The China Vortex

Search engine marketing is the main engine behind Google’s rise as a major online media player, and the product it is offered in is Google Adwords, which allows advertisers to directly target their online ads by selecting keywords, and then targeting them to relevant search results pages and to published pages (using Google’s publisher’s network, Adsense).

In China, the leading search engine company is Baidu, which started in the US, but came to China, and is now the most popular search engine among Chinese Internet users. It has been financially successful, and is listed on the US’s Nasdaq under the symbol BIDU

There are several reasons for Google Adword’s success, and the most important are two: PageRank, which measures the popularity of a web page by measuring inbound links which for the most part, are selected by humans and not computer algorithms, and introducing relevance into the keywords auction model. Under the Google model, paying the highest price for a keyword is not enough to insure clickthrus (for the most part, Google charges advertisers per click, or pay-per-click PPC), but it must be relevant. The more relevant it is, the more clickthrus it will get, and the less an advertiser will have to pay for a higher ranking.

As this excellent article makes clear, Google did not invent the keyword auction model, but it did perfect it. By perfecting the Google Adwords model, Google has become the hugely profitable online media machine it is.

As China becomes more important as a market, more advertisers are looking to sell directly into the Chinese market using Google and Baidu, the two leading search engine firms in the Chinese market. Baidu operates under a very different business model from Google, one which it has adapted to suit the Chinese market.

My understanding is that Baidu does not figure relevancy into its advertising fee structure, Chinese advertisers only pay for higher ranking. As far as I know, Baidu does not have anything like PageRank inbound linking algorithm to count inbound links either. Without these two elements, Baidu’s ad search looks a lot like the GoTo.com ad model. This makes it fundamentally different from the Google Adwords model.

I’m digging deeper into the search engine marketing business in China, and want to hear what you would like to know about. If you have questions, please post them in English or Chinese in the comments below.

Technorati Tags: adsense, adwords, auction, baidu, China, Chinese, engine, google, keywords, marketing, optimization, pagerank, ppc, relevance, relevancy, search, sem, seo

A while ago, a friend of mine who works in a 4A ad agency in Beijing said to me: “You think we’re all dinosaurs, don’t you, and that Google is going to own the whole advertising scene?”

I’m glad he challenged me, because he forced me to think things through on a deeper level as to what my prognosis on the ad industry is going to look like. I’ll try to outline those thoughts here.

First of all, advertising has been divided into roughly two camps; the “Google is going to own the industry” camp, and a more traditional advertising camp, which says that traditional advertising (TV, radio, print) are going to survive and prosper, albeit in a very different form. I must confess that for a long time I have leaned in favor of the former, or Google camp.

But having thought things through on a deeper level, I think that it might not be so simple. Here’s why.

Ad agencies were born in the 19th century; they did a simple job, they bought media inventory at wholesale rates from newspapers, and then sold them at retail prices to advertisers. Later on, as they developed, they also sold creative services to their clients as an extra value-added, generating more income. This is how the “mass market” arose; manufacturing had risen, and products needed to be sold to the masses.

The business model is simple: buy media inventory at wholesale rates, integrate them into media plans to sell to advertisers at retail rates, and charge them extra for creative services. Since media is controlled by a publishing group, or television or radio group, and is published at regular intervals, the media business is a lot like the airlines business. Airlines fly according to set schedules, and the more butts in the seats, the more profitable that flight is going to be. Media is a lot like that; media inventory are the seats, and advertisers are the butts (no pun intended).

The advertising industry is based on a version of information arbitrage; it knows better than publishers who the advertisers are and what they are willing to pay, and it knows better than advertisers the best ways to reach their audiences. Over time, a whole business ecosystem has formed, including media research and marketing, with their own businesses and terminology. This business ecosystem is deeply entrenched in larger company’s marketing departments, providing them with market information.

