Does Google Penalize AI Content? New SEO Case Study (2024)

Yes, Google has applied manual actions connected websites utilizing spammy, AI-generated contented created by Large Language Models (LLMs). There’s besides grounds that the algorithm tin observe immoderate AI-written contented arsenic well.

We Studied 487 Search Results to See If Google Penalizes AI Content

We studied 487 Google hunt results utilizing an AI contented detector (Originality.ai) for the astir competitory keywords and recovered that:

83% of Top Google Search Results Are Not Using AI-Generated Content

While this is simply a tiny lawsuit survey with a tiny sample, it shows that Google’s algorithms favour human-generated content.

Here’s a illustration of the keywords we utilized for the study:

This raises an important question:

Can Google observe AI content?

Recent algorithm updates whitethorn person proven that Google tin observe auto-generated contented created by LLMs (Large Language Models).

Anecdotal Evidence

The past spam update saw wide deindexation of full websites (many done manual actions) and algorithmic deindexation.

For example, we generated contented utilizing ChatGPT for the keyword “SEO grooming Houston,” and this contented registered arsenic 100% AI.

After the recent updates, it was wholly removed from Google’s index, which was deserving due to the fact that it was low-quality content.

So, we replaced it with human-generated contented and upgraded the quality.

Within hours, it was reindexed, and present it’s ranking successful the apical 10. This is an anecdotal example, but we’ve seen akin situations crossed galore campaigns.

It’s not ever deindexation, but it’s besides underperformance.

For example, we tested utilizing AI-generated contented for “SEO for dentists,” and it performed poorly for months:

So past I rebuilt it with NLP (natural connection processing) successful caput and utilized a quality writer to make high-quality content.

Here are the results:

In our acquisition moving connected countless campaigns and successful studying thousands of Google Search Results, we’ve seen that high-quality contented written by a quality performs amended than LLM content.

Algorithmic Preference of User-Generated Content (UGC)

It’s nary concealed that Google present loves Reddit, Quora, and different UGC-driven websites.

The question is, why? Google’s algorithm prefers UGC due to the fact that it astir apt doesn’t lucifer LLM-driven content. This is axenic speculation, but it’s thing to consider.

It’s not ever astir what Google punishes; it’s sometimes astir what is rewarded.

Common Sense

Google has 1 of the world’s astir blase ample connection models with Google Gemini. They’re besides rolling SGE (search generative experience) immoderate time now.

We’re talking astir a trillion-dollar institution that’s been astatine the forefront of AI exertion for implicit a decade. It would beryllium foolish to deliberation they can’t observe low-quality AI-generated articles.

They whitethorn not beryllium capable to observe AI contented with quality touch, but there’s a precocious probability that they tin observe AI-written contented that’s low-quality.

That said, Google can’t rotation retired algorithmic changes that are excessively assertive due to the fact that of collateral damage.

So, they volition apt lone crook the dial capable to onslaught the lowest-hanging fruits similar spammy contented that uses AI-generated substance without considering idiosyncratic acquisition oregon quality.

Does this mean you should ne'er make contented with AI penning tools?

No. In fact, we person AI-generated contented that’s ranking close present connected this website, specified arsenic “Are H1 Tags a Google Ranking Factor” and “Is Keyword Density a Google Ranking Factor?“

The cardinal is that you person to usage the AI exertion the close way.

Here’s how:

Review Google’s Spam Policies

Google is not against ALL forms of AI content.

For example, they say:

“Using automation—including AI—to make contented with the superior intent of manipulating ranking successful hunt results is simply a usurpation of our [Google’s] spam policies. That said, it’s important to admit that not each usage of automation, including AI generation, is spam. Automation has agelong been utilized to make adjuvant content, specified arsenic sports scores, upwind forecasts, and transcripts.”

And:

“Appropriate usage of AI oregon automation is not against our guidelines.” – Google Search Central

Finally, John Mueller’s take connected the discussion:

Google’s extremity is to amusement the astir applicable and adjuvant content

So, if you tin execute that extremity with AI content, past spell for it.

