Introducing Shanzhai Couture, a phenomenon that makes it cool to reject pricey fashion. This is starting with knockoffs, but it won’t end here. Just as China’s mobile handset industry began by making knockoffs and look-alikes but is quickly evolving beyond that, the Sino-fashionistas will use counterfeits as a starting point, but gradually move beyond look-alikes to create their own styles and labels.
Letting the Sunshine In
Much of my March was spent working with clients who are thinking through some of the issues facing the growing data center market in China. For the uninitiated, a “data center” is a place that houses anywhere from one to tens of thousands of servers. This blog sits in a data center, your bank information sits in a data center, there are a lot of them, and these places are growing.
Little wonder. One delightful quote from Smithsonian.com suggests why.
“From the year 2003 and working backwards to the beginning of human history, we generated five exabytes–that’s 5,000,000,000 GB – of information.
By last year, we were cranking out that much data every two days.
By next year, we’ll be doing it every 10 minutes.”
That quote was from two years ago. Draw the curve in your mind, and you can figure that, conservatively, today we could be generating five exabytes of data every five minutes. Not all of that is going to sit in phones, laptops, external hard drives, thumb drives, or those little SD cards that we stick in our digital cameras. Much of it has to sit in data centers.
The Great Heat Sink
Which is fine, until you consider that data centers suck energy the way blue whales suck krill: in massive quantities, and with large amounts of undesirable waste at the end of the process. In the case of data centers, that waste comes in the form of heat, which then demands more energy to power cooling, which in turn generates heat. The bigger data centers get, the more heat we are talking about. And data centers are getting quite large indeed, measured in millions of square feet of servers stacked like so much electronic cord wood.
Some data centers have started addressing heat as a resource, rather than a waste-product: IBM’s Swiss data center heats a pool; Telehouse in the UK is heating homes in London’s Docklands district; and Notre Dame’s Center for Reserch Computing is heating the flowers of a local municipal greenhouse with the heat from a rack of high-performance computing nodes.
Not everyplace where there are data centers needs heat, though. Some places simply need energy. As any engineer will tell you, where there is heat, there is potential energy. The key will be to capture enough heat so that it can be efficiently turned into energy, for example through steam turbines. Energy generated like this – through the waste heat of data centers, we will call “data-thermal energy.”
China is a natural place for the development of data-thermal energy. The country is early enough in the cycle of development for data centers to start designing its largest server farms to capture and channel heat efficiently. And scale will not be an issue in China. Leaving out government-run data centers entirely, some commercial data centers, like one 6.3 million square-foot beast under construction in Langfang just outside of Beijing, will have more floor space than the Pentagon.
The ability to capture and use waste heat efficiently also opens the prospect of cutting down on air-conditioning costs. If the heat can simply be blown – or sucked – away from the servers and into a central collection point for energy generation, the need to actually cool the air should abate a bit.
There is considerable engineering work to be done, but this is a worthy (if not essential) direction of thinking for the people designing and growing China’s server farms. It will demand imagination and discipline: the old way of doing things – stack ’em high, chill ’em down, and blow the hot air out the window – is cheap and pervasive. As the costs of energy grow and sustainability becomes more important, however, Big Data will need to start seeing itself as a utility, not just a customer.
Near People’s Square, Shanghai
Skyline in Silhouette
Walking the floor at both CES in Las Vegas and Electronica China in Shanghai within a ten-week space provides one with a clear view of how far Chinese enterprise has come, and, equally important, the degree to which international technology businesses have lost their former dominance in China.
One could conclude from these impressions that multinational tech companies are in a state of permanent decline in China: Beijing’s unstated but ongoing policy of import substitution has succeeded, and foreign companies are fighting a losing battle. You don’t need to go to trade shows for anecdotal evidence. Just look in purses and backpacks: ZTE, Huawei, TCL, Lenovo, and Yulong are five of the top ten mobile device brands, and they’re gaining on the global giants.
But if you dig a bit deeper, as you can at a show like Electronica, you find that the opportunities for foreign tech companies have not disappeared: they have evolved. To understand why and how, it is useful to start by looking back on how the tech business developed in China.
From Buy to Make
Since the beginning of reforming and opening in China in 1978, the nation has essentially gone through three phases of foreign involvement in technology-based industries.
The first phase was imports, when the government focused on bringing urgently-needed products like personal computers, telephone switches, automobiles, machine tools, and other technology-based products into China. The need for these products, most of which were essential to ease key bottlenecks in the development process, was so urgent that key ministries were permitted the use of precious foreign exchange to purchase those goods.
China’s leaders always expected, however, that the nation would begin producing these goods on its own, preferably in local companies, but realistically in joint ventures with global technology companies who would bring three essential ingredients: the products, with their component technologies; production know-how, with process technologies; and the capital to build the production facilities. This was the second phase: the shift to local production.
