At one point on Nov. 20, 2008, you could have bought shares of Web content delivery network (CDN) operator Akamai Technologies (NASDAQ:AKAM) for $9.25 a share. Today, you'd have to pay almost twice that.

It's tough to blame the bears and skeptics. That month, Akamai announced a 7% workforce reduction, leading speculators to conclude that increased competition from CDN pure plays such as Level 3 (NASDAQ:LVLT) and Limelight Networks (NASDAQ:LLNW), as well as new entrants AT&T (NYSE:T) and Amazon (NASDAQ:AMZN), was hurting business.

Even I was nervous, and I'm the guy who convinced Fool co-founder David Gardner to add Akamai to the scorecard for our Motley Fool Rule Breakers service in the May 2005 issue. But I also refused to sell.

"If bandwidth and hardware capacity were the only issues, if network design didn't matter, if software and services were irrelevant, then Akamai would have been dead long ago, killed by well-funded start-ups like BitGravity. You're wrong, Mr. Market. Again. Akamai isn't going anywhere," I wrote at the time.

A turnaround, delivered
My patience is paying off. Shares of Akamai surged last week after the company's reported results crushed Street estimates. My Foolish colleague Anders Bylund has most of the key numbers here.

Two he didn't mention would be 50 and $24,000. Akamai added 50 net new customers – including those acquired via ad network Acerno -- in one of tech's worst quarters since the dot-com bust. Average revenue per customer (ARPU), meanwhile, rose to $24,000, up 2% sequentially and 4% over last year's Q4 -- though that number will decline when Acenro's revenue contributions are included next quarter.

Gross margin fell due to pricing pressure but remained above 71%. Even better: Return on invested capital -- a measure of management effectiveness in creating value -- improved once again for 2008, to 8.1% from 7.9% the year prior. Fourth-quarter ROIC was 8.3%.

For Akamai President and CEO Paul Sagan, it was a redemptive report. "I think you're seeing the difference between PR and results," he said in an interview with me last week. He's referring to reports that upstart EdgeCast is winning big deals, including a contract with Deutsche Telekom (NYSE:DT).

Sagan used stronger words in responding to rumors that Apple (NASDAQ:AAPL) had chosen to shift some business to Limelight. Quoting from his comments to investors during the earnings call:

We continue to have a very close relationship with Apple including support for all of iTunes. Given our respect for our client's confidentiality needs, we won't be commenting further except to note that we recently extended our long-term relationship for another multi-year term. [Emphasis added.]

Should you buy these numbers?
Statements like that matter to shareholders. Numbers matter more, however. Some of Akamai's best numbers can be found on its balance sheet: $327 million in cash and short-term investments. An additional $440 million has been earmarked to long-term investments, but $250 million of that is stuck in illiquid auction-rate securities.

Cash also continued to flow, but not as organically as you might think. Of the $343.5 million in cash from operations reported for 2008, $57.9 million was derived from stock options exercises. Another $80.6 million stemmed from deferred tax benefits, credits for net operating losses that substitute for cash tax payments.

You might say that right now, Akamai is on the juice, using financial steroids to artificially boost its cash flow.

That's not as bad as it sounds; Akamai isn't A-Rod. There's nothing illegal or unethical about these steroids. They simply overshadow the effects of Akamai's core operations in generating cash. And they generated plenty: $89.6 million, or $0.48 per share, by my math -- good, but also well below the $228.1 million that actually flowed into Akamai's coffers.

That'd be a concern if not for two things. First, capital expenditures are higher than normal due to increased competition. Higher expenses mean lower cash flow. Second, Akamai's distributed model is holding up well. Here's how:





Incremental revenue

$154.5 million

$207.7 million

$145.6 million

Servers added




Revenue per new server




Capital expense

$115.4 million

$81.4 million

$56.8 million

Capital exp. per new server




Months to repay per server




Sources: Capital IQ, Akamai press releases.

There are two ways to read this. You could conclude that Akamai is trending poorly in terms of earning a return on its deployed servers. And you'd be right; today's Akamai needs roughly nine months of revenue to pay off a new server versus just five months in 2007 and 2006. Not all that encouraging, is it?

Not without further data, no. Yet there's more to consider. Limelight has a long history of burning cash, and Level 3, while cash flow positive, is saddled with billions in debt. That Akamai survived sweeping price cuts and still managed to earn a sizable capital return -- at least three months of big profits per server -- speaks volumes about the financial strength of its business and the discipline of its managers.

That's why Akamai has doubled since November, and why seasoned investors are still buying today.