The Search Engine: Chronicle of a Death Foretold

Gartner has recently predicted that the use of internet search engines will fall by around 25% by 2026 as they are replaced by chatbots and other virtual agents powered by generative algorithms, a trend that will change online marketing as we know it.

This is hardly a surprising conclusion: since the explosive growth in the use generative algorithms following the release of ChatGPT at the end of November 2022, we have seen how our information needs can be better met by a well-formulated prompt rather than foraging through traditional search engines.

The trend, which I have been able to corroborate by testing my students on a classroom laptop, makes a lot of sense: while a search generally directs us to a set of links or, in some cases, to some questions that try to provide more or less direct answers, the result of a question to a generative algorithm is, overnight, much simpler to manage. There is, of course, a certain element of novelty appeal in this, while at the same time, a complex relationship of trust has been established that to some extent has replaced critical thinking, as had previously happened with search engines themselves. At the beginning of this century, the first result of a Google search seemed to many people to be some kind of absolute truth; now, many unquestioningly accept a well-written ChatGPT answer, despite the tendency of algorithms to occasionally “hallucinate”, to simply make things up.

We shouldn’t forget that Google, despite having the technology to create generative algorithms, refused to make them available years ago, a perfect corollary to Clayton Christensen’s absolutely brilliant 1997 book “The Innovator’s Dilemma”, because it feared effects on the search business that were easy to anticipate:

1. Advertising through sponsored links, Google’s main source of business, stops working when the user, instead of receiving a series of links, some of them sponsored, instead finds a perfectly written paragraph.
2. The costs of a search in Google’s massive database are contained thanks to Google’s hyper-optimized engineering, while those of a query to a generative algorithm are more than an order of magnitude higher. This difference in costs can be assumed by companies with a relatively low volume of search activity, but they could completely destabilize Google’s bottom line, with a more than 95% market share, if the substitution process took place too quickly.
3. Fear of failures or “hallucinations”, that would be largely overlooked in the case of Google’s competitors because of their low market share, lack of experience or because they were seen as more novel, would have been unacceptable from a company with Google’s market share and prestige. And if they happened often, they could even compromise the perception of reliability of the search engine’s results.

Therefore, Google decided to apply the well known principle: if you don’t know what to do, do nothing at all, and this is why Google delayed the launch of its own generative algorithm, and why it was completely taken by surprise when OpenAI put ChatGPT on the market in November 2022. Christensen would happily have risen from the grave to give himself the pleasure of explaining this in a class, since it constitutes the perfect example of what he described. If you don’t act when you have the opportunity to do so, some competitor will come and steal your thunder.

For traditional search engines, which first appeared in 1993 with players such as W3Catalog, ALIWEB, JumpStation or WWWWorm, all long gone and substituted by Google, the writing was on the wall when generative algorithms with simple interfaces became available to the general public.

I’ve never been wholly convinced by Gartner’s predictions, but according to the trend I’m witnessing lately in the limited sample constituted by my MBA students, that 25% drop in a couple of years seems reasonably believable, or even cautious. It means that we will begin to see a whole new way of doing things on the web, transitioning from search engines with business models that have been grubbied by SEO, to finding new techniques designed to get generative algorithms to mention some brands or companies more than others. As always, the problem is not the technology, but human nature…

(En español, aquí)

—

This post was previously published on MEDIUM.COM.

***

You may also like these posts on The Good Men Project:

Escape the Act Like a Man Box

What We Talk About When We Talk About Men

Why I Don’t Want to Talk About Race

The First Myth of the Patriarchy: The Acorn on the Pillow

—

Photo credit: iStock.com