By the mid-1990s, though, another shift began to take place. As the global tech giants ramped up production in China to a mass-scale, local firms began manufacturing their own technology goods. Local firms began to dominate production, using a “fast-follower” approach: “maybe we won’t be innovators, or even the first to market with a given innovation, but we will come to market so soon after the innovation leader that we will still reap our share of the market.”
By last year, the payoff of this shift had become apparent. Chinese high-tech companies were long past needing foreign manufacturers to teach them how to build high-tech products, to help them implement cutting-edge production processes, or even to finance the construction of factories. Those local firms unable to bootstrap their own capabilities and finance now had a vast stable of local and foreign companies ready to provide the necessary technology, and finance, thanks to cash flow and capital markets, was no longer a problem.
Innovation, however, remained a challenge. While a handful of local tech companies – notably (but not limited to) Huawei, ZTE, Xiaomi, and Leovo – had begun to innovate, widespread innovation that would offer a more sustainable competitive advantage (and a larger share of profits) still seemed a ways off.
Enter the Innovation Platforms
And there it remains today.
This gap between efficient production and value-driven manufacturing is the heart of the next opportunity for foreign firms. While the days of foreign brands utterly dominating technology markets in China may be past, more than ever China’s manufacturers need a steady stream of innovations upon which they can base their own innovating.
Technologies that serve as the foundation that allows others to innovate are what we can call innovation platforms. Five factors make innovation platforms stand out from other technical advances:
Significant – The core innovation is a genuine advance that is both useful and relevant;
Substantial – There is a obvious, large, and diverse market for products based on the innovation that offer substantial profit potential, and the technology is easily commercialized;
Shared – The company promulgating the core advance is more interested in creating an ecosystem than a monopoly, i.e., it is content with focusing on supporting and enhancing the core technology and not getting into the business of its customers/licensees;
Stable – Any subsequent changes in the underlying technology are likely to be iterative, not major, for several generations of products. This makes it economically viable for companies to invest in R&D based on the innovation platform.
Supported – Rather than serving as a glorified patent troll, the companies that develop innovation platforms invest heavily in resources designed to assist product developers create viable commercial products, such as on-site engineering support, system validation labs, extensive documentation, or developer groups. In addition, the company continues to invest in improving the core technology.
Early Innovation Platforms
Many innovation platforms take the form of acknowledged industry standards. Examples like Wi-Fi, Bluetooth, and USB could be considered a form of innovation platforms, in that their technologies enabled the creation of products and even companies.
But when we talk of innovation platforms, we are really looking at products and technologies that spawn not only products, but companies and entire industries. Some illustrative examples:
The Xerographic Process: Invented by Chester Carlson and later commercialized by Haloid/Xerox, which begat the photocopier, the laser printer, desktop publishing, and many specialized sectors;
The Intel 8000 microprocessor family, that together enabled the creation of the personal computers, stand-alone video games, and a half-dozen major industries;
Qualcomm’s CDMA: CDMA enabled the commercialization of the internet, created the telematics industry, and is on its way to recreating the automotive, trucking, and healthcare industries, among others.
Each of these companies took an indirect lesson from the failure of Thomas Edison’s Motion Picture Patents Company, an industrial trust that tried to control the film business as well as the manufacture of cameras and film stock. It was, arguably, Edison’s greatest failure. By exercising a modicum of control over the core technology, supporting it, advancing it, and making it available on reasonable terms, Xerox, Intel, and Qualcomm each fostered the creation of immense economic value.
Platforms for the Future
In a world where industrial and engineering capability is a scarce quantity, the easiest way to make a return on a major innovation is to create a vertical industry around it, building the components, creating the product or system, and distributing it under your own brand. The Bell System did this for nearly a century with telephones, and IBM and a handful of other companies did this for the first three decades of the computer industry.
But when the ability to design, engineer, and industrialize complex products is widely distributed, as it is today, robust companies are built on either using innovation to enable industries, or in building on innovation to create industries.
For the time being, Chinese companies are (generally) comparatively better at building industries based on key innovations, and European and particularly US companies are (generally) comparatively better at consistently creating core innovations that can serve as the platforms for those industries. This does not mean that no core innovations will come out of China, or that the US is no longer capable of product development and commercialization.
But it does suggest that the richest opportunities in China for foreign companies, particularly those in science, engineering, and technology-based industries, lies in licensing and enabling Chinese manufacturers, rather than competing with them.
The question facing tech companies, then, is whether and how to make use of the company’s innovations – or an ongoing stream of them – in order to serve as a profitable and indispensable platform for Chinese innovation. And for those of us who watch this market, the pressing question is “in which industries will the next round of innovation platforms emerge?
