One of our beta testers is in Oz and sent me this story about Comcast putting a 250G limit on monthly downloads and added this comment:

Here in Sydney we pay $89.95AUD/month for 768K DSL with a 2G limit on combined uploads and downloads. Every 10M over that limit costs $0.25AUD. What in the world are Americans doing on the internet?

Wow. Bandwidth limiting.  And, no, I don't know how you could need 250G/month unless you were watching live streaming moves.  At one/night at 9G per movie, that would about do it, I guess.

25 years ago there were no video stores near me, 15 years ago there were a dozen, and now there are none again. 20 years ago I signed up for AOL (CompuServer was for noobs), 15 years ago they put in volume pricing plans by the minute, 5 years ago they took 'em off, now it is free.

How are ever gonna explain the recent past to our kids?

I was reading through the usual stuff this morning and hit this about a TechCrunch post that

ended up at the Yahoo front page, which leads to around 300,000 clicks per second. That’s a crap load of users.

Yes, it is. Or would be if it were true.

Do Some Numbers

I always like to run the numbers up and down until I get something I can comprehend. I look at Alexa a lot for that - and yes, I know it is imperfect.

Alexa says that yahoo.com got 45,000,000,000 (yes, 45B!) page views last month. Holy moses, did I do that math right?

Alexa also says that google.com got 26,000,000,000 (26B when it's home alone) which is in the same ballpark, but who would have thought yahoo would have more traffic than google?

Wait, there are more numbers.

3% of the yahoo traffic went to the homepage - 1.35B page views.

16% of google's traffic went to their homepage - 17B page views.

Ok, that starts to sound like I understand the world.

(Side note, 20B mail.yahoo.com vs. 3.6B mail.google.com page views. Hmm.)

Drilling Down Into the Numbers

Ok, back to the 300K clicks/second thought experiment. If Alexa is right and 3% of yahoo traffic goes to yahoo.com and another 3% to news.yahoo.com then we have 2.7B page loads/month.

There are (30 days * 24 hours/day * 60 minutes/hour * 60 seconds/hour) = 2.9M seconds/month. (Or you could google "seconds in a month!")

So, 2.7B page loads/month works out to around 1K/second for the home page and the news page.

I Call BS on TechCrunch Math

I dunno, if he'd said 300 clicks/second (which is non trivial!) I'd still have been impressed.

Now, did the story stay on the front page for an hour? That is 18K visitors, which is pretty impressive. Not 300K but still.

Don Is Smarter Than I Am

I asked Don to review my math and he pointed out that I just went the long way around the barn. If Yahoo could send 300K clicks in a second then that works out to 25B/day.

Another way to think about that: everyone in the world has to click on that story. Five times.

Compare Yahoo to 55,000 Digg Hit

I have this article on digg bookmarked because it is so well written and thoughtfully analyzes digg traffic, secondary spikes, etc.

And over a 4 day period he got 55,000 visitors from digg. He compares that surge to an earlier one where he got 0ver 40,000 visitors from digg.

How hard it is to get to the front page of digg versus yahoo news?  Easier, I think. 

Conclusion?

One thing we know is that digg sends traffic (a trickle, but still) forever on stories that get good vote volume. And you do get good feed additions too.

Now, if someone told me that Yahoo sent as many feed-friendly folks as digg I'd belive them, at least provisionally. But once the story is off the Buzz page I would be shocked if you continue to get traffic.

I'd love to opine on link love, but I don't have anything to base my opinion on. I know that is shocking on the internet, but there you go.

I started looking at this article to make sure my BS-O-Meter wasn't defective (it wasn't) but along the way I started to think about the relative size of social media "waves."

The otherwise very very smart Don Dodge posted this gem that gets some bits right:

I talked to a Facebook App developer at the ReMix conference. He told me his app is generating 300 million page views per month. Wow! Then I asked what kind of CPM (Cost Per Thousand) ad rates he was getting. He shrugged and said somewhere between $0.02 and $0.05 per thousand. That pencils out to between $6K and $15K of advertising revenue per month for those 300 million page views. Pretty good for a couple of young hacker/coders with very low overhead, but not the kind of business that commands million/billion dollar valuations.

