How to evaluate energy conservation for magnetization? A permanent magnet can attract a certain magnetic material mass, giving a certain amount of energy $U_{mag}$.
If we first use this magnet to magnetize $n$ ferromagnetic pieces, we would have $n$ extra magnet and so be able to attract $n$ extra masses. That system would give the energy $(n+1)\;U_{mag}$.
How could we explain this?
 A: Well, be warned: My answer a simple of ridiculous modelling from a much more complicated phenomena. =D. =). Also, lets have this equation in mind: $\mathbf B = \mu_0(\mathbf H + \mathbf M)$. 
If we take a material and apply work on it, the variation of the magnetic field structure because of the work input is:
$$
dw = \mathbf H\cdot d\mathbf B = \mu_0HdH + \mu_0\mathbf H\cdot d\mathbf M
$$
The first term is independent of the material (sometimes called vacuum work). The second term contributes to a change in the material's magnetization vector $\mathbf M$. So, make no mistake: To magnetize a material, you need to apply work. In a closed hysteresis cycle, the later term is non-vanishing, meaning we are dealing with "inexact" differentials, and work is transformed into heat. So, be careful when integrating the later because of considerations of thermodynamic reversibility.
So, if you take a magnet, to magnetize a ferromagnetic material, we now know we need work. The magnet is going to supply such needed work somehow from his own energy. The volumetric density of magnetic energy is:
$$
u = \frac{1}{2}\mathbf H\cdot\mathbf B = 
\frac{1}{2}\mu_0\mathbf H^2 + \frac{1}{2}\mu_0\mathbf H\cdot\mathbf M
$$
So, the magnet will lose its energy by magnetizing the material. The H field is again, independent of the material, meaning, the magnet will decrease its magnetization vector, meaning a weaker magnetic dipole moment, meaning a weaker magnet. Again, what I am telling in somewhat simplistic. In reality, be careful with thermodynamic reversibility, heat, entropy, ferromagnetic domains, temperature, and such things. Sometimes even quantum effects (after all, ferromagnetism is largely a quantum phenomena).
I think the energetic considerations somewhat confusing (my opinion). I prefer to treat in terms of fields. Remember: The induced magnetic field opposes what is causing it. Let $A$ be the magnet, and $F$ the ferromagnetic material. $A$ is magnetizing $F$. So, $F$ has now its own magnetic field which will oppose what is causing it: The magnetic field of $A$. Then while the field of $A$ is magnetizing $F$, we have the field of $F$ demagnetizing $A$. In the end, $A$ become weaker. And in summary, energy conservation laws hold.
A: When a material is magnetized, its potential energy become lower, not higher.  The loss of potential energy creates kinetic energy as they attract each other.
A: We know that a magnetic material like Fe, Ni can be converted into a permanent magnet by aligning their domain or weber element along the axis of the magnetic material. For doing this we have to do some work which remain in the material as potential energy. Every permanent behave like a magnetic dipole and which has a magnetic dipole moment.Now when we brake a magnet into two, its magnetic moment becomes half because number of molecular magnets divides into tow parts.As a result, the Pole Strength of the total magnet is grater than the small part.
Hence the polar strength of the total magnet is approximately the sum of polar strength of the tow part magnets. So, the total potential energy remains conserved. 
