AI powered content: does it work for SEO?

Artificial Intelligence (AI) has come on in leaps and bounds since Alan Turing first asked, “Are there imaginable digital computers which would do well in the imitation game?” in his 1950 paper ‘Computing Machinery and Intelligence’.

In the 70-odd years since, computers have gone on to defeat chess grandmaster champions (IBM’s Deep Blue vs. Garry Kasparov in 1997), successfully complete the DARPA Grand Challenge in 2005 (race for autonomous vehicles over 100km in the Mojave Desert) and win against humans at Jeopardy! In 2011.

AI use in digital marketing is rising steeply

While these milestones highlight AI’s rapid progress, the technology is now starting to make a real-word impact.

When it comes to digital marketing, the enormous potential and central role AI can play is clear. In fact, according to Salesforce’s sixth edition of the State of Marketing report, the vast majority (84%) of marketers now report using AI, up from just under a third (29%) in 2018, a huge 186% adoption increase in just two years.

These stats won’t come as a surprise to many, considering machine learning algorithms enable marketers to filter useful insights from enormous amounts of data and deliver enhanced customer experiences, making it an invaluable resource in a data-driven world.

AI and SEO content generation

Covering AI’s role in the field of digital marketing within one session would be an impossible feat for the best of us; because of this, we will instead discuss and explore AI’s place in the SEO content creation process.

AI has long been a hyped-up buzzword in the SEO industry; but as is the case with many avant-garde technologies, opinion is split, and cases can be made by SEOs for and against the use of AI.

While AI is undoubtably an extremely useful resource in the research, data analysis and planning phase of content creation, when it comes to the actual writing itself, the waters begin to get murky.

On one hand, using AI to generate content is significantly less time and resource consuming. However fast, a human would never be able to write a 1000-word article in the space of a minute like an AI tool can. This means a certain amount of budget needs to be allocated to producing content which could instead be reallocated if AI were used. If quantity were the be-all and end-all, the benefits of using AI are clear.

On the other hand, in a time where ‘fake news’ and misinformation are rife, can widely available AI currently be used to generate high-quality, accurate and engaging content? And how will Google and other search engines treat AI generated content? Google’s mission is to “organise the world’s information and make it universally accessible and useful”, so what does that mean for the endless wave of content that can be churned out day in, day out by AI tools?

To get a better understanding, we used widely available AI content generation tools to create a number of articles around different topics, such as sport, fashion and pharmaceuticals. Using different topics allowed us to analyse how well the tools performed when varying levels of difficulty and accuracy come into play; aside from the fashion failure, many wouldn’t be too fussed if an article suggested wearing the wrong type of shoe with a certain dress, but inaccurate information could be extremely costly when it comes to the type and quantity of medicine to take for a headache.

The AI-generated content was then analysed and evaluated based on a variety of factors, from readability to accuracy, and compared to like for like content created by professional writers to see how the two matched up.

Alex Bova, SEO Operations Manager at Threepipe Reply.

Article first appeared in The Drum.