“What Is Ethical SEO? - Search Engine Journal” plus 1 more
Posted: 14 Aug 2019 06:45 AM PDT
Disclaimers from the edge of the ethical rabbit hole
Humans crave ethics. Ethically-sourced. Ethically-raised. We're drawn to the idea of ethics because it gives us a framework to understand the impact of our existence.
Humans study moral philosophy whether they intend to or not. We use our judgments about the rightness or wrongness of actions as a foundation for making our own decisions.
The ethics we adopt help define us personally and as a society.
Faced with our limited imprint on time and vastness of existence, let's do what everyone does.
Let's Google it.
Hey Google, What's the Definition of Ethics in SEO?
If we expand out on this thought, there are some interesting nuisances.
Bots aren't "moral" devices. Neither are websites.
Each search engine has its own version of what is acceptable. Two search engines could have acceptable strategies or techniques that directly conflict. You only have one site.
Ethical SEO for one search engine would be unethical for the other.
This is where real-life humans come in.
Whatever decisions are made, the code will inherit their stands of good as bad. It will carry their blind spots and biases.
Every digital asset inherits the ethics and power of the business they represent, the individuals contributing code and commentary, as well as behavior patterns of their users.
This ethical imprint is in how we differentiate optimization from manipulation. Does this engine accept the practice?
Both involve research and code changes.
Structured data markup is a fancy tech term for slipping a search engine a cheat sheet. That cheat sheet would be approved by some search engines and not others.
Does structured data have moral value? It is an action and it has a consequence.
Creating a Human-Centric Definition of Ethics in SEO
In general, ethical questions don't have a right or wrong answer. Instead, it's about the thought experiment of taking a situation and holding it against a moral framework.
Plenty of theories, frameworks, and principles exist.
For the sake of brevity, let's look at the work of Immanuel Kant. He boiled down the ethical value of actions to a two-question framework.
If the answer to either of these questions is no, the action is unethical.
Code can't inherently decide whether to return a controversial app that allows Saudi men to control the movements of dependent women.
Is this app an example of modernizing misogyny or another form of relativism? The arguments for either side lead to deep rabbit hole full of human rights and technology that simply won't fit here.
Let's step back and take in a larger view.
If we use the same framework adopted in the definition of SEO, morality is relative to the guidelines. This app is relatively ethical to the Google Play store.
Apply Kant's framework.
Could we rationally will that others prioritizing actions relative to the guidelines that most benefit us?
You can argue that human rights override the morality of relative guidelines. By doing so, you would be exercising your self-agency. Good for you!
Chase that intentional moral choice by asking yourself when do human values override technological guidelines? Do ethics now have an order of operation?
There will never be a right answer, but you'll ask better questions every time.
In early 2019, TurboTax, a major tax U.S. tax service blocked search engines from indexing a page that would allow low-income users to file their federal taxes for free.
In business terms, they identified a low ROI audience (users making $34,000 or less, or are active duty military who earned $66,000 or less). It leveraged established brand recognition and cross-platform marketing messages to upsell.
The human consequences played like a bait and switch, if you searched for free federal tax filings, you landed on a page like this.
It's for a product known as the "Free Edition" and distinct from TurboTax Free File Program, which is part of the IRS's Free Filing Alliance. Despite the page's grandstanding, the user will pay $60-120.
ProPublica documented their experiments in creating various users. Each time, their faux tax filer would end up at the same dead-end requiring an upgrade before taxes could be filed.
You could argue that the practice targets military members and the economically disadvantaged. You wouldn't be alone. Lawmakers have openly challenged the company with the FTC and IRS.
The practice of using design and conversion pathing to coerce users to do something they didn't want to do is called dark patterning.
TurboTax used organic search as part of their strategy by adding robots noindex directives to the actual free tax service.
Only 3% filed using the free product – even though 70% of TurboTax's users are eligible for free filing.
Can I rationally will that everyone act as I propose to act?
Does my action respect the goals of human beings rather than merely using them for my own purposes?
Search Engines Are Imprints of Power
Websites and search engines are imprints of the ethics, biases, and norms of those with. Those that are successful become artifacts of power – or less somberly, brand recognition.
The longer that power endures, the greater its influence on culture and other developing technologies.
There's no better way to scale our ethical framework than to look at power and language.
In 2019, 54% of the internet is in English despite only representing 25% of users. The second most used language on the internet is Russian at 5.9% of all content.
That gap is so wide, it should have its own guided tours and gift shop.
That gap is so wide it impacts a lot of lives in ways we may not think of.
Language forms our reality. The words we use impact how we conceptualize the world.
When a new concept enters our vocabulary, we give it a name.
No better example of a well-seasoned SEO than seeing page view entered into the Merriam-Webster dictionary six months ago.
