FIVE things you don’t want to hear about search engines

by | Apr 17, 2019 | Tech

Reading time: 8 minutes

Why Chrome silently adding DuckDuckGo to the search engine’s alternatives to Google is a big deal and what it tells us about Search Engines.

It should come as no surprise that, whatever it is that we search for, be it NBA news, sex scandals, celebrities or political opinions, we tend to come across our preferred team, the scandal of the celebrity we follow and the news outlet we have always relied upon. Search engines do this, normally, in good faith. They want to serve you the results you are most likely to go for because, if they don’t, you will use another search engine.

Hence, it’s not that easy to come across diverging opinions. All these search engines are going to analyze your activity, categorize you, match you with thousands of other users, and serve you the results all of you, as a tribe, are searching for. The result is that your internet activity tends to be reductive and heavily influenced by confirmation bias. Whenever you hear about the “bubble”, “alternative facts” or “post-truth”, this is what they are referring to. But then, why is it that so many people use Google almost exclusively as the go-to search engine?

That’s one of the problems of the indexed web: it’s not a democratic system, it is a pure oligarchy. Search engines were not invented with a philanthropic nor a pure analytic-driven interest in mind. They’re a company. One with a product to sell, and valuable one at that: YOUR ATTENTION. And as such, their primary directive is to get you coming back and viewing whatever it is they want you to see.

That is why I was tickled the other day, reading that DuckDuckGo recently made it to the Search Engine giant browser, i.e. Chrome. If you never heard of it, DDG is an alternative to Google that, unlike them, does not analyze your past history, nor tracks you even on your way to work. It just serves you the results of your search based on the words you are using and the popularity of the sites with those keywords.

But why would Google, the largest internet company in the universe, with the de-facto standard in search, give up control of what made them famous? And furthermore, give it to another search engine that does not track users? What is it, they don’t want any more money?

Now, do you remember The Matrix? In that universe, humans had become a source of energy at the service of machines. Well, in this one, we have become a source of data at the service of machines. DuckDukcGo, say no. They refuse to turn you into the AI’s and Deep Learning Algorithm’s meal. Starting in 2008, they now handle one billion searces per month. Google and bing are no longer your only options.

The baseline is: happy news for the users! Chrome now allows you to choose your favorite search engine (*clear throat*). No, Google Search is still the default one and you should still switch to Firefox if don’t want to be tracked constantly, but it’s a good start.

The fact that now we have a choice is a major event and it gives us a very important clue: Google is diversifying the ad-based revenue model, and has are several reasons for it. On its side, Android will switch from free to paid license-based service which will generate a lot of income from device producer such as Samsung. On the consumer side, we notice a very loud trend: ads blocker, very useful tools that enable the user to physically block the view dynamic advertising online, averaging 16% of browser’s installation, bots, click-farms and automation having a negative impact on real user’s interaction on media and prediction models. All these suggest an upcoming ad’s Armageddon.

Source: Statista

Why is this important? Mainly because ads are crucial for the sustainability of those services. There are several case where it is quite useful for the user to receive a suggestion or ease the information. In all the other cases, advertising reach us way to quick to understand what the user is truly looking for. Also add that some brand use it as an annoying interruption and distraction especially during a simple task such as a search query. So, are search engines still on our side?

1. Search engines aren’t free.

Google, Bing, and even DuckDuckGo are all Ad-based services. All search engines use the user profile to track the search history on the web as a target pool for potential advertisers. With the introduction of AI algorithm software, search engine services are trying to push it further by predicting what users might, or often ‘will’, need via related keywords based on history. It’s quite easy for instance to match a user looking up flu symptoms with a link to paracetamol products. Or a user trying to find an illegal copy of ‘Mein kampf’ by Adolf Hitler may suddenly find his Google searches inundated with Trump election ads.

In the case of a ‘no-ads’ option provided by DuckDuckGo, we can assume that a higher respect for every users’ privacy will be a strong selling point, pun very much intended. But what about the business? Well, here’s the catch. DuckDuckGo also collect other search results, including Yahoo, Bing and Wikipedia, to list the most relevant results. It’s still not saving the IP or any other sensitive data to link the user profile. This is definitely good news for users and it does respect the privacy, especially when ads are out of the picture. All this has a major impact on the user experience though, which have its own pro and cons, i.e. not being geo-localized.

2. Your user profile does not define you.

We are used to fast and easy answers. And most of the time we are not looking for the truth, but something that reinforces our convictions instead. In the end, it is the less consuming way to get our queries answered. Humans developed a peculiar trend in getting things done as fast as possible and we became quite sensitive on response reaction’s time. That’s why we have the kitchen so close to the living room, if not in the same room. But this behavior comes at a very high cost: we confines ourselves in our very own walled garden.