Basic Assumptions

What I don't know about Facebook could fill Wembly Stadium with just enough room for a Flock reunion, so let's assume:

• This is app is at the median of the top 50 Facebook apps
• The actual income is at the top range of $15K/month
• The top app earns 10X what the bottom app earns and it's a smooth distribution

So, the total revenue for all 50 companies is around $12M a year.

No Money In the Ecosystem

Take these three data points:

• I have a friend who runs a three person shop that supports Cisco routers for a fairly large school system. He billed $3M last year. That is $600K/person.
• We work with a small Oracle outsourcing shop (5 guys in the US and 25 in India) that billed $7M last year. That is $230K/person, but if you cost average that by salary dollar the number is $700K/person.
• There was a local business paper article about a 12 person company that does custom Microsoft VP/Apps/whatever programming and just broke $7M in total revenue or $580K/person.

What's the point?

There is a LOT of money in the Cisco, Oracle, and Microsoft ecosytems. I'm sure all three of these remora companies are too small to even barely come to the notice of Microsoft/Oracle/Cisco.

If any of them were doing Facebook work they'd be #1 with a bullet.

But I bet the 25th largest Facebook Widget Maker can get some 1:1 time with whomever they choose.

Let's review: $12M/year in the Facebook ecosystem and, what, a couple billion for each of the big three? Hmm, well, there may be a network effect from the users of Facebook, but the outside world is starving to death.

Wiser Words

Let's parse one particular part:

Pretty good [$15K/month] for a couple of young hacker/coders with very low overhead, but...

It may sound good to make $15K/month, but that is only a buck eighty a year. Take 50% out to pay for taxes and basic health insurance and you're making $90K. Hope you don't have any hosting or other costs....

So, really, it's krep income, even when you come near the top.

Valuations

The next bit is even more interesting:

... not the kind of business that commands million/billion dollar valuations.

Good lord, Facebook is "worth" $15B to Microsoft and a host of other people and it's certainly going through money like this bubble will never pop.

Let's look at the valuations of our ecosystem companies for a second:

• Oracle - Market Cap=$112B, Rev=$20B, Operating Margin=34%
• Microsoft - Market Cap=$280B, Rev=$58B, Operating Margin=40%
  
• Cisco - Market Cap=146B, Rev=37B, Operating Margin=25%

(By the way, if you aren't impressed with a HARDWARE company like Cisco having a 25% operating margin, you should be.)

So, Facebook is worth 10% of Cisco or Oracle and 5% of Cisco? I think we know that is crazy talk. But if that is the talk, then why is Dan dissing the 300M page views of a Facebook widget maker? Isn't Facebook just a bigger aggregator of page views?

Do Some More Facebook Math

What did Zuckerberg have to say about Facebook revenue on a con call in January of this year:

Revenue for Facebook for 2007 will be $150 million, as has been widely reported. But for 2008, Zuckerberg projected revenue to be increased to $300 million to $350 million.

Currently they have 450 employees, so this year their revenue was $333K/employee, which is not bad.

Next year they plan to have 1,000 employees ... yes that keeps their revenue/employee pretty much flat.

Remember the 25th most popular widget maker pulling in $180K gross/year? He's not looking too stupid, relatively.

How does this math work again?

Summary

There aren't very many people making money adding value to Facebook, it monetizes it's users poorly, and management plans to ramp staff and keep revenue/employee flat this year.

It is a truism that "you have to have the right traffic" to convert.  Convert people to what? Well, it depends:

• RSS readers
• Subscibers
• Ad clicker-throughers
• eBook downloads
• And so on....

At the end of the day it is about converting people to a vaulable asset.

And I don't disagree with any of that because we have been experimenting with traffic a bit ourselves.

First You Need A Goal

When we started this division of company the goal was to build a good base of regular readers by offering targeted and high quality blog articles (*cough*) and unusually useful free tools.  We would then convert some percentage of this group to paid subscription members interested in advanced SEO tools.  We estimated that with 1,000 unique readers a month we could flip the switch on our subscription service and have the business go profitable due to natural conversion.

