Amerika

Why The Dot-Com Crash Is Still Going To Happen

For those of us who witnessed the first dot-com crash, it became clear that another would follow because the model had not changed at all. That one seemed to end in 2007 when mobile computing became huge, but even so, the model never changed.

We knew that the internet would fit into something like one of the models we already had: telephone, newspaper, television, cable, or radio. As it turned out, the internet went after the newspaper model, where advertising pays for almost everything, but mixed in video to try to make it thrive.

Although internet ad revenues keep rising, the effectiveness of internet advertising has always been in doubt. Advertisers pitch to where the users are, but it takes them years to find out that what they are doing has not been successful, and this is steadily the case with the internet.

Eventually, the false advertising money will run out, and we are going to have to confront the fact that Facebook, Netflix, Instagram, Reddit, and Twitter, et al., are worthless in practical terms.

Dot-Com 3.0 depends on Web 2.0 “push” interfaces to keep stuff scrolling across your screen while you click, but it turns out that very few of the ads actually sell anything. Rather, like Antifa and mousy cat ladies with hand-lettered signs, they “raise awareness.”

As it turns out, internet ads are probably worthless:

Last year, a group of economists working with eBay’s internal research lab issued a massive experimental study with a simple, startling conclusion: For a large, well-known brand, search ads are probably worthless. This month, their findings were re-released as a working paper by the National Bureau of Economic Research and greeted with a round of coverage asking whether Internet advertising of any kind works at all.

“We know a lot less than the advertising industry would like us to think we know,” Steven Tadelis, one of the eBay study’s co-authors, told me.

The problem, according to Tadelis and others, is that much of the data websites generate is more or less useless. Some of the problems are practically as old as marketing itself. For instance, companies like to run large ad campaigns during major shopping seasons, like Christmas. But if sales double come December, it’s hard to say whether the ad or the holiday was responsible. Companies also understandably prefer to target audiences they think will like what they’re selling. But that always leads to the nagging question of whether the customer would have gone and purchased the product regardless.

The biggest pitfall however comes from the tendency of online services to select their own audience. Facebook, for example, appeals to people with a lot of time to scroll and click, so tends to favor the young, the retired, stay-at-home moms, and people with easy jobs that involve a lot of waiting or doing low-impact activities with frequent breaks.

The “Natural Born Clickers” study from 2009 pointed to a tiny audience, about 8% of the population, being responsible for 85% of the clicks, and illustrated how most of these were people with below $40k/year in household income. You can advertise mugs and tshirts to this group, but they are not in the market for cars, homes, or competitive consumer goods.

We are heading toward a future where many users employ adblockers and almost all of them operate within niches. The idea of one big site for everyone — Facebook, Twitter, Reddit, Instagram, Pinterest — has effectively died under an onslaught of censorship and partisanship. People want to be with those who are already like them in order to avoid the toxicity and repression that comes with the general public typing away on their phones.

Still, some see the Dot-Com bust as never happening:

With the rest of the economy limping along, the Bay Area felt like Gomorrah before the fall. But the party just kept going.

It kept going in part because vastly more money was sloshing around than people initially realized. Source one for that cash: foreign investors, particularly ones based in China, Saudi Arabia, and Japan. Firms such as SoftBank have funneled astonishing sums into the pockets of start-up founders and early investors, helping prop up tech valuations and allowing early-stage investors to cash out. (It seems worth noting that tensions between Beijing and Washington have led to a collapse in Chinese investment in tech in recent months.)

Source two: the giants that form the highly profitable headwater of the broader tech ecosystem, such as Amazon, Google, and Facebook. These companies have flushed hundreds of billions down to smaller, younger firms and their investors in the form of acquisitions and acquihires. Facebook spent $20 billion on WhatsApp and Instagram; Microsoft spent $26 billion on LinkedIn and $7.5 billion on GitHub. Deals such as these were possible only because those acquiring companies were themselves so very profitable. More broadly, the tech dura-bubble has coincided with a continued boom in corporate profits that has lifted valuations and increased the sums spent on mergers and acquisitions throughout the whole economy.

In my view, the above illustrates how likely it is to happen. The foreign investment is drying up, and the big companies despite having a lot of money have fundamental problems with their own business models. This ceasefire will not last.

Further, people should get very wary when they hear about “mergers and acquisitions throughout the whole economy.” Mergers and acquisitions generally occur when a market is waning, which means that margins narrow and there is no longer room for competition because nothing can be won on thin margins, as occurs when technology gets established.

If Silicon Valley could identify the name of its crisis, it would be that its products when new commanded high margins, but now are as commonplace as running water, telephone lines, and credit cards. People want their gadgets to just work and do so for less cost than when they were cutting edge objects.

When Facebook gets recognized as daytime television and internet advertising is seen as ineffectual as those business card ads found on the higher-numbered pages of newspapers, this market will collapse just as surely as media seems to be collapsing.

On the broader front, we should wonder if consumerism itself has not collapsed. Starting in the 1940s, people had more money to splash around, and so everyone upgraded to the new labor-saving devices, televisions, and gadgets. The internet — smartphones and computers — was the last gasp of this.

Now that none of these devices are new, people do not want to spend money on them in any substantial degree. They want to have these things available like other matured technologies. With no new advances, the technology becomes a fixture, not a fascination.

That in turn means that the Big Tech companies are overvalued. The market still treats them like cutting-edge innovators, when in reality they are more like maintainers of past technology. At some point, the markets will recapture that lost value, and send the economy wobbling as a result.

Tags: , ,

|
Share on FacebookShare on RedditTweet about this on TwitterShare on LinkedIn