It was similar like this when I was using Dragon NaturallySpeaking to compose my blog posts:

 Here's an example of what I used to do:

First draught dictated into DragonPad with my cordless Bluetooth headset connected to Dragon Naturally Speaking. (Occasionally, I would use the dictation box, but it was just for shorter sections since you couldn't save from the dictation box.)

After I completed my draft, I would check through it again to make sure there were no transcription problems that I had overlooked the first time around. After that, I would calculate the word counts and determine how accurate the voice programme was for the particular post.

I would then copy and paste my work into a blog post (using Internet Explorer, Google Chrome, or another similar browser) and make some last edits, such as adjusting the formatting and adding hyperlinks, before publishing it. If the article was particularly lengthy, I would divide it up into a series of pieces, such as this one about using the new iPad 3 in the classroom or this one on the many possibilities for a student response system, for example.

When I was finished, I would push the publish button, and the post would go online.