Getting Started

We did the obvious stuff - we identified where our target market (SEO'ers, e-commerce store owners, etc) congreated and started, politely, building a presence.  This was designed to complement our market research and help us trim and adjust our rollout schedule.  It has also helped us identify some additional places where we can deliver tools and automation.

And so a trickle of really high quality traffic starting coming in, with little increases here and there and no drops.  Reassuring!

Moving Stepwise

In the early days we got a few Stumbles and a bit of a Digg rush, but the base of our readers didn't really increase very much.  Duh, no tools and not much content.  So we decided to keep up our social authority efforts at the SEO watering holes and ignore traffic for a while.  This let us concentrate on building out the initial SEO tools and making modest announcements about our intentions.

For example, our Free Yahoo Store SEO Analyzer is an extremely poweful, Yahoo store specific SEO tool.  This brought in a lot of Yahoo merchants and got a lot of attention from the SEO community focused on that market.

And several thousand people have used our Digg Friend Finder to increase and improve their Digg friend network.  Presumably a fair portion of these folks are people using social media for SEO fashion.

Each of these tools bumped up our regular visitors, as we expected.  We suddenly had both content and tools, and people were responding to it.  I discussed how powerful this was in my post on SEO Tools vs. SEO Content.  Basically tools deliver a lot of repeat traffic and to the extent that they are unique they provide stickiness.

Time to Get Some Bulk Traffic?

People love lists, so we released lists in the last two weeks:

• 117 SEO RSS Feeds Worth Reading
• Downloadable List of 2,162 Social Bookmarking and Networking Sites

BAM! as Emeril would say (sold to Omnimedia for $45M, BAM indeed!), thousands of new people show up.  And keep showing up in a reducing curve over the next week.  Which is exactly what happened before, so we were curious to see if the outcome would be different.

What's New Pussycat?

Our RSS readers took a big jump.  And the number of people navigating to our site directly from bookmarks (that is so 2001!) jumped considerably too.  We're watching the waves of Digg and Stumble traffic subside, but unlike in the early days, we're seeing more readers washed up on shore: there is enough here to keep them coming back.

Google SERP Inflection Point

I know we all live in the google matrix and so we have to try to deduce causes from the visible effects, but there is no question that our google traffic started jumping as well for the keywords we're most interested in.  We notice that, right before we released the list that we'd gone from pariah to PR3 in three months.  And we had been seeing some additional google traffic so that change didn't totally surprise us, but it was really almost a week after the first list that we saw the dramatic increase in google search traffic.  So clearly there was some confluence of events going on in google-land and it was some aspect of the traffic/linkfest that tipped us over.  We'll probably do a few more Stumble/Digg things in the next month or so and see if the google traffic increases after that - after all, it might be a mighty coincidence.

Monetizing

We've reviewed other tools and criticized them for not having a clear monetization strategy, terms of service, or a privacy policy. We've had several people tell us that we could be doing a much better job of monetizing rather than using Adsense (horrible) and Amazon Affiliate links. They're completely right, and we're going to be experimenting with some other monetization strategies on our free and freemium tools. But we've been very clear that our real goal is to build up a community that will be receptive to paying for a subscription for our premium tools when we roll them out. We figure that if we offer free and freemium tools that are a cut above other free tools ( i.e. they actually work and produce actionable results) and we offer some things for free that other sites require payment for, when we launch our premium service people will see the dramatic increase in value and will sign up.

As Time Goes By

And obviously that sticky effect should continue to improve as we have more tools online and more content.  If we kept, say, half a percent of the Digg/Stumble traffic as regular readers in this wave, maybe we can get that up to one percent in a few months with some of the new tools we've got in test.  Since we're building SEO tools and not MySpace widgets I wouldn't expect that the upper end of capturing readers will ever be very high, but that is ok - if you're running a gossip site you need 40K readers/month, but our targeted sales community is much smaller and more cohesive.

-----

I ran across that Buck Converter circuit graphic quite some time ago while I was looking at table saws and I have been looking for an excuse to use it!