Where online media is different is that Google’s Adwords/Adsense model disrupted the whole model. When it comes to search advertising (ads placed on search results pages), Google owns the page (media) where ads are placed. But when it comes to overlaying ads on third-party websites, Google (and the other ad publishing networks) do not own the page. Instead, they pay a portion of the money generated by each ad click to the content publisher. So who determines the targeting of the ad campaign? What is the definition of media inventory? Sometimes it is a large 4A ad agency working for its advertising client, but it could just as easily be a man sitting in his living room in Shanghai targeting his customers with a Google Adwords campaign.

Google has empowered advertisers with the tools to reach their advertisers directly online, without any need for a third-party advertising agency, if that is indeed what the advertiser chooses to do. Furthermore, he/she can set the budget and tweak the campaign according to his needs.

Even more interesting: Google does not have to buy media inventory. It just charges a commission per click.

This has worked very well in an online world, where Google does not have to own the media inventory it runs its ads on. But does it work in traditional media (TV, radio and print) where reach and frequency are the main metrics? Not as well. This is because these are all backed by large media organizations owned by corporations, or in China’s case, by the government through large state-owned media organizations.

No pay-per-click model would work for them. And I’m sure that they would not go for a model where a third-party paid them a commission determined according to how many sales were made. Their problem: their audiences are eroding as more means of delivery become available (webcasting for television, podcasting for radio and more print content moving to the web). It will be interesting to find out if electronic paper will support overlaid ads. This will be a major factor in determining whether the future of print will be more like the Internet or paper.

A common criticism of advertising is that it is unaccountable; there is no way to know which ads directly lead to sales. The only media where it is possible to establish a direct connection between advertising and sales are direct mail advertising and its online younger brother, email advertising (spam). Now Google is trying to establish a direct relationship with scripts which monitor user behavior directly through the checkout process online.

A common marketing rule of thumb is that a person has to hear about a product seven times before he/she makes a purchase. What’s changing now is where they are hearing it from; now the referrals come more from Google Reader and Facebook than television. Social networks are taking more and more of my time. There is no way to establish to establish a direct connection all the time; any advertiser who claimed that they could determine which impression resulted in a sale would be a liar. The promise of Google and online advertising is that eventually it will become possible to trace a direct cause and effect relationship between ads and sales; this forms a threat to the business ecosystem of online advertising.

So far, advertising can measure reach and frequency, but it cannot measure purchasing intent. To reach a higher level of personalization, advertisers (especially of big-ticket items) need to know where each potential customer is in the sales cycle.

More and more, for niche products/services especially advertising is about balancing quantity (reach and frequency) with quality (clicks and other metrics which lead to sales).

If Google goes into traditional advertising, it will have to buy media inventory to sell to its clients for offline campaigns. If it does this, its cost of doing business will go up considerably, cutting into its earnings.

This is why I see Google’s possible foray into the traditional ad agency business as a defensive move, not an offensive move. Online advertising revenues are reaching saturation, and Google has to show its investors that it’s doing something. The real challenge to Google’s ad revenue model will come from Facebook, which has probably already eroded search engine advertising revenue. This is why the real battle will be between Google and Facebook for ad revenue.The age of the “mass market” is over; now we are well on the way to making sales the old-fashioned way:one-by-one.

Both traditional advertising and online advertising are playing out a classic game of “crossing the chasm”. For traditional advertisers, the challenge is how to build a new business ecosystem for online advertising; for online advertisers, led by Google and Facebook, it is how to learn the terminology and and be accepted by the traditional advertising business ecosystem.

Then of course, media publishers might decide to sell their advertising space in online exchanges or auctions, instead of just selling wholesale. CCTV (China Central TV) already does this; it will be interesting to see if others follow suit. If they think that they can get more revenue by adopting a more efficient buy/sell mechanism, then they will do it.

This raises the question of why Google and the 4A agencies are trying to build their own exchange networks? Instead, why don’t they build a new business ecosystem which media inventory owners and ad buyers could simply plug into using standard APIs? Wouldn’t that make things so much more efficient? Why doesn’t WPP or Publicis buy Facebook? Buy the network. Then why not charge both buyers and seller a service fee plus commission for the wider exposure and ad targeting the network brings? It could be argued that this is already what Google Adwords/Adsense already provide, but it does not yet provide a wholesale backend solution for plugging in large amounts of inventory.

