Although this may sound alarmist, content professionals are nonetheless keeping a wary eye on the automated journalists and digital content creators currently edging towards our jobs.
With robots historically confined to the online retail, logistics and manufacturing industries, ‘Robot writers’ are a fairly new innovation- one that those of us in the content industry are not particularly welcoming of. After all, no algorithm could possibly emulate the sophistication and sensitivity of a human writer. Or could it?
In fact, the market has recently seen an influx of content production platforms, like Wordsmith and Articoolo, which promise to do just that: write human quality content on an automated basis. But like most new marketing tools, it’s worth taking the time to assess their usability and how they respond to our client’s needs.
The case for automation
When it comes to automation, our gut reaction is inevitably one of distrust. There are multiple factors to take into consideration when crafting content: marketing objectives, tone of voice, target audience, empathy and emotional context. It’s hard to see how a computerised solution could take all of this into account when churning out an 800 word article.
And yet it’s easy to see why there’s an increased interest in automated writing. Content marketing can present a conundrum to many businesses: namely, producing enough fresh, new content at a high enough standard on an ongoing basis within budget.
Marketers drawn to the benefits of automating their customer management, social channels, lead generation and email marketing have made a logical conclusion – that content is next. It’s time consuming and resource-heavy. Automation can help with all of this.
And as usual, investment starts with early adopters such as Facebook and Google. The latter party have always traditionally frowned upon programmatically generated content, but Google are investing in developing automated content solutions. In responding to the modern appetite for digital content, particularly news articles which lend themselves well to algorithms, these companies are essentially giving endorsement to other automated platforms currently offering the same computer-generated content services. And Econsultancy’s report from 2016 on The Future of Content Marketing indicates that although customers in the UK don’t like the inauthentic nature of automated content, they are willing to engage with it if it’s designed and delivered with intelligence.
From a production perspective, automated solutions remove the tricky aspects of crafting content: time management of writers, dealing with loss of creativity or ‘writer’s block’, sourcing reliable content professionals – and that’s just with written content. Image and video content requires additional, specialised skills and software, not to mention hours of painstaking effort, to create.
Again, it’s not difficult to understand why marketing management, struggling with full schedules and ambitious output targets, might increasingly see automated solutions as an answer to some of these issues.
The human touch
The problem with all of this is obvious: an algorithm simply cannot replace a human being when it comes to tone of voice, empathy and an innate understanding of an audience’s needs. Well, not yet.
What we’re talking about in a nutshell is the underlying principle of content that converts: authenticity. A robot writer by its very nature lacks authenticity. For the increasing number of businesses who rely on sincerity to inspire brand loyalty, being an authentic voice in the marketplace is what works, and it’s what is going to continue to work in the future.
It’s the equivalent of speaking to a human being on the end of a telephone instead of an automated helpdesk recording. We crave human interaction in most areas of our lives, and robotic content often leaves us feeling underwhelmed and uninspired.
And we’re seeing evidence of this in how people in the industry are discussing automation in relation to content. This agency admits to being approached by automation companies who readily admit that their words are lacking a certain something, and want to add back in the ‘human touch’.
To be fair to the machines, they can do a pretty good job of generating content that conveys numerical data, such as market reports and sports results. As a support to a busy outreach schedule, this might be invaluable- providing the content is edited by a human.
What robots can’t do is find the personal angle in a story – which, after all, is what makes content compelling in the first place.
What’s on offer
An algorithm that can churn out fresh, unique content that appeals to search engines and humans alike is the content marketing equivalent of the Philosopher’s Stone of alchemy – much sought-after, but stubbornly elusive.
Automation began with the old-fashioned ‘article spinners’. These take input text (e.g. a previously written web article) and ‘spin’ it into new, unique pieces of writing – in theory. In actuality, the results tend to be pretty underwhelming. Here’s what happens when we put the previous paragraph through a high-profile online article spinner:
A calculation that can produce new, interesting substance that interests to web indexes and people alike is somewhat similar to the substance showcasing likeness the logician’s rock of speculative chemistry – much looked for after, yet wilfully tricky.
‘The logician’s rock of speculative chemistry’ is wonderfully entertaining, but it illustrates our point nicely. In this instance, the robots have failed.
However, a new breed of potentially game-changing content creation services that use artificial intelligence to mimic human writers have materialised.
One such service is the aforementioned Articoolo, which promises to create ‘unique, proofread, high-quality content from scratch, simulating a real human writer’. Simply describe your topic in two to five words, set the length, and the platform rustles up an article for somewhere between £1 and £2.
Articoolo uses a type of AI called Natural Language Generation, or NLG. Essentially, NLG is the practice of applying context to analytics to come up with a ‘human’ voice- we’ve seen this already with innovations like Siri, who can decipher spoken commands.
We decided to put Articoolo through its paces, and tasked it with writing an article on content marketing:
While the end result was remarkably proficient for a computer-generated article, it did require a lot of editing to make it sound less…well, robotic. And while the article conveyed most of the information that we required (apart from actually defining ‘content marketing’), it was missing a certain je ne sais quoi: humour, flair, and pizzazz. When you’re writing about B2B topics in particular, those things are necessary to avoid a rather dry read.
The automation revolution doesn’t just stop at text- even video is getting the automation treatment. Platforms such as Wochit give you access to a huge library of stock footage, infographics and video content from news agencies like Reuters and Getty, which you can then format and overlay with your own corporate branding and messaging. Again, experienced video marketers would tell you that this misses the point- video is one of the best mediums through which to communicate your brand’s story. A mashup of stock footage is unlikely to compel or convert.
No need to worry just yet, human writers
The insatiable desire for more and more content has arguably already resulted in a ‘blanding’ of our public discourse, as quantity threatens to prevail over quality.
But this is exactly why we writers and content specialists shouldn’t worry too much about the robots just yet. In an increasingly vast sea of ‘stuff’, consumers respond to content that stands out, content that compels, inspires and educates. And so it follows that our clients still need us humans. They need unique brand voices, they need insight and experience, and they need thought leadership when it comes to the words they want us to write.
We humans aren’t going anywhere for a while.