The cause of the Web 2.0 bubble

Web techies are lazy when it comes to considering the business aspects of their latest Web 2.0 ideas. They spend even less time developing creative revenue models when they see that web businesses with little real revenue can be valued at $15-billion.

When Adwords, Google’s paid search-engine advertising program, launched at the beginning of 2003, it wasn’t met with rapturous applause. Display advertising was the most prevalent form of online advertising since the late 1990s and those in the industry knew that it wasn’t very effective — browsers developed banner blindness and eventually didn’t even see the display banners at all. It was unusual at the time for a traditional marketing agency to put a significant portion of its clients’ marketing budgets into online advertising placements.

Google Adwords changed that.

Suddenly, you could track your marketing spend more effectively, because Google disclosed how many times your ad was displayed (impressions), which of your chosen keywords generated the impression, and you only paid every time that someone actually clicked on your advertisement.

Over the course of the next three years, companies poured more and more money into these paid search-engine placements, to great effect, because early advertisers focused on direct response metrics — they measured their Adwords spend and they compared it with the direct revenue generated from that advertising to calculate whether their efforts were effective.

Shortly after the launch of the Adwords program, Google launched the Adsense program. This allowed every web publisher out in cyberspace to monetise its content by displaying ads from the Adwords program alongside its content. The response from webmasters was overwhelmingly positive — finally they could make money out of their online content! Advertisers were prepared to experiment, because the same ads seemed to work so well alongside search-engine results. Unfortunately, the Adsense program met with limited initial success, because the clickthrough rate of text ads next to content is many times lower than the same text ads next to relevant search results, and so too is the conversion rate of that traffic.

But then in about 2005, brand marketers started talking about the branding effects of these search-engine and content-network text ads, and the fact that an advertiser would receive the benefit of unaided brand recall (the brand marketer’s nirvana) from these placements. Browsers viewed your ads, even when they didn’t click on them: this was like free advertising! Accordingly, these marketing gurus suggested that advertisers could pay a premium to their direct marketing cost of sales, because advertisers generated sales from their search-engine marketing campaigns that couldn’t be tracked directly (direct domain types, repeat searches and so forth).

By implication, while you were at it, you could get even better brand recall if you featured your brand prominently on a visually appealing, contextually targeted display banner and spread those banners as far and wide as possible across the web. At the same time, search-engine ad placements became increasingly competitive and expensive, so search-engine advertising budgets leaked into contextual banner-ad placements on the various content networks.

In other words, advertisers quickly unlearnt everything that they had known was wrong with banner display advertising.

Fast forward two years and we now battle to find a new Web 2.0 start-up that doesn’t have display advertising as its primary source of revenue, because apparently people on the internet want everything for nothing and Google has proven that you can make money out of display advertising. These start-ups are almost invariably under the illusion that in order to replicate the Google Adwords cash machine, one has to create a large online community and place contextually targeted, visually appealing ads in front of it. The social networks, most notably Facebook recently, have suffered under this illusion and had dismal results from running banner-ad placements alongside profile pages.

There are three main reasons why text-based ads are so effective on Google.com and other search properties, compared with the poor results of banner advertising next to site content:

the relevance of ads next to site content is not as high as ads next to a SERP;

the intent of a page browser is not the same as that of a searcher; and

text ads are less intrusive than banner ads.

So really, what Web 2.0 start-ups lack in their display advertising revenue plans is an appreciation of push versus pull marketing.

Searchers on search properties like Google.com are looking for advertising; they expect to find advertising and as long as it is relevant and not intrusive, they will respond to it. People who are browsing the internet are not looking for advertising; in fact, they teach themselves to ignore it (banner blindness), and any superficial attempt to place advertising in front of them is usually seen as being intrusive. It degrades their user experience and does not assist in creating brand equity and, consequently, favourable unaided brand recall.

Even if these browsers can be convinced to click on your push-advertising message, the chances that they are going to convert into customers at your site, at that time, is very low. Browsers will likely be at a very early stage in the buying cycle and as a merchant, it will require much effort for you to push them through a whole gamut of conversion funnel elements, of which some will fall outside of your control — for example, they may want to get the opinion of a friend on your product before they purchase.

In my opinion, the banner-advertising-next-to-page-content cycle persists as a result of two flawed assumptions:

the advertisers keep putting money into these ads, because they believe that although the clickthrough and direct conversion rates are low, they benefit from branding, which leads to sales that they cannot track back directly to this spend. This leads to rampant overspending on display-advertising placements, which content and applications owners believe they can exploit at higher CPMs the bigger their user base gets and the more it can be segmented; and

the content or application owners believe that they have no other way of monetising their application or content, because people want everything online for free.

Incorrectly valued display-advertising placements and pervasive free access to content and applications result in large online social networks developing around this free content or application that don’t generate enough display-advertising revenue over time to substantiate the network’s size or valuation. If, as a content or application owner, you try to break this cycle by charging for usage before you have achieved scale, then you are unlikely ever to achieve scale. Consequently, Web 2.0 venture capital keeps replicating weak revenue models across a limited number of opportunities, resulting in a valuation bubble.

If advertisers stopped overpaying for display-advertising inventory and content and application developers understood why display advertising is not effective as a primary revenue model for them, then the fallout from this bubble would be considerably reduced. Unfortunately, the Prisoner’s Dilemma from game theory tell us that game participants will not act in the interests of all, and casualties from Web 2.0 are inevitable.