ai-content-header
Best Practices

Automated Content Writing: Can It Really Save You Time?

7 December, 2017
4 min read
Magnus Linklater

There’s a growing sense that Artificial Intelligence (AI) poses a serious threat to job security, with rising concern among a broad spectrum of society; from retail workers to lorry drivers, town planners to legal secretaries, the foreboding message is clear: the robots want our jobs.

However, while automation is ubiquitous throughout the marketing world, from targeted email campaigns to scheduled social media posts, it’s hard to believe an algorithm could ever conjure up a successful content marketing strategy.

The creativity, imagination, analysis and planning involved means the majority of marketers won’t be lying awake at night fearing R2-D2 handing over their P45.

Indeed, a 2015 BBC report, citing an Oxford University study on the future of employment, revealed marketing professionals are among the safest third of workers, with only a 33% chance of automation causing unemployment (compared to a 99% likelihood that telesales staff imminently face mass redundancy).

BBC Report 2015 - Likelihood of automation for marketing associate professionals

Incidentally, the same study notes how marketing directors face a measly 1.4% risk of automation, highlighting how, when it comes to strategy, robots simply can’t match our real world experience (not for now, at least).

However, worryingly, there is an area of marketing that’s increasingly being outsourced to AI: content writing.

 

Automation stations

I fully understand the attraction of robotic writing, as it promises to save time and money. There are a million other things I could be doing right now instead of writing this article, important things that could help grow my company, and while I heartily advocate the multiple benefits of business blogging, there’s no denying the writing process is incredibly labour-intensive.

It’s one thing having a clear idea of what you want to say, and being able to communicate it in an effective, eloquent manner that ‘speaks’ directly to your audience while simultaneously ticking all the crucial SEO boxes. Not to mention the amount of research and editing required to produce outstanding content.

Thus, if you stumble upon software that promises to make the creation process ‘quicker, cost-efficient and much more pleasant,’ while also guaranteeing ‘unique, proofread, high-quality content from scratch, simulating a real human writer,’ you may be forgiven for having your head turned.

Those are the claims of Articoolo, a New York-based automated content machine that’s been the talk of the tech world.

However, upon reviewing their homepage, any hope you may have of this being the answer to your content creation prayers should immediately dissipate. Check this out for ‘proofread, high-quality content’:

Grammar and spelling mistakes on Articoolo's website

It’s likely these grammatical mistakes are the result of human error,  but they certainly don’t bode well. For a service that supposedly prides itself on quality content production, you’d think they’d have the nous to use a rudimentary spellchecker when reviewing their sales copy.

Still, let’s see what’s on offer when we sign-up and ask for an article on the topic of ‘automated content writing’. Here’s an excerpt from the piece that took less than a minute to deliver:

Content Mash-Up vs Truly Distinctive Writing Previous generation article writing applications generally based on Mash-Up techniques for Search engine optimization content creation.

Is that even a sentence? And what’s with the flagrant use of uppercase letters mid-sentence? Here’s another articulate snippet on the theme of ‘content marketing’:

A highly effective content advertising plan can certainly helps you earn more money than you did in 2016. A road map for success in business in 2017 is a digital map path.

After reading that, I find it hard to believe an automated writing service could ever ‘helps’ me earn more money, no matter how easily it may slot into my ‘digital map path’.

To be fair to the people behind Articoolo (assuming real humans run the company and it’s not all AI), they say the platform provides a ‘quick, coherent starting point for your articles’, the subtext being they’re aware that what you end up with won’t be ready for human consumption.

But doesn’t that undermine the whole automated writing concept? On the evidence of the above excerpts, the text requires substantial editing, rendering the whole exercise rather pointless. Not to mention the fact that statements aren’t backed up with research, and there are no sources referenced or linked to.

If all you end up with is spluttered jargon, why bother? The ‘quick, coherent’ claim rapidly goes out the window.

 

Spinning yarns

Despite repeated warnings on the dangers of duplicate content, Google recently raised eyebrows by investing in a content automation programme, seemingly giving legitimacy to robotic writing.

Ultimately, search engines are on a mission to return only the most valuable resources at the top of the SERPs, so any form of duplication causes confusion, meaning your copied content is likely to be filtered out of search results.

A shady way around this is ‘article spinning’, whereby you use a piece of software to exchange words for synonyms, giving the impression of unique content. Many local news websites regularly undertake this practice, using content published on other sources and ‘spinning’ it for their own use, and this is the area Google has now stepped into, working on ways to turn raw data into palatable, readable content.

So while ‘spinning’ does, on the face of it, remove the risk of direct duplication, it still seems inherently dishonest. For example, I could take Sendible’s piece on quick ways to create new content ideas, and ‘spin’ it in a few seconds to publish on my own blog.

However, upon using Spinbot, this is the quality of material we’re left with:

It's anything but difficult to become mixed up in subjectivity while making new substance points. A few thoughts feel extraordinary when they're initially considered, at that point neglect to perform.

Different circumstances, thinking of new substance thoughts includes excruciating – and significant – hours spent in gatherings jotting contemplations into a scratch pad, never to be seen again.

The shared factor among most substance creation forms is that they're frequently established completely in subjective feeling, without selecting the capable effect that confirmation and information can have on making new substance thoughts.

I don’t know about you, but my ‘substance thoughts’ on this are, “What a load of rubbish!”

Regular content production inevitably presents a business challenge - it’s time-consuming and expensive - but if you rely on automated content, you risk harming your reputation. Would you be happy publishing the above under your name? I certainly wouldn’t. Does it save time? Not if you have to completely rewrite it before it makes sense.

 

Robot wars

To reiterate, I’m not a complete Luddite campaigning against progress. I fully appreciate and embrace the many wonders of machine learning and automation, such as using AI to improve social media campaigns. However, when it comes to content writing, you can’t replicate the human touch.

Even if the technology improves in the future, it’s hard to believe robots could ever get to grips with your brand’s tone of voice and create meaningful content that intrigues and inspires.

Content writing isn’t simply sitting down and spitting out words. There’s much more creative and strategic thinking involved, looking for ways to add value to your target audience in a manner that qualifies your expertise.

It takes time to get right, but the payoff of great content - written with SEO best practice in mind - can be substantial, breeding loyal customers that keep coming back.

No algorithm can match your wisdom, so take out a scratch pad, jot down your contemplations, and focus on publishing your own ‘substance thoughts’ rather than relying on robotic words.