I gave a "main tent" talk at last week's SIMChina conference in Beijing (BJ). Nice, relaxing trip: BJ is only one hour from Qingdao (QD); QD conveniently intersects three of China's four largest ITO centers. Billed as the largest conference on IT outsourcing in China (trust me, it wasn't), I want to touch upon some lessons for Westerners.

The first talk by someone other than a PRC government official was a talk by the head of a software industry association in the States. His talk was on software as a sevice (SaaS). In his talk, the association director touted the wonders of SaaS. Listening to him, you might even think that SaaS will solve the problem of world hunger. I'm bullish on SaaS, but he missed the key point: SaaS doesn't have much of a future in China, sans MNCs. His talk demonstrates the typically flawed reasoning of many Westerners and Western firms.

One could say that since there's an opportunity in the States, there must be an opportunity for outsourcing to the States. This may very well be true in many cases, but certainly not in all cases. And this is one of those situations where there is a huge disconnect in the market. What the association director failed to realize is that firms in China are very reluctant to have their data hosted. Second, even if hosted data wasn't an issue, the method of doing sales in China is very different than in the States and doesn't fit an enterprise-class SFA solution. By necessity, guanxi must remain an art and a mystery, not something to be tracked. "Click on the blue icon for payments to municipal officials; click on the green icon for payments to provincial officials." Not likely. For these two reasons, don't expect to see hosted SFA in China. And if there's no opportunity to build expertise in SaaS solutions in the domestic market, it's not likely going to fly as a viable outsourcing sector. Of course, there's another major problem: There isn't a lot of money for solution providers in this space. Some SaaS vendors offer huge incentives, but at the end of the day, it's not really a viable market for solution providers, especially China's solution providers. This is not to say that a firm or two won't try to take advantage of the incentives: Many firms have no strategy to speak of and are purely opportunistic. But the better firms, well, know better.

Another lesson is to be wary of efforts to attract foreign direct investment and all the variations on this theme. There are officially 29 cities in China with skin in the outsourcing game. But only a handful are viable options for U.S. software and services firms. Some of the presentations were downright goofy, like the head of Ningbo's Software Park talking about the bridge (a real, physical bridge) which will be built over the next few years to cut the travel time to Shanghai to merely two hours. Well, this is a good enough reason for me to set up shop in Ningbo. In all fairness, Ningbo is a beautiful city and it's even a logistics hub, but it's not a top spot for IT outsourcing.

Now to my presentation. My presentation at SIMChina was focused on management issues, versus my meatier presentation later this week at CSIO which will be focused on marketing issues.

Solution providers in China tend to emphasize the targeting of American and European CIOs. This is rather odd since it's not the best strategy, but given that it is the primary focus for U.S.-facing outsourcing, it's necessary to address. Many firms believe that they can do this by remote control, i.e., that they don't necessarily need a physical presence in the States. Others believe that a 28 year old engineer-turned-sales person staffing a shared secretarial office is good enough. Few are willing to make the necessary investment in staff: Sales, project management, and technical support. Matter of fact, when choosing China's best in serving U.S. end users, only a handful have adequate staffing in the States.

Another hot area is contract software development. But the only China firms with success in this space are those which are not legally China firms, i.e., they're really U.S. firms with operations in China. They are also headed by Americans, usually ethnic Chinese. Why? IPR, of course. This is also an area where Hong Kong is making a play; HK gets IPR.

The domestic providers don't want to acknowledge that the primary reason for going with a China solution is cost. It's all about cost. All the surveys show this; cost is the only variable which shows up in every survey. Technical expertise ranked as the leading factor in this year's ITO survey by DiamondCluster. Yet, few domestic firms are willing to commit the resources to develop a specific technology or portfolio of related technologies. It's all about breadth and no depth. But U.S. CIOs want depth. Clearly, another disconnect.

Another oddity is that domestic firms have no idea what metrics to use in managing their company. Although there are many possible metrics, the best four are revenues, revenues per employee, headcount delta, and utilization. Some may argue that the blended hourly billable rate is better than revenues per employee. In general, I agree. But it's too easy to fudge the blended rate, so I prefer revenues per employee. Utilization is the funniest: When I query the CEOs of domestic providers, their answer is usually 100% (sometimes they even joke that they're running at 110%). Wrong answer!! When they seriously think about it, it's usually 90+%, but this is still way too high. The optimal seems to be around 75-79% and some of the Indian globals can run as low as 65% when there are a lot of new hires. But the utilization rate says something: Few firms in China are willing to train their employees. Of course, they have a pretty good reason; many employees jump ship once they receive training. Even People's Daily chimed in that this is a serious problem, citing that a "backward education system resulted in students weak in programming capability, while enterprises employing them have been unwilling to provide relevant trainings." (A bit of Chinglish, but you get the point.) Joseph Hsu, CEO of Symbio (and coincidentally the speaker before me at CSIO this week), noted that "there are many computer scientists in China, but not many software engineers."

I pointed out that some of the best opportunities are in those areas that are not mission critical to the U.S. client, but where there is a huge disparity in wages. I've said it before and I'll say it again: Programmers in the States are an endangered species (sans defense-related work). Java programmers in the States make at least $75 per hour; in certain cities in China, they make $3 per hour (even less in some places - but you're risking your life if you go there). Yet they're billable at $15 to $20 per hour. This is clearly a win-win, well, not for American programmers. This isn't so simple in that there's definitely a quality and rework issue. But for a lot of custom application development, the numbers make sense. A solution provider in China can make at least $15,000 per year in operating profit from each programmer, even at 65% utilization. Not huge numbers by U.S. standards, but not bad by domestic standards, especially when a firm has at least 100 billable programmers. Trust me, these numbers scale very favorably. Noting the potential for significant cost reduction, many of all ilks in the States are thinking about captive or BOT (build-operate-transfer) models. Yes, these are perfectly valid. But be prepared to struggle with all sorts of political, economic and legal issues. Matter of fact, I find that I'm spending more of my time dealing with political, economic and legal issues versus vendor evaluations.

I also warned domestic firms to avoid CIO surveys for the purpose of setting strategy. They're great for getting an idea of how a firm's technology portfolio may fare over the next few to several months, but they're too variable for setting strategy. Another thing to avoid are forecasts by IT advisory services. The Kensington Group has found that most forecasts are so awful that one would really, truly get better results with a dartboard and dart-throwing monkey. (This is not a common expression in China, so it usually generates a lot of chuckles.) That's not to say that the analysis provided by the advisory services is worthless, only the forecasts. Fact is, the forecasts are really for show; nobody really cares about the forecasts when we're in briefings.

In lieu of CIO surveys or bogus forecasts, I recommended tracking IBM Global Services (IGS). From what I've seen, IGS will be the big winner in China. Morgan Stanley recently went so far as to title one of their reports, "When IBM sneezes, IT services catches a cold." A variation on an old theme, but just as true today. I also warned the domestic providers that they better watch their back. Quoting from the 26 May issue of Morgan Stanley's India IT Services Daily, "... it is likely that the largest exporters of services from China could be the Indian and multinational companies, in a few years." Bingo!!

Finally, I told everyone to go read one of Michael Porter's books. No, not Competitive Strategy or Competitive Advantage, but his Competitive Advantage of Nations Competitive Advantage of Nations (or, the related article published in HBR, available in digital format ). Bottom line: It's all about manufacturing - and out of the top 100 solution providers in China, only one lists manufacturing as its key vertical. Can you scream, "Disconnect!!"

David Scott Lewis is the president of IT E-Strategies, a consultancy focused on IT outsourcing opportunities in China, and the editor of the David on IT Outsourcing in China e-newsletter.