For the last few weeks I’ve been hoping this meme would die. I saw some commentary on some blog by Jeremy Liew (not surprised he turned it into a post). Now, Jeremy’s a wickedly smart guy and definitely knows his numbers. He’s also a VC at one of the VC firms I have nothing but respect for: Lightspeed Ventures.
Which is one of the reasons I wanted this to die. I basically hate disagreeing with wickedly smart, incredibly successful people who I have a huge amount of respect for (including Scoble, PageRankWhore, Screenwerk, Ross Dawson and Hillel).
But, now’s the time.
First, let’s summarize Jeremy’s point (see his original post, as well as followup 1 and followup 2):
- Building a media business to scale is hard.
- “Scale” is defined in 50M$/year in revenue.
- Getting to 50M$/year in revenue is basically impossible.
Okay, Jeremy didn’t say impossible 
What he did say is:
- General audience stuff won’t get more than 1$RPM (revenue per thousand pages, including all display advertising). This would require 4 billion pages per month.
- Demographically specific stuff won’t get more than $5RPM. This would require 800 million pages per month.
- Huge, in-demand stuff won’t get more than 20$RPM. This would require 200 million pages per month.
First off, if you own a media company or site that isn’t able to break 2-3RPM on even the most general audience stuff, give me a shout. At b5media, we currently average between 2-5RPM on general audience traffic and far more on specific demo or high-mand content.
Which really brings up the issue. Jeremy’s figures are only true if you rely either entirely on Google AdSense or on generic deals from the big few ad networks. They don’t hold true if:
- You do more than 25M AdSense impressions per month (you get a better cut, exclusive deals, etc).
- You work with a boutique ad network (like Federated Media).
- Are able to get some remnant inventory providers (which can go from 0.50-2$/unit CPM, giving you an RPM of 1-4$ with just 2 units per page).
- You do any non-performance/metric advertising (like sponsorships).
- You do text links.
- You have any internal sales team at all (which’ll sell those 1$ units for 2-5$, giving you an RPM of 4-10$ on even the most generic traffic).
Basically, Jeremy’s figures won’t apply to any media company that has a serious execution strategy. The core of this is that the numbers Jeremy’s mentioned can very easily be translated to the following numbers, once a company passes about 10M pages/month in “real traffic” (ie: seen by humans):
- For generic traffic, expect a 3-5RPM. To get to 50M$/year in revenue would thus require 800 million pages per month.
- For demographically specific stuff, expect an RPM of 12-15. You’d need about 300M pages per month.
- Huge in-demand areas like cars and sports can net you an RPM of 40-50. You’d thus need about 100 million pages per month.
Now that’s not to say these numbers are easy. But a media company that balances the above 3 properly, does sponsorships, syndication deals, content licensing, text links, feed ads, etc could potentially achieve 50M$/year in revenue on about 200 million pages per month.
Again, not a small number, but neither is it impossible.
In fact, I’d venture to say that it’s not as hard as folk (like, y’know, the New York Times, for goodness sakes) would like to make everyone think.
Again, NOT EASY. But here are some thoughts (I haven’t built a 50M$/year business, so take this wiht a truckload of salt):
- If you’re at 10M$/year in revenue, you should be able to acquire a couple of companies per year in order to increase your growth rate by at least 2x.
- As you hit certain milestones (10M pages/month, 25M pages/month, 50M pages/month), entirely new advertising opportunities open up. Anyone have any idea what the RPM is for TechCrunch? Somewhere near 750$.
- As you build a major media company, you start to build infrastructure that you can license to others or that can be used to rep others. All of a sudden your core traffic doesn’t matter as much.
Anyways, core of this is: if your media business is actually in the ranges Jeremy’s talking about AND you’re generating tens or hundreds of millions of pages per month, give me a shout. There’s a whole hell of a lot more you could be doing.
Hell, if your site is generating 1M pages/month and you’re getting the kinds of numbers Jeremy’s talking about give me a shout. There’s STILL a whole hell of a lot more you could be doing!
