What’s next for Intelligence Automation in advertising?

Charlie Spargo's picture
by Charlie Spargo

Jason Gossett, Digital Strategy Director at Reprise - IPG Mediabrands' digital-first agency - address the challenges that exist within the AI market for companies and brands who are trying to innovate in this area, and why it matters in the advertising sector.

Throughout the 50s and 60s, there was an immense enthusiasm for AI research, and in the paper ‘Computing Machinery and Intelligence’ Alan Turing considered the question, “can machines think?” - something that still poses some considerable challenges today.

AI has been widely portrayed in cinema as something that embodies and exhibits human-like qualities. However, it is now the basis for a wide range of mainstream technologies including web search, smartphone applications, medical diagnosis, speech recognition and most recently, autonomous vehicles. 

A multitude of definitions 

The common assumption today is that AI is going to replace all of our jobs and automate everything. However, we have to look beyond automation and realise the potential of the technology to evolve the computing landscape, and exploit our creativity to innovate in the applications of AI and augment our own performance.

The concept of AI is perceived as complicated because the term ‘artificial intelligence’ has been used to cover a range of different technologies - and the meaning of AI has changed over the years since it was coined back in the 50s. Broadly speaking, there are two classifications that exist: 

  • Human-style intelligence: sometimes referred to as ‘general’ or ‘strong’ AI, this is very much the anthropomorphic representation we see on screen.
  • Task-orientated intelligence: sometimes referred to as ‘narrow’ AI, this refers to technology focused on a particular range of tasks with the aim of solving specific problems, from medical condition diagnoses to marketing personalisation. 

Opportunity lies in how we utilise the latter to augment human intelligence and improve our ability to take action or make decisions. 

Initiatives driving the growth of AI 

It’s evident that big tech companies can see the economic and social benefits AI could provide, in terms of increasing productivity in existing industries, and for creating wholly new products and services: 

  • Google created ‘TensorFlow,’ which is an open source AI framework. They offer machine learning (ML) pre-trained components through Google Cloud, and last year at their I/O conference Google introduced the ‘ML kit’ which is a software development kit for developers to integrate pre-built ML models into their apps, which can support functions such as text recognition and face detection.
  • Microsoft also opened up its AI translation capabilities in its Microsoft Translator app to developers. 

These types of initiatives are important in creating a sharing culture around AI, which should help to make it more accessible and easier to incorporate when considering a test case for a company looking to explore the possibilities and sell into the wider business. 

Brands augmenting intelligence 

Two-thirds of the opportunities to use AI are in improving the performance of existing analytics use cases. So how are companies incorporating AI for human augmentation? 

AI improves personalisation 

  • Thread is using AI to help people buy clothes; based on data from style quizzes and online stylists, customers receive personalised recommendations that they can vote up or down, and the algorithm uses this data to find patterns in what each customer likes and tailor its recommendations. Thread’s AI augments the approach for their stylists, sorting through volumes of clothes to find the best version of the item picked by the stylist. 

AI driving the connected customer experience 

  • Starbucks uses AI to link up with their rewards members' accounts and take into account things such as order history, customer preferences, weather conditions, time of day, holiday and even birthdays to make drink and food suggestions. They use weather data so granularly that they can predict tiny variations in demand store-by-store, adjust stock and display and drive sales accordingly. 

Long-term thinking is essential for AI to flourish 

Considering that almost 90% of the world’s data is estimated to have been produced within the last two years, with an increasing volume of information being collected from a greater range of sources, at greater speed than ever before - the need to develop meaningful insights from this data is incredibly important to the advertising industry, especially for the brands wanting to understand their customers on a much deeper level and provide the connected experience that people desire. 

Whether it’s Amazon trying to include more intensively automated shopping experiences with Amazon Go, or a Reprise and UM collaboration creating an enhanced emotional listening product using IBM Watson to determine the mood of customers using social data, there'll be an element of risk involved when deciding which areas of the business AI would benefit and the resource needed to implement a test case.

There are companies like Phrasee, who specialise in niche creativity for email marketing headlines, whose AI-optimised language is always in a brand’s voice; and Idio, who have a content intelligent engine designed to drive demand in the B2B content space for brands.

Such forward-thinking products offer huge potential when it comes to exploring the possibilities of AI, and we’re going to see more and more of this over the next few years as augmented intelligence continues to amplify human innovation.