My sites are focused on localized topics, like news and weather. Each site focuses on a different area in the United States, but all within a fair proximity of one another. So 99% of the traffic for each site will come from people that live within that specific region.
(We actually block most non-US IP addresses at the firewall, since we do not target any traffic outside of the US.)
The one site is doing great, but the potential traffic for the site is limited by local population; we've reached a 100% penetration in that region, so we'll never have more traffic than we have now. This is why we added new sites; we cover more cities, and we increase our potential. The problem, though, is that the new sites have a very low RPM as compared to the larger site.
Obviously, my employees work to maintain the larger site, as that is the majority of our income. But as the company owner, I'm trying to find ways to make the other sites grow, too... with the hopes that at least some of them will become as large as the original.... or, hopefully, larger.
By marketing, though, you're 100% correct about paying for traffic, but it's a catch-22; if no one knows that a site like mine exists for their region, then they don't search for it, and will never find it. So we do pay for Google ads (with fair results) and Facebook ads (with terrible results, if any), targeting each specific region, but in the hopes that the people that click on the ads will return to the site regularly.
That can be ridiculously expensive, though. At an average of $2 /click, we pay $3,000 to reach 1,500 new users. If every single one of them returned to the site every day and looked at 8 pages a day (which is about a normal rate for us), and an average RPM of maybe $1.50, it would take about 6 months for us to break even on the campaign! But all 1,500 don't return daily, so that's obviously a losing strategy. Worse, some areas are in higher demand, so that PPC can be up to $10!
We also incorporate things like billboards, giving out promo items, sponsoring local events, etc, but that's very time consuming. Covering 60 regions means that we could spend a maximum of 6 days in each region per year. Currently, we're doing exactly like you suggested (focusing on 3 at a time), but at the rate we're going, by my count it will be about 20 years before they're all effectively marketed.
Technically speaking, all of these sites are really a single site, and the domain acts as a mask; eg, www.keyword.com is a mask for www.lousydomain.com/keyword . So in theory, I could redirect every user to lousydomain.com/keyword and possibly double the RPM, but then I lose the SE-friendly and memorable domain, etc. Worse, I haven't been able to find a good umbrella domain; the ones I have simply aren't that memorable and do not test well.
So that brings me back to the original question. Other than consolidating everything in to a single site, can you think of any other way to maximize the RPM on them individually?