Don’t be afraid of AI, argues Jon Stubley (pictured below), VP ANZ at GumGum, you’re already using it and simply didn’t know it…
I am reading a whole host of ‘AI is the future of marketing’ articles at the moment.
Undoubtedly, AI is rapidly developing and will become a much bigger part of our personal and working lives; soon to be accelerated as and when the rumoured Apple Neural Engine chip is incorporated into next gen iPhones.
However, what many people don’t realise is that AI is already an integral part of the digital ecosystem as it exists today.
If you work in advertising, marketing or sales and after sales, then odds are you already use some form of AI every single day of your working life (and probably at home too). In nine ways to be exact. And here’s how:
AI allows search engines like Google (does anyone use a different one?) to display the right results, even when the user can’t articulate precisely what they are looking for. Google, for example, uses an AI system called RankBrain to guess what users mean when they’ve entered a query the search engine is unfamiliar with. It uses machine learning to translate the unknown words and phrases into more familiar phrases that have a similar meaning.
Programmatic campaigns use computational advertising (a series of algorithms) to deliver the right ad at the right time. According to Nielsen, two thirds of all advertising in Australia last year was done programmatically.
US lingerie brand Cosabella has gone so far as to replace its media agency with an AI platform called Albert. The platform optimises campaign spending autonomously based on real-time results, and makes creative recommendations when certain ad units outperform others. The brand has seen a 336 percent increase in ROAS and a 155 percent increase in revenue.
In-image advertising relies on a machine learning technique called a neural network, a series of learned mathematical functions that process information in a manner similar to the human brain.
By feeding millions of labelled images into a neural network, my own company (GumGum) has trained its AI technology to identify all kinds of objects, people, colours, concepts and brand logos. This way, the technology can place an appropriate ad inside every image. For instance, a parent might see an ad for notebooks within a photo depicting a school.
Brands are using large-scale data analysis to target customers based on their demographic information, past purchases, offline behaviour and online browsing history.
Coles and Woolworths use this type of predictive analysis for their weekly specials. It enables marketers to identify when consumers are going through major life events like pregnancy and moving home, so they can target them accordingly. It uses the same AI technology that led to a father in the US finding out his teenage daughter was pregnant after she received offers featuring baby products from Target.
Virtual sales assistants use artificial intelligence to communicate with potential customers. It means companies can collect information about potential customers at scale, before bringing in their human sales teams.
In Australia, telco Vocus recently announced plans to introduce a range of AI driven solutions to better manage and understand its customers, including using it for lead nurturing.
If you are one of the many people in Australia who already purchase from Amazon, then you will have been recommended a product with the message “customers who bought this item also bought….”
The personalised product recommendation uses collaborative filtering, an AI solution that links site visitors to other consumers who have similar tastes.
If you’ve taken an Uber from any CBD at midnight on a Saturday then you will have paid a premium as a result of AI powered dynamic pricing. It uses machine learning to set the optimal price point for a seller’s goods and services at any given moment, based on what people have been willing to pay for the product under similar circumstances in the past.
In the work sphere, Google’s dynamic price floors automatically adjust the minimum price a publisher will accept for a given impression, based on what buyers have previously paid.
Social media monitoring
Brands would be lost without AI when it comes to social media. Data analysis algorithms identify which customers are driving discussions online, while natural language processing uncovers the sentiment behind the thousands of messages people post about brands each day.
This means brands can be alerted to potential issues before too much damage is done.
Using the same technology, an ad agency in the US has created the Trump and Dump Bot that identifies Trump’s tweets mentioning publicly traded companies. Tweets likely to trigger a stock drop are shorted; any profits are given to an animal charity and the bot made this year’s Cannes shortlist.
Chat bots are now widely used for sales support functions across most sectors, with a study from Transparency Market Research estimating that the global chat bot market will be worth US$7.9 billion, up from an estimated US$627.7 million in 2015.
Jetstar, for example, launched ‘Ask Jess’ back in 2013 using natural language processing to conduct chat conversations that mimic human interactions.
If I had a bigger word count, I could also mention how it is being used in sports sponsorship negotiations, to craft marketing messaging and so many other applications.
For sure, AI is going to further transform marketing, media and advertising but it has already made significant in-roads.