yes, I can. Especially when the prediction is ten years hence and made about a product that doesn’t exist predicated on a launch frequency nobody has attained. 1mm users means core leaps in launch frequency, beam shaping, operations and sales came together. Furthermore, sitting at the deadline, the 20mm isn’t unrealistic. For hard technology venture, that’s a good prediction. (A lot of tech meets its order of magnitude threshold only to learn the goal is unrealistic, ever.) Not spectacular. But also not bad.

Within an order of magnitude would be 2M users. This is not within an order of magnitude.

> This is not within an order of magnitude

10^6 vs 2 x 10^7. An order of magnitude. Order of magnitude isn’t literally divide by ten. Granted, there is a qualitative difference between 1 to 10mm and 1 to 99mm, but they’re both one order of magnitude apart.

Thanks!

just because investors have the presumption that presentations are lies and founders have the incentive to lie during the presentation doesn't somehow warp reality into thinking that prediction was a good one.

It may have constituted a good business decision to lie during presentation day, but whether or not that's true is wholly independent of the state and accuracy of the prediction itself; that prediction sucked.