I'm afraid your question is too vague. To answer your question, it would be useful to know the following:

-What is your mobile phase? i.e. what solvents are you using?

-Do you know what your compound of interest is? Is it polar or not? Is it a large molecule or a small one?

-How many other compounds are there in your sample? How similar are they to your compound of interest?

You say you're using a gradient. That means that the concentration of each component of the mobile phase is changing with time. If I understand you question correctly  (and I'm not sure I do) you wish to know if starting out at the solvent ratio in which your compound eluted would make it elute faster. This isn't necessarily the case, and furthermore, the bigger problem would be if it would elute alone or mixed with other compounds.

You wish to optimize the separation. What does that mean to you? Do you wish to make the process faster, or to use less solvent, or to achieve a cleaner, sharper chromatographic peak? Some of those options may exclude the others.

Lastly, if you wish to monitor how your compound behaves using a particular combination of solvents, you may want to try thin layer chromatography, in which you can monitor the elution at different times (granted that you have a method to visualize your compound on TLC, like Dragendorff or vanillin staining.)

Mobile phase is Hexane:ethyl acetate in the order of increasing concentration of ethyl acetate.

Compound is polar and seen to elute out towards high (70%) ethyl acetate.

The sample is a crude extract and I'm not necessarily interested in ensuring that my eluted sample is itself pure, but would atleast like to get rid of the previous non-polar compunds. Frankly, this isn't an issue.

I wish to make sure I can use the same protocol and always discard the previous fractions and use only the fractions in which my compound eluted out. (for the same type of extract). Thus will also make it faster with multiple extractions so I don't have to monitor fractions in eac\h case, and will also tell me how much solvent I need to use for elution.

Can I do that always?

If you are using the same column without making sure the silica gel is clean or at the same hydration state I doubt you would get away with having good reproducibility over more than a few runs. If you are packing a fresh column each time you should get the same elution characteristics, providing your sample is the same (especially if it has the same amount of water and retained impurities). Keep in mind gradient is volume (not percentage) so as you scale up the column volume you will want to scale the volumes of eluent used by the same factor. 

So, what if I have the same column volume with same sample loaded and I increase the volumes of eluent used (of each fraction) ...

Will my compound still elute out first at the same fraction?

No, the compound will elute after a specific volume of ethyl acetate has passed through the column. Increasing the volumes of eluent will increase the volume of ethyl acetate that has passed through the column. If you pass enough ethyl acetate through the column your compound will eventually elute in a lower percentage (of ethyl acetate) fraction. Saying the compound elutes in 25% ethyl acetate for example is not really what you should be looking at when you run a gradient; instead on a given column it is more appropriate that say that the compound eluted after 250ml of ethyl acetate (in hexane) were passed through the column. So 1 liter of 25% ethyl acetate or 1/2 liter of 50% ethyl acetate should both have the same effect in causing the elution of a compound using a step gradient. 

Thank you so much! This is the answer I was looking for. In fact, it makes a lot of sense :)