3 ways AI will change marketing as we know it
WE Communications Blog: Technology, WE Communications Blog: Social Innovation
As a millennial with a day job in digital marketing, I generally spend more time engaging with social media than I do with many of my friends or family.
This isn’t new or surprising. Since phones gained a personal pronoun in 2007 and photos started making it to Instagram in 2010, social media has grown into an integral part of our lives. Instagram currently has over 1 billion monthly active users (for context, as of February 2019, world population is 7.7bn) and Facebook as a whole continued to report growth towards the end of 2018, despite controversies.
But is what we see on these platforms chosen by us? Or is it chosen by Artificial Intelligence (AI) in order to sell us things in some obvious and some not-so-obvious ways?
The role of AI in advertising
Marketing and technology have long had an important relationship. While common public perception of AI is robots and virtual assistants such as Siri it didn’t come from mad scientists. AI was borne out of a close relationship between corporate decision-making and state power. Nowadays, one way AI is used is to help marketers make conversions, by giving consumers the best content at the right time, in the right place. This is known as hyper-personalization.
Hyper-personalization could mean targeting people on Instagram during payday with in-app shopping links. Or, it could mean serving personalised Google ads for dog food to people who have talked about their pets on Facebook. AI can also spot trends in your online shopping habits, and serve you adverts with direct purchase links as you’re about to click “buy”.
This is the huge benefit of AI. While it might take a team of humans months to analyse consumer behaviour and test activity against this, AI can learn as it’s used. It can segment customer groups as needed, creating unique algorithms from hundreds of thousands of data points (combining things like purchase behaviour, referral sources, on-site interactions and previous actions). This ultimately gives marketers a detailed picture of their market and actionable insights, in record time. For consumers, this means they see the things they’re interested in, when they want to see them.
However, AI does still get it wrong sometimes. One Amazon shopper was inundated with adverts promoting toilet seats after a one-off purchase. As she explains, “I do not collect them. I am not a toilet seat addict. No matter how temptingly you email me, I'm not going to think, oh go on then, just one more toilet seat, I'll treat myself.” But the more AI learns, the more it seems to know what consumers want to buy even before they do.
AI’s impact on social media
If you run a social media page, you’ll know all about the struggle to be seen. But appearing in newsfeeds can often be less about the filter you’ve chosen or how many times you’ve posted that day, and more about if you’re giving the algorithm – or AI – what it’s looking for.
Because the workings of algorithms are such a fiercely guarded secret, learning what makes them tick is a matter of trial and error. Engaging with similar accounts, adding a written description to images and talking to everybody who engages with your content can all help to let AI know what you’re about.
Sophisticated AI finds patterns in user behaviour and uses those patterns to find more content of the same type. In addition, the patterns AI learns can also be used to accomplish a staggering number of content-related tasks, such as replying to comments or tagging photos. Companies today use AI to both create and curate content, discover hidden trends, identify anomalies and engage with consumers.
Other important roles AI can play in communications include its use in determining whether people perceive a brand positively or negatively, by identifying key words and common themes in social posts and articles. Using this information, brands are then putting AI in charge of managing their social media crisis communications.
An example of this is KFC’s response to running out of chicken in 2018. Thanks to the careful use of strategic social data surfacing tools, KFC was able to respond quickly and appropriately to the supply chain glitch: “The chicken crossed the road, just not to our restaurants,” tweeted the official account, several days before any news coverage of the story went out. Individual social media users are often responsible for breaking news before traditional media outlets, and viral social updates can rapidly amplify the negative reaction. So what might begin – in the case of KFC – as a simple supply chain issue can escalate into a public drama in a matter of minutes if you don’t take action to shape the narrative – which is exactly what KFC used AI to do.
“Hi I’m Cora. How may I help you today?”
For two common concerns that every brand has – staying secure and keeping up – AI has a very important role to play.
Data breaches aren’t going away anytime soon, despite regulatory change such as the implementation of the EU’s GDPR. One of the biggest problems is a lack of cybersecurity experts. It’s been reported that the gap currently stands at around 3 million, while a report by ESG shows just 9% of millennials are interested in a career in cybersecurity. Where millennials can’t (or won’t) step up, AI will, with machine learning in cybersecurity technology planned to close the gaps in both numbers and diversity of skills.
From your bank’s chatbot to Facebook replying to DMs on its own, AI is helping brands fill gaps in customer service by giving marketers the power to provide the round-the-clock service that is often expected in today’s ecommerce landscape.
Through machine learning and advanced language programming, marketers can make sure the massive onslaught of customer service needs coming through to their platform are taken care of at a rate that would be impossible for a single human customer service rep. You can set up customer service bots to reply to your DMs out of hours, to answer FAQs, and to stay on top of replying to relevant comments.
So, what does this mean for digital marketers?
- While technology is rapidly advancing, and AI is getting quicker, smarter and more sophisticated, what was true of the very first computers remains true: a machine cannot – yet – create or replicate the emotion, experience or intuition of a human.
- As covered in WE’s recent position paper, “Artificial Intelligence and the New Imperative for Communicators,” AI is still a machine built by humans; it needs humans at the helm to make judgements and tell brand stories.
- For now, AI’s primary role is to help remove more mundane or repetitive work – and so it’s believed that there will always be tasks for which humans can’t be replaced by computers.
- Digital markers and other communicators need to understand its implications, look ahead, and build new strategies to help their brands to make the best use of AI.