Final export doesn't reflect mix?

What you hear in SONAR does not matter, because what you hear depends on volume settings and such. What matters is the readings of the volume meters (RMS, not peak) and the look of the waveform that is created.

If you import a commercial track in the same project, and A/B compare them, you'll see the difference.
Bounce your project on one stereo track inside the project, and compare the audio waveforms to the commercial one.
Also compare the frequency curves with an analyser.

If you have not done compression and volume automation before the final export, it's very unlikely you can compress the project in a satisfactory way in one go with Boost 11.

Getting the "commercial" loudness is not a simple thing. Every decision you make during tracking and mixing matters. Thinking about it after the final mix is too late.

There's a million posts about mastering and compressing on this forum, and googling can find you a million more, then there are books like Roey Izhakis "Mixing audio" and books by Bob Katz, which are very, very recommendable.

You also might want to have a look at your routing inside Sonar. What sort of bussing structure are you using?

If you have multiple tracks including drums, guitars, vocals, keys etc, then ideally you will benefit from having your tracks output to separate busses. But most importantly, you  must also set up a master bus to which all your sub-busses feed.

The point being that NO track is feeding directly into your Main Out/soundcard because any such track will be bypassing whatever processing you have on your master bus, and your export will never sound the same as what you're hearing when you play the project.