For all its strengths, Google is not a user-generated network; Facebook is. It is already selling advertising space, and it is in a better position than Google to create a whole new online metrics for advertising because it owns the network.

At the end of the day, advertising is largely a numbers game, except for the creative part. And then the creative is just to support reaching the right target audience and increasingly now, getting them to convert to buyers.While Facebook now has only 50M profiles, it is growing exponentially.

So, to my friend, I would say that we’re all dinosaurs basking in the glory of our late Cretaceous period just before the asteroid strike, and the world is changing faster and faster. The main question is how human and computer work will be allocated; this is a question which meets in a very intriguing way in the advertising industry.

If this period in advertising was Chinese history, we would call this the Warring States period. The question is whether the ad agency will still be recognizable in 10-20 years’ time? I would say that it is; as long as there are people who spend a portion of their lives reading and absorbing content both online and offline, and if agencies remodel themselves into generalist marketing organizations equally comfortable with both online and offline, there is room for agencies, both large and small.

And what will ads look like? I don’t know, but if more look like Fight for Kisses, I’ll be pleased.

I found out about this very entertaining video commercial from a blog I read on Google Reader. Enjoy.

Technorati Tags: 4A, ad, adsense, advertising, adwords, Beijing, business, cctv, central, China, Chinese, crossingthechasm, ecosystem, exchanges, google, offline, online, television

Alibaba has chosen Google China as its main advertising platform partner for its online advertising service Alimama over Baidu.
Alimama provides roughly the same advertising campaign targeting and service delivery capabilities to advertisers as Google’s Adword service worldwide, with the biggest difference being that Alimama is targeted at the Chinese domestic audience.

Alibaba had been in secret discussions with both Google China and Baidu. The discussions with Baidu broke down for undisclosed reasons, and soon after, Alibaba announced its partnership with Google. This agreement is important because Alibaba is the owner of the largest B2B platform, Alibaba.com, and also China’s leading online auction firm, Taobao. Taobao has successfully defended its online auction presence in China, forcing eBay China to hand over its operations to Tom Online while it rethinks its China strategy.

This is a major blow for Baidu since Alibaba has the capability to spend a significant amount of revenue targeting search users and publishing networks with ads. In the US, eBay is one of Google’s biggest Adword’s clients, but the relationship has recently become rocky because the two companie’s have competing online payment systems. While online payment systems are not the most sexy online products, they are highly profitable since they usually operate on a commission system, taking a cut of the total transaction, instead of a flat fee.

Google has introduced Google Checkout in China, and Alibaba has its own payment system, Alipay. It is likely that in the advertising agreement both payment options will be offered to campaign buyers. For observers, it will be interesting to note whether Google Checkout or Alipay will achieve “preferred service provider” in future revs of the service. This will obviously be a source of major competition between Google China and Alibaba even though they are cooperating on this advertising solution.

Baidu has 62% search marketshare in China and is the market leader, while Google has only 20%. Baidu, even though it is widely seen as China’s native son in the search market, has significant problems which I have discussed at some length in an earlier article on the Chinese advertising market.

Baidu’s single greatest challenge is coming clean about click fraud. A major reason for its inability to tackle the problem is that as a public company, any attempt to clean up the problem would hit its earnings, and may even lead to litigation about past performance. It would naturally avoid coming clean about the issue and push it off to future management to tackle. The trouble with this approach is that click fraud becomes a slow rot, and advertising clients will choose to shift their adspend to competing search engines which have more effective anti-click fraud mechanisms in place.

Click fraud has become a major drag on the development of the Chinese online advertising market, which is poised to pass 90 billion yuan this year.

This may well be the background to Alibaba’s decision to partner with Google China. Alibaba is planning for an IPO listing later this year.

Technorati Tags: advertising, adwords, alibaba, alimama, alipay, baidu, checkout, China, google, online

Microsoft’s recent purchase of online ad market platform AdECN Exchange highlights the rise of neutral ad market platforms as a new venue for the buying and selling of ads between content publishers and advertisers.