That absent definition did not phase us or impact our reality because we know what it means.
The phrase's transformation was powered by the digital community. Our language has the ability to adapt and welcome new concepts in.
Others are not so lucky. When new concepts and information aren't available in a language, it means fewer members of that society can engage or adopt them.
This is dangerous because some critical fields move at break next pace. Two such fields are medicine and technology.
When linguistic researchers studied an HIV prevention program sponsored by U.S. companies they found that over 80% of subjects in certain studies did not understand concepts such as placebo, randomization, and the ability to withdraw from research participation.
Can you ethically include a subject in a clinical trial who does not know what a placebo is?
The landing page for potential participants is indexed by Google and scores well for performance and SEO on web.dev. It's optimized and designed to bring in new users. At the time of this article, I was only able to find the content in English.
Kant would ask, "Does my action respect the goals of human beings rather than merely using them for my own purposes?"
Can we respect another human when asking they are consenting to an agreement they don't understand?
Translations, Cookies, & Communities
Can they agree to be tracked and remarketed to?
Do they know they might be given personalized content recommendations based on what they like? That the platform would what is free speech?
If your site hosts user-generated content, how can it address hate speech in a language no one on your team speaks?
Facebook's lack of oversight directly continued to the Rohingya refugee crisis. More than 70,000 people fled their homes. Here is Facebook's apology:
One could argue that platform growth without adequate translation or regulation has a larger societal impact than a translation budget.
A Translators Without Borders study details the impact of language on the European Refugee Crisis. Events requiring humanitarian organizations to step in are rarely polite enough to provide adequate resources and planning.
When thousands of asylum seekers are fleeing and the dust hasn't yet settled, the team adapts with their boots on the ground and a prayer in their pocket.
The best resource they have is being able to communicate to the humans impacted by war, disease, or disaster.
Not all migrants, asylum seekers, or refugees impacted by an event will speak the same language.
In a six-month space, 14,000 refugees arrived in Italy by sea from Nigeria – a country estimated to have 520 languages contained in its borders.
Knowing which languages and their respective percentage of the whole group could allow these teams to get the right translators and resources in place.
The study found that without objective information in their language, refugees relied on word of mouth and social media posts for critical information.
These refugees had internet access but couldn't find objective information in their own tongue. Imagine relying on a Facebook meme to get out of a war zone.
One could argue that engaging native speakers for content creation and community moderation is an ethical requirement for internationalization.
The human consequence of that ethical decision could mean picking a dusty and forgotten translation ticket out of the backlog and bringing it to sprint planning.
Paying for Rank by Mortgaging Trust
It's the ubiquity of search that gets us.
No official figures have been shared since 2012, but back then Google processes 40,000 search queries every second – even on an average lazy summer afternoon.
If search queries were snowflakes, we'd be buried in a blizzard we barely seem to acknowledge.
Users trust search engines because we believe they know the answers. They have the power to scour the web and the authority to sort out all possible answers.
SEO professionals are part of that process.
We nudge the system with better page titles and cleaner code to manipulate what answer is shown where and to whom.
We play with the fabric interconnected knowledge and fuss with the ubiquitous pillars of knowledge shoved in everyone's pocket. Isn't that oddly wonderful?
It also means we're on the front lines of the battle for user trust.
Freedom of Speech vs. Misinformation & 2020 Presidential Election
A user searching for [Joe Biden 2020] is likely interested in the presidential candidacy.
The SERP result directly below the candidate's real website is a "parody" site. It's the work of a consultant employed to make videos and other digital content for President Trump's re-election campaign.
This first-page parody put in the effort. While performance is painful, it nails a perfect 100 score on web.dev and an impressive 93 inaccessibility.
In the 2016 election cycle, we watched Facebook take the brunt of fake news backlash. Facebooks ads have a fast turn around.
As far as I know, no one has published data on the index coverage of these sites by Google.
Ranking beside your competitor for their branded query would be impressive for any SEO professional. I asked Twitter for stories from SEO professionals who've worked on political campaigns.
I expected to see tales of more wildly inaccurate news sites. Instead, SEO pros shared their stories of misinformation.
SEO Will Be Judged by Its Ethics Regardless of Whether We Talk About Them
I'm writing this now because we need to talk about it now.
SEO has gone from obscurity to established practiced and may be preparing for new levels of scrutiny.
Broken down, these presidential tweets assert:
These individual statements have to be aggregated with each unique person's understanding of the speaker's tone. It is not hard to read that unfavorable search results will be discredited.
If search loses credibility with users the most sanctimonious algorithm can stop digital strategies from triggering real-world consequences. Without a trusted source of truth, humans are more vulnerable to misinformation.
Ending the Critical Myth of an Unbiased Algorithm
Algorithms come in many flavors and all of them have biases.
