Ever noticed that when you click a link on Google, then come back to try and click the next link, a box with similar search queries spawns in about a second later, right as you're about to click the next link, causing you go be sent off to an entirely different search query?

It's been around for quite a while (early 2018), and despite constant complaints from users, to the point there's even an extension that disables it, it doesn't seem to be on its way out any time soon, either. Why, when this "feature" is among the most hated parts of all of Google Search? Why, when Google themselves don't just advise against doing this, but actively lowers pages that do this in search results? Even if they're an Agile™ company, an error like this wouldn't fall through the cracks, let alone with such hypocricy, right?

And the more you look, the more strange events like these pop up—take for example the categories ("Web", "Images", "News", "Shopping"...); it is almost never to anyone's benefit to randomize the order of these should-be-stable elements, let alone stuff a bunch of other junk in there (indistinguishable from the thing you wanted):

What's going on here? These aren't reverted like the usual "oops"es, so why is Google actively trying to divert you from your task, from the answer you seek? And while we're at it... doesn't it feel like Google Search used to be...better? A lot of results nowadays are just (seemingly) autogenerated spam, made specifically to cram as many ads in and get the highest search ranking... Where's the content?

Where Growth Ends

To understand Google's decisionmaking, we need look no further than their revenue breakdown—the great great majority comes from ads. In fact, almost all ads you see in browsers are Google ads. Google has practically monopolized the advertising space on the Web with their intense monopoly: a monopoly on search, as well as a monopoly on browsers, one they are happy to keep pushing. In TETR.IO's case, a jaw-dropping 95% of players use Chrome, or a Chrome-like browser (e.g. Brave, Opera, Microsoft Edge, etc.). They all run modified versions of Chromium (the open-source base of Chrome, also developed in-house by Google), and thus inherit any changes by Google. For all intents and purposes, they are Chrome reskins. (Not that it matters, as Chrome reskins are still in the minority, making up only about 9% of players.)

So, in terms of occupying the advertising and browser markets, Google has no way to increase their revenues and keep investors happy. The growth of the advertising market on its own isn't something Google can truly influence (and increasing their cut would only harm this). So how else can Google grow?

It's a simple question, but since Google relies on ads, we need to take apart when exactly Google makes money. In specific, Google mainly makes money when it can either show ads directly on its own content (and get a 100% cut), or when it can send you to any page that serves ads through Google's network (and get a smaller cut). The more ads on the result page, the better. The problem is, people don't spend 100% of their waking days on Google properties, or even on the "general Internet" at all. Since this is the only place where Google can increase their revenues.... what can we do about this?

Optimizing "engagement"

Let's get to work then! Let's take an example user. They might be doing something productive (let's say, writing a document), and suddenly they need to look something up. Let's make a chart for that:

Here, the user searches something, looks through the search results page, clicks a result, then finds the answer they were looking for. In this time, Google got to serve 2~4 ads themselves (in the Search page), and the site owner of the result page got to serve another maybe 3~4 ads. That's good, and makes for a good user experience. The user finds what they were looking for in no time, and can go back to being productive.

...

This isn't very lucrative. What if I told you we could turn those simple "3 or so" digits into maybe, 10 times as much? Keep the user searching for 10+ minutes? No reason not to, there's no alternatives anyway! After all, the point of ads is to market in human attention span. Let's redo the chart to something better:

Look at how much more time is being spent on Google and sites with Google ads! Total searches are tripled, link clicks quintupled, and above all, our revenue from this interaction has increased manyfold! Now, the user searches for something, gets to a spammy AI-generated site filled with ads we get a cut from, tries a bunch more, then accidentally clicks to a new search term (allowing us to serve new full-price ads at the top), gets to some more spam, tries another query, and only after a long time of being "engaged" leaves Google!

This might sound dumb and exaggerated, but this is the reality that comes up when Agile and data-driven development becomes the norm. Data-driven development is just a fancy name for ignoring the big picture and hammering down on optimizing specific numbers (KPIs)... and those will always include revenue and meaningless stats like "engagement". It is not a coincidence that the hitbox for the related search terms exactly overlaps the hitbox of the next search result:

Side note on those spam sites: why would Google remove them? It's a question with a simple answer: they wouldn't, as they actively increase Google's revenue.

The Death of Web 2.0

If you think this might be an isolated case, look again. This is all around us on a massive scale. As soon as a company has grown to a size where it cannot gain new users, it instead looks to gain more "engagement" from said users. It's pretty logical when put like that—but when applied to search engines or other productivity tools, it goes against their purpose. This also explains another aspect of why search results are so low-quality: many modern sites don't want to be indexed anymore. How often have you clicked a search result to end on a loginwall? Sites want to protect their own growth by locking you into their platforms, too. They don't want you jumping in, getting the info you need, then leaving.

In essence, we're witnessing the true death of what we used to call Web 2.0 - the web as a publicly available writeable interconnected network of information, where anyone can get the information they need within a few clicks. Instead, the Web has become a set of massive islands, purposefully isolating themselves from all others, and making sure you stay there for as long as possible.

After all, a product is a means to make money. There's no reason to care about the overarching quality of your product when it doesn't increase your KPIs in the short term!

Have you run into similar things like this? I'm wondering about other people's interactions with KPI-focused UI design like this, and foregoing all user experience or even the core of a product for bottom line gains. Let me know your experiences in the comments!