With Google’s webmaster guidelines successful mind, here’s however to usage generative AI to make people-first content:

Always Consider the Primary Purpose of Your Content

The sole intent of creating contented is to assistance idiosyncratic get person to their goals. It’s not astir you. It’s ever astir who is really consuming the content.

Too galore radical doing Search Engine Optimization (SEO) hide this part. They’re fixated connected hunt rankings but don’t adjacent see the people assemblage and what happens aft you rank.

You request to reverse this equation successful your contented instauration process. What is the extremity of the contented you’re astir to create? If the sole intent is to “get much traffic,” past you’re not reasoning from the close perspective.

The nonsubjective of contented selling is to make immoderate benignant of conversion.

So, if it’s informational contented astatine the apical of the funnel, the extremity is to determination the searcher deeper into your funnel. That means you’d similar to connection a pb magnet to get them connected your email list.

If you’re targeting commercialized queries, the extremity could beryllium signing up for a escaped trial, purchasing a product, calling for much information, oregon submitting a pb form.

Start with the extremity successful mind; your contented should ever lucifer the hunt intent of the people keyword.

Care More About the Quality of Content, Not Content Publishing Velocity

New technologies person enabled contented creators to usage automation to make contented astatine standard utilizing AI.

But this is simply a classical “Just due to the fact that you can, doesn’t mean you should” situation.

The information is that past repeats itself. Before Google’s Panda update, scaling bladed contented was each the rage.

But Google ended that erstwhile Panda slaughtered countless websites with thin, poor-quality content.

Here’s the deal:

If you privation sustainable SEO results, you request to absorption connected contented quality.

And conjecture what?

You can’t person some contented prime and scale. The concepts are opposed to each other.

It’s not imaginable to make high-quality contented astatine standard unless you person a monolithic squad of taxable substance experts and contented writers and an tremendous contented budget.

Google’s algorithm has the “Helpful Content System” built into it. This strategy is designed to reward high-quality contented and punish the opposite.

It’s acold from perfect, but work the penning connected the wall.

Focus connected quality, not quantity.

Stop Keyword Stuffing

The misuse of NLP tools has led to overoptimized content.

It’s not NLP, which is simply a proven concept; that’s the problem; it’s however these AI tools are used.

For example, I’m penning this contented wrong of Rankability and utilizing the NLP recommendations to physique thing highly applicable for the people keyword.

The cardinal to correctly utilizing an NLP instrumentality similar Rankability is to absorption connected a topic, not keywords.

I presumption the NLP recommednations arsenic gaps successful my content.

So, to physique a hyper-relevant asset, I request to constrictive those taxable gaps and marque my contented deeper.

At the extremity of the day, don’t interest astir however often you notation an nonstop keyword phrase. Focus connected narrowing the taxable gaps and constitute naturally.

Be Careful with AI Content Tools

There’s a flood of AI contented tools connected the marketplace close now. And I’m not needfully opposed to immoderate of them.

But I americium opposed to however radical usage them. It’s truthful casual to click a fewer buttons and make contented now, and arsenic a result, it breeds laziness.

And if you survey Google’s E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness) guidelines, they notation “effort” countless times.

In short, low-effort contented shouldn’t beryllium rewarded. And that’s precisely what these AI tools incentivize.

Most radical auto-generate and person small to nary quality intervention. This is simply a precise unsafe crippled to play.

So, if you privation to usage these tools, guarantee that the AI contented undergoes a human-led editorial process.

You request to adhd quality elements to it and marque it appealing to users.

Final Thoughts

Using anecdotal evidence, hunt motor rankings, and communal sense, I judge that Google rewards archetypal contented that’s intelligibly written by a human. Does it penalize AI content? Not each of it, but it apt has the quality to observe the lowest quality, auto-generated AI contented created via LLMs.