It isn't at all impossible to get $10-$20 CPM for basic standard banner adverts, then you can charge higher rates for, as an example, video pre-roll adverts, and you can work on bespoke marketing campaigns to complement banner advertising, which again can provide higher profit margins.
The above is from personal experience. For example, the company I've been with for the past few years (essentially since they decided to expand away from being a single website) manages to employ a handful more staff than Reddit on 20m monthly page views.
If your ad space will be profitable for the advertiser, become your own advertiser. Pivot to generate more direct revenue. I'm doing that on a number of sites of mine and it's working well.
Example: Google is getting into the loan lead generation business (http://adwords.blogspot.com/2009/10/introducing-adwords-comp...) and the airlines business (http://www.google.com/press/ita/) because they know that by letting advertisers get access to their audience, they are leaving a lot of money on the table.
I see their trend towards trying to get more of the pie continuing over time to expand into banking, movie tickets, car rentals and a gazillion other industries until they start coming close to the antitrust lines because it's a huge source of revenue.
With B2C, your revenue grows over time as you grow your userbase. It's pretty predictable, and not too risky.
With B2B, it's more binary. You might go months with 0 revenue, then get some deal and suddenly be profitable. But then there's added risk, as losing a single client can have a massive impact on your bottom line.
First, neither is easy. But "B2C is predictable and not too risky" is patently absurd. It requires attracting users, keeping them and figuring out a way to monetize them in some economically sensible way. And this is while competing with 1000s of other services trying to attract attention as well. VC portfolios are littered with carcasses of B2C startups that never caught on (and those are the well capitalized ones)
With regards to B2B, I think you're describing B2B in a consulting context vs. B2B SaaS. In SaaS if done right, you have people with a pain and with real money to solve that pain. As a result, you don't require millions of users to make good money.
Of course, one is not better than the other. Depends on what you know, what gets you excited, etc.
It doesn't even have to be a niche. If you provide basic infrastructure stuff (email, hosting, IT, whatever) there's billions to be made on business customers.
It really just comes down to marketing and sales prowess and the know-how to deliver stuff for less than the big players.
So if $100mm = 100 billion page views/yr. Then by that logic reddit should be making ~12 million a year, which doesn't seem to be that case.
And that would be in addition to the subscriptions people are buying.
Still, your point is right. You can't just have a blanket number that works for any site.
http://www.quora.com/What-is-considered-a-significant-number...