I leave the first question to the companies themselves. For the second question, my early research points to transportation, healthcare and biosciences, construction, energy, and the environment. I know: I have my chips on a lot of spots on the roulette table. In the coming months, I look forward to sharing with you why I think things are going that way.
Blue Moon and Justified
When challenged to come up with examples of innovative Chinese companies – or those that might start innovating soon – many of us are hard-pressed to come up with names beyond the obvious Tencent, Huawei, and Lenovo. To help remedy this, and to make a balanced case for China as an innovator, I am going to start highlighting select Chinese companies that I believe are moving in that direction.
One company to keep on the radar is Hisilicon. Formerly Huawei’s application-specific chip (ASIC) division, Hisilicon has developed a system-on-a-chip (SOC) product line designed for mobile devices. The recent announcement that Huawei will be using Hisilicon chips in its upcoming flagship Ascend P7 mobile phone offers no surprise – on the surface. In fact, a skeptic might suggest that Hisense winning a spot on a a Huawei device is so much internal self-dealing.
The skeptic would be only half right. Huawei’s mobile device team are a loyal bunch, but the company’s leaders are no idiots. To risk the company’s tenuous reputation among consumers in an insanely competitive market merely to engage in some gratuitous dogfooding is uncharacteristic of the firm. Something else is going on, and it is likely that Hisilicon is sneaking up on the better-known MicroTek in its ability to provide the processing power for complex smartphones. If that is the case, Hisilicon is about to pop onto the radars of both Qualcomm and Intel as well.
Before we add Hisilicon to the ranks of mobile chip powerhouses, however, we need to add an important caveat. It makes good sense for Huawei to buy from Hisilicon if it can, but it probably does not make as much sense for other manufacturers. Putting a chip into a phone design involves more than just buying processors off the shelf and sticking them on a printed circuit board. Smartphone testing and development demands close cooperation between component providers, essentially letting everyone in the process into a lot of proprietary secrets.
If I were a smartphone manufacturer, I would look at Hisense SOCs in the same way that I would look at Samsung memory: whatever the virtues of the silicon, I am giving my competitor a close-up look at my mojo. In a world where Samsung and Huawei are pulling out all stops to lead the smartphone business, that’s writing an invitation to my own funeral.
So that is why I am watching Hisilicon. The technical capabilities are growing to the point where the company is likely to become a nexus of innovation, but the commercial challenges it faces are interesting indeed.
“Mobile Advertising Lags China’s Smartphone Explosion“
January 24, 2014
Reporting from Shanghai, AdAge’s Angela Doland writes a thought-provoking piece on how mobile e-commerce continues to outpace the growth of mobile advertising in the world’s largest smartphone market. As a percentage of all e-commerce, mobile is creeping into the double-digits, reaching as much as 21% during major holiday promotions.
At the same time, after years of effort, the most optimistic projections would have mobile advertising reach 3% of total ad spend in China this year. Given that Chinese users spend some 40% of their media consumption time staring at their mobile screens, you can understand the frustration of the advertisers.
Mobile Advertising Done Right
On the one hand, this trend should not surprise us. History teaches that effective advertising techniques for any new medium emerge only after an often extended period of trial and error. E-commerce initially grew much more quickly as a percentage of Internet-based revenues than advertising, and advertising was slow to find purchase in print news, radio, and television.
What this suggests is that the problem is not whether advertising can be adapted to mobile. The problem, rather, is that advertisers have yet to find an approach that makes the channel compelling.
Mobile Ad 1.0
There are three ways to approach mobile advertising. The first is to approach it as another channel for online advertising. This is where you talk about text-based advertising, display banners sized for the mobile screen, mobile search-based advertising, and ways to insert clever ads into music, videos, books and games consumed on a phone or tablet. Let’s call this “Mobile Advertising 1.0.”
My experience is that this has been the common approach in China, but that the challenges involved in making it work across three carriers, a half-dozen operating systems, hundreds of devices, and thousands of apps have made it difficult to get economies of scale. This alone might, in fact, explain why we are yet at such low numbers. Would it be easier with one carrier, one phone, and one operating system? Indeed. But I suspect that is not the real problem.
Perhaps, instead, we are misunderstanding the channel, and need to rethink how we do things. Back in 2006, I was in the room when my friend and former client Ian Chapman-Banks explained to a Japanese reporter that the reason that mobile advertising was having so much trouble was that we had failed to understand the value proposition.
Mobile Ad 2.0
Ian’s point (and I am paraphrasing heavily here) was that advetising as we know it was based on reaching out to chunks of people with similar characteristics at a given point in time. Mobile, Ian noted, had the ability to enable us to deliver a specific message to a specific person at a specific location and specific time.
