Consider $N_1 = \sum_{k=1}^K kr^k$ and $N_2 = \sum_{k=0}^K kr^k$ (assuming $r \neq 1$).
These two equations are equivalent since the first term in $N_2$ will always be zero due to the multiplication by $k$.
However, solving now:
$N_1 = \sum_{k=1}^K kr^k$
$N_1 = r\sum_{k=1}^K kr^{k-1}$
$N_1 = r\left(\sum_{k=1}^K r^{k}\right)^{'}$ (derivate trick)
Using geometric progression standard form:
$N_1 = r\left(r\frac{1-r^{K}}{1-r}\right)^{'}= r\left(\frac{r-r^{K+1}}{1-r}\right)^{'}$
Now solving for $N_2$:
$N_2 = \sum_{k=0}^K kr^k$
$N_2 = r\sum_{k=0}^K kr^{k-1}$
$N_2 = r\left(\sum_{k=0}^K r^{k}\right)^{'}$ (derivate trick)
Using geometric progression standard form:
$N_2 = r\left(\frac{1-r^{K+1}}{1-r}\right)^{'}$
Clearly, $N_1 \neq N_2$ based on this expansion. Where is the flaw in the logic?