In this perfect mechanism were we find that exact news, that expected picture and that immediately familiar Facebook group, we are lost in our own cognitive bias. It basically makes us stupid and self-referred. And that is a problem that can imply something futile such as not being able to try a new restaurant. Or find a new brand that we would love to wear. Or use another search engine to open our mind to a different idea and potentially save us from Brexit.

3. They share and sell your search history.

We could argue that selling any information that we generate ourselves entitles us to get a cut, right? Wrong. Firstly, there are pages and pages of users’ agreements (the one you half-skim through in four seconds before clicking ‘Yes, I’ve Read and Agree), and secondly, we are using what essentially amounts to an exchange currency. The most disturbing part is the sharing activity that search engines do. So, even if your sensitive data is kept safe (*clear throat*, sorry this cough seems persistent) your user profile is constantly populated by your interests and used keywords.

I honestly am not concerned about my sensitive data. I don’t use email for new products, and spam filters works extremely well nowadays. Almost every phone has a blacklist to stop sales executives or call-bots annoying you during the day. I’ve also moved house every third year in the last 12, so good luck finding me. But this access to data becomes risky when we want to keep some information private. Let’s say I am 14 years old and I accidentally became pregnant (*clear throat louder*). I wouldn’t want my father to know it, presumably. Well, since 2012, Target have had different ideas. Now look at the desk where you’ve put your Amazon Echo and ask “Alexa, are you constantly listening to me?”. If said device starts laughing, don’t be surprised, as you’ve probably given the 7,000 people listening to your conversation a giggle.

4. Search engines are not the best place for Advertising.

Entering the click-farms. Ever heard of them? They are commonly found somewhere in south-east Asia, full of shelves with connected phones that, controlled by a central unit, open and close Instagram accounts, web pages, Facebook groups, to generate an enormous amount of clicks. Normally I would say there’s nothing wrong with that. The more click or like a page receive, the most visible it will become. But let’s consider the scope of the advertising.

Advertisers pay for traffic, and they expect a conversion rate of users/clicks for their money. They value the clicks based on this conversion rate. Let’s be clear: clicks generated by bot are worthless, mainly because they will never lead to a purchase or work for any branding purpose. The non-human interaction makes it very difficult for the bot on the other side to analyze the data and create realistic user profiles. The whole data-set is fed with tons of junk-data, diminishing the value of an otherwise powerful source.

We are labeled through our searches and clicks. It’s a painless way to share our preferences and needs to the market. Our actions are the product that ads service will monetize. As such, it should aim to create value.
When millions are invested on polluted traffic shortcuts to reach only fake users the total value generated of ‘null’ seems a bit short on potential. Profits skyrockets, yes… in return, it totally screws the data-based profiles that advertisers were looking for.

5. Analytics based on search engines are not in any way reliable.

Did you ever search for a project management solution? Please do this exercise with me. You will be overwhelmed by ads for project management software on any platform. I can’t unsee it. And I chose a tool months ago. Ads are late, unreliable and irresponsible. And starting to be annoying.
I noticed that even if I react to the ads, they won’t stop. That tells me there is no way to count my click on the ads, which is probably sunk under the sea of bots. That, or the Wednesday’s marketing meeting titled “Let’s seriously bother our potential customers”.

I wanted to get some data on that. Based on a 100,000+ clicks campaign run on an external, publicly accessible, advertising page (hard-coded), we experienced a discrepancy of almost 10% between the click counts and the users (actual real person) reaching the dedicated landing page (which was this same Web-zine). This is influenced by different technical factors, among which we have the connection speed, the server availability, the current traffic, the request contingencies, etc.

Nonetheless, this differs quite a bit from the value that Google and Facebook declare of (clear throat did someone say fake news?) between 12% to 15%. This gives us 1 bot click every 10, to remain conservative. Albeit the number seems small, even just 5% of rubbish is not easy to clean up. Every predictive algorithm needs to rely on a statistical model which, with such a margin of error, the risks to show the user non-accurate advertising run quite high. Take now into account that irrelevant or wrong communication can immediately translate into loss of reputation for brands and businesses. And millions of dollar in wasted resources, loss of opportunity and, ultimately, job losses.

My two cents thought

I don’t think ads or ads-based businesses are evil. I see the value of communication and understand the will of creating value for the users. I just don’t want to be scared by click-bait, not be able to tell if news are authentic or fake, and confuse nude pics for ads. Historical data is what allows progress to continue and I want it to make my life better. Internet enables me to choose and gather information in a vastly free ecosystem. I want to keep that freedom over anyone else’s opinion. I want to be curious and open to change whenever I can prove it’s worth it, and pour more energy into learning new or different things rather than moving the couch further from the fridge. If you feel me, go and try something different from time to time. Do not confine yourself to your comfort zone and let someone else make the choice for you. Try a different search engine. From time to time.