Oh, and, Jeremy, if your portfolio companies are hitting these walls, feel free to give me a shout as well
March 19th, 2007 at 11:07 pm
Im a firm believer in the ‘No Money FOR Advertising”…*cough* http://adgridwork.com
March 20th, 2007 at 1:26 am
Whoa! $750 RPM for TechCrunch?! Are people really paying close to a dollar for just one human view? Seems like an awfully high number.
March 20th, 2007 at 3:45 am
Jeremy,
First, thanks for the kind words.
I agree with a lot of what you say. At “middling size” CPMS can be higher than what I laid out. But as you get to real mass scale, you start to run into difficulty in maintaining CPMs. Take Myspace for example, which according to Media Metrix did 40bn page views in January. Given that its media metrix, this number is almost certainly low. Forbes estimates Myspace monthly revenue at $28m. That suggests an RPM of around 70c. The problem is that when you get big, CPMs start to dip again as you start reaching the same users over and over.
Your overall point though is true and I don’t dispute it. There IS money in online advertising. Plenty of people can make a very nice business out of it. But if your aspiration is to build a $50m revenue online media business, then as you conclude, it will be hard!
Cheers
J
March 20th, 2007 at 5:50 am
That’s a lot of shouting, Jeremy.
I’ve believed for a while that depending entirely on traffic and blog network-type ad deals will eventually pay the mortgage and for a modest lifestyle. But nothing special that you couldn’t make journoing in the print world.
As you say, it’s the wider offshoot business that brings in income, at least until you’ve reached a tipping point, maybe 2 – 3 years in.
Until then, you need to survive, hang in there, and keep pressing the buttons.
Stoicism is the most under-rated entrepreneurial quality.
March 20th, 2007 at 9:39 am
Jeremy: Thanks for stopping by. Great point on achieving sustainable CPM’s at the greater-than-200M pageview level.
I guess my perspective on the whole MySpace thing is that it’s a problem they created for themselves. MySpace is just like forums: lots and lots of pageviews per user. Way above industry norm. They’ve trained their users to click and click, which has driven the efficiency of ads down, thereby driving down prices.
In addition, because of the glut of inventory, they can’t sell fast enough at 2-3CPM to fill it up, so they’ve reverted to selling more inventory for less.
You can actually buy campaigns on MySpace for as little as 0.10CPM. Reachin 1M people for 100$. Craziness.
Realistically, these are problems they’ll solve. After all, MySpace knows more about its users than anyone else (even Yahoo/AOL who are huge into datamining).
I’d expect MySpace and other massive social networks to begin leveraging this knowledge in the next few years to start inching towards Yahoo/AOL’s 10RPM average. It’ll probably take 5-7 years to get there, but they should be able to do it.
Greg: The 750$ was back-of-the-napkin math on revenue of about 100K/month with about 1.5M unique visitors. It’s probably closer to 250$RPM. But still, that’s just incredibly high. Possibly one of the top 10 highest on the planet.
March 20th, 2007 at 9:41 am
My impresion of Jeremy’s article was that he wasn’t so much talking about media networks like b5, but moreso a single entity with ppc and cpm ads as it’s single revenue stream. Network-type media companies are a different animal, like you showed, because of the shifting RPM’s per niche and per reach, as well as their buying power.
I think the “long tail” theory is an excellent approach to a media network because as you expand further into far reaching niches, you may not get the same traffic, but in some cases, the RPM’s are higher. When (if) traffic increases in those niches, revenue increases exponentially. If traffic doesn’t increase, you still have a decent revenue-producing site (hopefully) with low overhead.
March 20th, 2007 at 9:50 am
Steve: Yeah, I know. Really I wasn’t arguing with Jeremy’s core points, but against the growing meme that’s basically saying “aha, you want to do an ad-based business? Well Jeremy Liew says you can’t!”.
(I’ve actually heard some VC’s say this, minus the exclamation mark and imaginary finger thrust into the air)
The truth is that Jeremy’s right: most small media companies do face these challenges. And many mass-scale media companies do as well.
I just don’t agree with the extrapolation the MSM and many bloggers are making is all