At first glance, it seems like the best option to create content campaigns is to start with placement campaigns.  This is because you get to pick the exact sites your ads appear on.   However, it's important not to ignore keyword-based campaigns.

In fact, I almost always start with keyword-based content campaigns. Each keyword (usually around 50 and at the most 100) goes into its own ad group.  The keywords shouldn't just be taken from search, but should be double-checked to make sure they describe the theme of your site.

After you've done this, let your new campaign run.  Then, every so often, check the placement performance report.  It tells you what sites Google has decided were relevant and ran your ads on.

I try to do this every 100-200 clicks, depending on the market.  So, this might be checking your placement report every other day, depending on your budget.

The reason this is so important is that the content network is massive and if you only use placement targeteting, then countless opportunities get overlooked.

Once inside the placement report, I then go through the sites with clicks and if one is relevant, I place it under a new placement targeted campaign.  At the same time, I add it as a negative in the keyword-targeted campaign.  In addition, if a site has clicks and is NOT relevant, then I also add that as a negative site.

When looking at the Placement report, prioritize sites first by the number of conversions... followed by clicks... and finally impressions.  This means your conversion code needs to be installed when you begin.

By doing the optimization process this way, over time, you will have built-out a detailed placement campaign with winning sites and have tested ads from the keyword-based campaign and early site targeted campaigns you can use in future ad groups.

To summarize, the key take away is this... The content network is HUGE... and to get a handle on it and make sure you don't overlook opportunities... you cast a wide net initially and then optimize it frequently.