Online ad market platforms represent the next stage, or second generation, of ad networks. The first generation was represented by Google Adsense, the company’s successful platform for publishers, which provided ad inventory from independently-published websites for Google Adwords, the ad targeting and delivery system, and Google’s cash cow.

First generation ad platforms such as the Adwords/Adsense platform have used real-time indexers, or spiders, to scan content for keywords, and then match up advertisers with inventory. Technologically, this is great non-trivial technology, but there are also problems with it.

• If you are an ad publisher or advertiser, you need to join a network (Google, Yahoo!, MSN, Baidu, etc.). By joining a network, you automatically lose advertising and revenue opportunities with other prospects who may not be members of the same network.
• Advertisers and publishers are entrusting a third-party to act as their facilitator and to act in their best interests. While search engines make claims to be objective and neutral; this is in fact impossible. Just do a search on a term of your choosing across several different search engines, and compare the search results.
• As search engine companies go public and come under pressure from Wall Street and investors, management’s strategy is always to blur the line between organic (free search) and pay-per-click (PPC) search. As revenue becomes more important, search results become more skewed to favor sites which belong to their publishers’ network.
• Click fraud is a major problem which the search engines have never been able to come clean about. Aside from waffle statements to the effect that “click fraud is a minor problem which does not affect most users”, all search engines, even Google, have been reluctant to provide independent third-party statistics about click fraud. This reluctance to come clean has led many to believe that the problem is greater and would affect their revenue more than they want investors to know. In a worst case scenario, it could be manipulated into a Ponzi scheme.
• In certain markets such as China, where keywords are sold through distributors, there is even wider room for abuse through distributor collusion. This is why advertiser groups have formed organizations such as Fanbaidu who have challenged charges for advertising clicks made to their accounts.
• As the Cluetrain Manifesto made clear, along with Seth Godin, marketing and blogging are becoming increasingly about conversations. Blogs are nothing more than linked conversations on a given topic, and sometimes they ramble on by themselves. For this reason, blog content resists a “one size fits all” approach, hence the attractiveness of the long tail approach. For unique content, neutral ad platforms where buying and selling is done by human buyers and sellers online work better than networks which have their algorithms continuously tweaked. Since the most knowledgeable seller is the creator of the content, this means that more and more, content creators will become marketers and publishers of their own content. After all, the main task of a publisher is to attract good content creators and market their work.

This is why the ad exchange system is the trend of the future; it works best for unique content and for the long tail. Compared to ad networks, they are more transparent. Click fraud collusion is made much more difficult because the market is real-time and more dynamic, and the content creators and publishers would have it in their own best interests to fight and resist click fraud. Transparency rewards the honest over the long term. Exchanges are not perfect and Ponzi schemes can also develop in exchanges, but this has more to do with human nature than exchanges.

The problem with the advertising industry, as it exists today, is it is built for a world where advertisers and inventory are comparatively static, and where audiences are defined as being “mass market”. In today’s online market, where peoples’ needs, care and interests are constantly changing on a real-time basis, the question should become “Is there a mass market anymore, and what is its definition in quantitative and qualitative terms?” If the answer is no, then the main currency of advertising becomes attention, which would then have to be translated into monetary terms not only on an individual, but on a time basis. Pushed to its logical outcome, advertisers would need to pay consumers for their time and attention.

In an article on Ogilvy China Digital Watch, Kaiser Kuo raised the question about why ad exchanges were slow to take off in China. Although there may be many reasons, I believe the most important single reason is that content creators want to just create content, and don’t like the idea of marketing, buying and selling their own content or becoming publishers. They want to write for someone else and be paid, and don’t want to take the risk themselves. This problem is not unique to China; it will affect takeup of the online ad exchange model all over the world.

Of course, the market always tend to reward the individuals who see and act on opportunities before others.

Technorati Tags: AdECN, adsense, advertising, adwords, baidu, google, marketing, microsoft, msn, ogilvy, yahoo