Portrait of the emerging business as an older dog

Xbox, one of MSFT's most expensive "emerging businesses" to date, is now some 5 years old and no longer a young dog. In 2001, when the original unit was launched, there was already concern that MSFT was being far too optimistic about its prospects:

When Microsoft expects to break even is moot. Henry Blodget, Merrill Lynch's celebrated internet analyst estimates it could take as long as five years for the company to recoup its investment - essentially the life-cycle of the product.

In the same article, Robbie Bach, Xbox division head, took issue with Blodget's assessment, saying (among other things):

Microsoft does not go into businesses that make losses for five years. I have my own profit-and-loss account.

"Five years" have now passed, and the Xbox division has made annual losses over that entire period (albeit that it did have one profitable quarter on the back of the Halo launch). So, while props to Blodget are due, even his assessment ended up being recklessly optimistic.

Before determining how long it will actually take MSFT to "recoup the investment", let's try and determine what that investment has been. That's more difficult than it might appear, since MSFT didn't break out segment P/L in the earlier years. However, according to this article in '05, Forbes estimates losses for the first 4 years at approximately $4B. Add to that what we know from MSFT's subsequent earnings reports (which do break out segment P/L), and that total rises to at least $5.3B and possibly higher (depending on what Forbes included in their calculation). So, minimum $5.3B investment to date and counting, since the division still isn't profitable.

Now to the return part. How much profit could MSFT potentially generate from Xbox? Well, we can't look to MSFT for input because they've never bothered to even roughly estimate it publicly. However, we can look to SONY, the segment leader, for some idea of what's possible. From what I can see (and admittedly I couldn't locate the source I recalled), in their best year ever, SONY appears to have generated some $1B in profit from their PlayStation-led gaming franchise. Let's be hugely generous to MSFT by assuming they can supplant SONY tomorrow, and get to that former level of record profitability (SONY currently being expected to lose as much as $1.7B on the back of the PS3 launch). At that rate of return, it would take another 5+ years to breakeven, for a net total of 10+ years just to recoup the investment - twice what Blodget predicted, and more than twice Bach's estimate. But wait, it gets worse. In reality, while MSFT might ultimately be able to generate more creative options for monetization than SONY (TBD), assuming $1B in profits, or even half that, any time soon, has a likely probability of zero when rounded to the nearest decimal place. So, actual time to recoup the investment, assuming it ever occurs, will likely be measured in decades. Keep in mind that Ballmer's characterization of this business, as recently as May 31, 2006, was:

I'll tell you that was one of the great creations of shareholder value of all time.

My question would be: Based on what evidence?

All of which raises the question of what MSFT's broader strategy wrt emerging businesses is, and how success is measured. Are they worthwhile standalone efforts, meant to be profitable in typical industry 3-5 year timeframes (or less), but MSFT just can't execute properly? Or are they always, at least in part, defensive plays aimed at protecting the existing crown jewels and only justifiable (financially at least) on that basis? I think the answer is a combination of both, which explains why MSFT's "investments" continue to be viewed by many externally as hopelessly unsuccessful, especially when they weren't sold to the street or shareholders on that basis. Net net, MSFT needs to start rationalizing its efforts and ensuring that if new ventures are being touted as making standalone financial sense, that they in fact do so, and over timeframes that most would consider reasonable. Otherwise, the street and investors will continue to focus on the order-of-magnitude better risk/reward investment choices of competitors (e.g. AAPL's in Ipod, GOOG/YHOO's in search/advertising), and conclude that Ballmer and team are either inept, bullshitting about the true purpose of these investments, or working on timeframes that even the longest of "long-term" investors would balk at.

Update: note that in the linked webcast, Ballmer infers that Xbox losses hadn't yet reached $4B (as of May '06). Specifically, he says:

Let's say we go negative $4 billion before we start going positive

That's at odds with Forbes' estimate and seemingly MSFT's own reported financials, but possibly true if there's some discrepancy in what H&E losses were directly attributable to Xbox (let's ignore the possibility that losses may well have been understated by charitable interpretations of what costs to bill to G&A and R&D versus the unit). Suffice to say, that $4B vs $5B (or somewhere in between), while substantial, doesn't change the underlying analysis that materially (i.e. minimum payback period becomes 9+ years vs 10+, and likely still decades under any more-realistic scenario).

Update #2: And the expectations for Xbox profitability get pushed out even further:

When asked how long it will be before Microsoft starts making money on Xbox, Bach replied, "To be clear, we have said that in fiscal 08, entertainment and devices makes money. That’s not exactly Xbox. We don’t break profit down by business. And there are parts of entertainment and devices that make money. Xbox doesn’t. Xbox has to make significant progress to enable E&D to get there."