Mix those initial blindspots with machine learning based on human interaction and you can conjure the worst in humanity. It took less than 24 hours for Microsoft's Twitterbot to morph from tween to filterless racist tirades.
These are learning pains that we have to acknowledge.
We need more languages ethnicities and gender identities represented on the teams building these tools. Failure to be represented could be catastrophic.
The lack of ethnic diversity in datasets for facial recognition software could play as out false identifications.
Facial recognition software in the U.K. has been highly criticized after it failed to include a "safeguard" (a.k.a. a set threshold that disqualified results under a matching threshold). Non-white subjects were more likely to be falsely identified.
Ethics requires that we step beyond our immediate environment. Disregarding users from datasets now can devalue or limit their contributions in future iterations.
"Works on my machine" is not an option.
Promote Transparency About Personalization
When you're only given back results you agree with, it impacts your perspective.
The code splits you into buckets. You become persona A or persona B.
The result of this false dichotomy is palpable.
What Is the New Definition of Ethics in SEO?
You were warned there would be no answers here.
As SEO professionals and all of tech have conversations, more resources are coming together.
You can check out the EthicalOS toolkit to learn about the major ethical risk zones and thought experiments to help future-proof your tech.
Apply an ethical framework to a proposed action.
Ask if we could rationally will that others prioritizing actions relative to the guidelines that most benefit us.
Create civil discussions about when human rights override relative guidelines.
By doing so, you would be exercising your self-agency. Good for you!
Chase that intentional moral choice by asking yourself: "When do human values override technological guidelines?"
Do ethics now have an order of operation?
There will never be a right answer, but you'll ask better questions every time.
All screenshots taken by author, August 2019
Posted: 12 Aug 2019 07:32 PM PDT
Google users who searched for various terms related to the recent mass shootings in El Paso, Texas, and Dayton, Ohio, saw virtually no right-leaning news sources in the three "Top Stories" results that Google shows as the top results for news-related queries, according to a study by AllSides.com.
Instead, Google returned results favoring the left-leaning CNN, the study showed.
AllSides, which attempts to provide balanced perspectives by mixing sources biased left, right, and center, searched words such as "Shooting," "El Paso," "Dayton," and "Guns" multiple times on three days between Aug. 5 and Aug. 7 in the aftermath of the shootings.
Of all the "Top Stories" results, 70 percent were from sources on the left, while about 18 percent were from sources AllSides considers to have center bias. Some of those, such as NPR, USA Today, and The Associated Press, have been characterized by others as left-leaning, including in the subjective chart created by investigative journalist and TV host Sharyl Attkisson.
About 8 percent of the results came from sources that haven't yet been rated for bias by AllSides. The remaining 4 percent came from sources biased to the right.
The study noted that a quarter of all the results came from a single source—CNN. Second was The New York Times with 14 percent. Third was The Washington Post at 7 percent.
The search for "Trump" was skewed the least, showing around 57 percent of sources biased to the left, around 35 percent to the right, and the rest in the center.
AllSides acknowledged that the study used a small sample size, but noted that it "tracks closely with prior data on Google bias," including its own report on Google News bias from last year (pdf) and the data put forward by researchers from the Northwestern University.
AllSides said the results of its study don't mean that Google is "suppressing voices from the right" in relation to the shootings intentionally.
"It may be that a lack of right-leaning news media overall accounts for the huge difference between left and right-leaning appearances in Google News, or is perhaps an unintended consequence of Google's algorithm," it said.
Google has denied infusing political bias in its products but acknowledged that Google News gives preference to news sources the company deems "credible."
The company's workforce skews heavily left, with more than 90 percent of their political donations between 2004 and 2018 going to Democrats, according to data collected by Business Insider and political data firm GovPredict. It showed a similar situation at other tech companies, including Facebook, Apple, and Amazon.
Evidence of Bias
Intentional or not, the evidence that tech companies favor left-leaning politics in their products has been growing for years. Internal documents, interviews with insiders, and hidden camera footage have shown that Facebook, Google, Twitter, and Pinterest suppress content using vague, sometimes secret criteria that reflect the political preferences of the companies.
Robert Epstein, a psychologist who has spent years researching Google's influence on its users, has published research showing that just by deciding the sequence of top search results, the company can sway undecided voters.
This led to 2.6 million votes shifting in the 2016 presidential election to Trump's opponent, former Secretary of State Hillary Clinton, Epstein determined. He warned that in 2020, if companies such as Google and Facebook all support the same candidate, they will be able to shift 15 million votes—well beyond the margin most presidents have won by.
Trump has reportedly been working on an executive order to address politically biased censorship by social media companies.
|You are subscribed to email updates from "how to know the most searched words on google" - Google News. |
To stop receiving these emails, you may unsubscribe now.
|Email delivery powered by Google|
|Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States|