In other words, what was keeping mobile advertising from being effective was that we were not using what made it fundamentally better than mass media advertising. This is the first time in history that advertisers could reach a person of their choosing at the time and place of their choosing, and all advertisers seemed to worry about was where to stick the banner on a small mobile screen.
Mobile advertising would be effective, Ian implied, when we figured out a way to make these capabilities work for the advertiser. Clearly, we are still looking for that combination, yet given the speed with which mobile is evolving and the innate conservatism of the advertising industry, this should come as no surprise. The key was to experiment and to keep experimenting.
The Mobile Ad 2.0 argument, then, is that if we want to figure out how to make mobile work for the nearly 1 billion mobile users in China (not to mention the rest of the world,) we have to experiment. Ian, who at the time had a generous marketing budget at his disposal, had allocated 10% of it to what he called “R&D:” money to try new channels of advertising and marketing that would not be evaluated alongside traditional channels, but that were just there to make sure that when something new worked, the company would be ready to exploit it.
So we aren’t at Mobile Ad 2.0 yet, but if we stick with it, we will get there eventually.
Is there a Mobile Ad 3.0?
Late last year I wrote a post that summarized why there are a number of ways to approach social media, each of which is guided by the marketing or technology silo from which one has emerged: practitioners who come out of advertising see social media as an advertising medium; people who come out of direct marketing see it as a direct marketing channel; PR people see it as a means of delivering messages; and so on.
What is different with mobile is that, in part because the challenge in putting mobile to work is, at the moment, much more technically intensive, the companies, departments, and agencies playing in that field have been those with lots of money. In short, it has been the advertising people. For that reason, we tend to talk about mobile as an advertising platform.
That exposes an assumption that is not necessarily supported by the facts. Zooming out of our ad-focused myopia one step further, then, we have to ask this: does mobile marketing need to be advertising-based, or are we missing something?
It’s Mobile Marketing, Jim, But Not As We Know It
In addition to allowing us to target an individual based on habits, time, and location, mobile also allows us to engage that individual in a conversation at a specific time and place. Mobile market research is based on that premise, and some of the early results hae been promising. As long as market researchers do not bombard us to the point of insensitivity with intrusive polls, and provided that we make it worth someone’s while to respond (good information is never free), this is likely to be a fruitful channel for some time to come.
Mobile has great value for point-of-sale applications based on near-field technology that go beyond completion of a sale. I walk into a hotel, and I am already getting notes on Foursquare about specials in the coffee shop. That’s a good start: it would be better if those specials were relevant to my dietary needs (e.g., “hi, David! We have great vegetarian options for you today!”)
Or how about direct-response on demand? When driving from city to city, I could tell Google’s Waze app on my phone that I needed a Sinopec station, and it would tell me distance, directions, prices, and offer me a coupon for stopping in.
I could go on, but you get the point. If there is a Mobile 3.0, and I think there should be, the opportunity is to start from the targeted user’s wants, needs, location, situation, and time, and work backward to the advertiser. This demands an intermediary who can make the match, of course. That’s why I think services like Criteo are going to translate well into the mobile space, and, in the long run, so will Baidu and possibly Tencent. The real gold rush will be for those companies who have the mass of advertisers on the one hand and the mass of users on the other.
Hence, Baidu’s ongoing interest in mobile. IF there is a single Chinese company that should make mobile advertising 2.0 or 3.0 happen, Baidu is it.
No PR Playground
What I am still trying to figure out, though, is where public relations has room to play in mobile. I have heard a few ideas, but I don’t see anything compelling so far. Classic advertising and classic PR don’t yet have roles to play in mobile to the degree that advertising does with online and PR does with social.
Yet every time I sit down and watch another compelling mobile technology demonstration, I am reminded that the tools we are creating today will be hopelessly antiquated, irrelevant, or both in five years. At some point, we are going to figure out how to make a connection between a company and a mobile user work out well for everyone. But we aren’t there yet.
…to Michael Galeotto, who was kind enough to include Silicon Hutong in his list of worthy Chinese bridge blogs.
As he notes, and as readers will acknowledge, I don’t post here daily. I figure that my readers, like me, have lots to do and read in their lives, and I try to post here only when I have something to say that I figure is worth reading.
…and to you. For every article I post here, I still spike (reject) two others. I will continue to do that, because I figure it is the least I can do to thank you for eleven years of paying attention to this forum.
A quick note – in order to make this blog available to people in China without a VPN, I will be shifting servers in early April. If you are following siliconhutong.com, you should experience no change. But if you are following me through WordPress.com, please note that you will probably need to change your settings when I make the switch to a wordpress.org-based site.
Update: Added a link to Michael’s post.