CMO.com recently published this article by our Innovation Director, Tom Ollerton, looking at the automation of the creative process. They've been kind enough to let us reproduce it below.

Imagine if you had a tool that could automatically research your audience, find the insight, write your brief, come up with the idea, write your copy, design your ads, edit your video, write the music, and put it live for you — all in a few minutes. Well, that future is already here for those who know where to look. Whether it’s something you want to implement for your brand — that’s a bigger question.

The term “artificial intelligence” is currently all over the marketing press. It’s often pitched as a bit scary for marketers—machines will be using the swathes of consumer data now available to us through social channels to deliver what a marketer can, but more quickly and efficiently. So what are brands actually going to do with this threatening force, and will it really be able to replace you or your agency?

AI Miracle?
Global lingerie brand Cosabella was recently on the receiving end of a lot of high-profile PR for ditching its digital agency and using Albert.ai to automate its social with great results, such as an increase in social return-on-ad-spend of 50% and a decrease in adspend of 12%. However, if you look behind the PR spin, the real story is that Cosabella provided all the creative assets, and the tech was focused on distribution. While it’s seductive to think we can automate the creative process, there are a lot of considerations, and limitations, to doing so.

The two stand-out opportunities for AI in marketing right now are automation (like the Cosabella example above) and its best pal, personalisation. In both these areas, there are realistic and useful solutions for brands that can make a difference to a digital marketing strategy. I’ve spent the last six months meeting over 50 startups who offer some incredible technologies in these areas, such as Zo, which can create Facebook content based on a variety of data points as well as create your next post in front of your eyes so you can tee it up and watch it fly, and Picasso Labs, which uses AI to analyse your market’s social images to tell you which visuals, colours, and composition will drive the best ROI.

Personalisation is, arguably, even more exciting — I see Dynamic Social Video (DSV) as the next biggest opportunity in marketing. This technology, from companies like Idomoo and Photospire, allows you to send hundreds of different versions of a video to narrow target audiences, without having to make numerous edits.

No Match For Human Brain
But while all these technologies can potentially make marketers’ lives easier, AI is far from a complete solution. There is no machine that can replicate a human brain and truly understand our needs, wants, and desires — after all, it’s hard to replicate a system we don’t fully understand ourselves. The spectrum of emotions required for advertising to stimulate its customers range from the rational “50% off — click here” to complex tasks such as getting under the skin of consumers and identifying culturally relevant insights that underpin your whole strategy.

So how should brands use and enjoy the benefits of new technologies? First, learn from past examples. The social media community, in particular, is unforgiving, and will seek to disrupt brands that get their strategy wrong in whatever way they can. Brands like (our client) Domino’s have had success by creating a chatbot character Dom with its own tone of voice that can’t be corrupted by trolls with bad intentions. Just like crowdsourcing, when you’re dealing with any element of mass interaction, boundaries are very important.

How To Make AI Work
Let AI deal with the less emotionally complex tasks, freeing up time for the best human talent to spend on work that requires empathy and a contextual understanding that a computer can’t attempt to recreate. Recognise that to make the most of these opportunities, the perfect solution is currently the synthesis of human emotion and machine efficiency.

Seriously consider investing in personalisation. The key to becoming more personalised is utilising your own first-party customer data. Plugging this information into social ad platforms will allow you to target specific users, and will give your brand the opportunity to develop personalised messages for your target audience.

And, finally, don’t ditch your agency. A good agency will be encouraging you to experiment with new technology, not hold it back due to fear that it’ll take their jobs. Their true value comes in understanding how to make your business resonate on an emotional level with consumers, and this is one area the machines are just not ready for. Yet.