This question is motivated by Section 3.2.3 in Griffiths' book on Electrodynamics.
I am trying to calculate the force of attraction exerted on a point charge by an induced charge, from an infinite grounded plane. Using the method of images, we can calculate the potential to be $$V(x,y,z)=\frac{1}{4\pi\epsilon_0}\left[\frac{q}{\sqrt{x^2+y^2+(z-d)^2}}-\frac{q}{\sqrt{x^2+y^2+(z+d)^2}} \right].$$
I am unsure where to go from here. Griffiths suggests using the relation $$\mathbf{F}=-q\ \nabla V,$$ but I don't see why this is the attractive force. I cannot understand why this should be be the force exerted on $q$, rather than the force exerted $by$ $q$.
But even if $\mathbf{F}=-q\ \nabla V$, I still don't know how to calculate the force. The components of the gradient are given by \begin{align} &\frac{\partial V}{\partial x}=\frac{1}{4\pi\epsilon_0}\left\{-\frac{qx}{[x^2+y^2+(z-d)^2)]^{3/2}}+\frac{qx}{[x^2+y^2+(z+d)^2]^{3/2}} \right\},\\ &\frac{\partial V}{\partial y}=\frac{1}{4\pi\epsilon_0}\left\{-\frac{qy}{[x^2+y^2+(z-d)^2)]^{3/2}}+\frac{qy}{[x^2+y^2+(z+d)^2]^{3/2}} \right\},\\ &\frac{\partial V}{\partial z}=\frac{1}{4\pi\epsilon_0}\left\{-\frac{q(z-d)}{[x^2+y^2+(z-d)^2)]^{3/2}}+\frac{q(z+d)}{[x^2+y^2+(z+d)^2]^{3/2}} \right\}. \end{align}
Evaluating the gradient at $(0,0,0)$ gives $\mathbf{F}=-\frac{1}{4\pi\epsilon_0}\frac{2q^2}{d^2}\widehat{\mathbf{z}}$, but Griffiths tells us the force is $\mathbf{F}=-\frac{1}{4\pi\epsilon_0}\frac{q^2}{(2d)^2}\widehat{\mathbf{z}}$. The only way I see to arrive at this result is to evaluate the gradient at $(0,0,d)$, but the $z$ component of the gradient ($\partial V/\partial z)$ is singular there. And I am not even sure why one would evaluate the gradient at $(0,0,d)$, rather than $(0,0,0).$
I have searched around google and this site, and have been unable to find the answers to my questions. I realize these questions are pretty simple, but I don't have much of a physics background. Any help would be greatly appreciated.