[ QUOTE ]
Ok. Exponential is definitely not a good fit. A much better fit is parabolic. Here's a couple of slightly different fit techniques and projections for the month. According to this projection, the target would be missed by about $1.2M.
However, I expect a big surge near the end of the month. Don't know if they'll make it. Like I said; $12M for the quarter was too aggressive, imo.
I'm pretty sure I could fit this better by adding a weekly sinusoidal modulation, but I'm lazy.
[/ QUOTE ]
Parabolic is almost never a good fit for projecting the future, and to a trained eye your graph doesn't look like a good fit at all. Polynomials are very good at fitting points, and very bad at forecasting.