Artificial Intelligence is set to have a significant impact on our industry, and is already starting to transform how we do business. At Zenith, we made a big bet in the AI space two years ago with insurance client Aviva in the UK. Our pioneering data scientists worked out how machine learning could be used to attribute sales conversions to specific digital interactions and automate the most complex and time-consuming aspects of digital planning – ultimately resulting in increased efficiencies.

Since then, we have continued to evolve our capabilities in machine learning, earning ourselves a seat at the table with some of our top clients across Zenith and Publicis Media, developing machine learned models across a multitude of areas.

I had the privilege of participating in the International Advertising Association’s 80th Anniversary Conference in New York, as a panellist on the Artificial Intelligence & Cognitive Computing seminar. The panel also including a leading advertiser Diageo, and software technology experts, CA Technologies.

Clear themes emerged from our discussion: key things for marketers to consider before venturing into AI for the first time, or further down the path.

What does AI mean to you? What came across quite clearly in a discussion that looked at the point of view from agency, client and product development, is that everyone is still mapping out how AI, and specific techniques, can best serve their business and to what end. Defining how AI will bring value to an organisation is critical in developing some genuine capability. The most common examples of application of AI within companies were machine learning and deep learning, around content and personalisation.

Consider the personalisation opportunities. We have a unique opportunity to use machine learning models to deliver a more personalised experience based on user behaviour, content consumed, and other external factors (e.g. weather). Diageo developed a cocktail recommendation engine that suggests a drink based on the types of alcohol you enjoy. If we are playing the long game, getting personalisation right means brands develop better relationships with their consumers and ultimately drive growth.

What does AI mean for brand authenticity? As Uncle Ben once said, “with great power, comes great responsibility.” The more we all know about our customers, in theory, the truer to them we can be. Having said that, we discussed the need for people and brands to act responsibly, and don’t blame the technology when things go wrong. Be aware of what you are building and how it will be used. It’s also important to consider who has access and what ultimately ends up in market. If CA technologies is going to use machine learning to help support a more personalised message to their very niche B2B audience, it’s important the experience they deliver is  right, as they are in the business of selling software that is designed to do just this.

What about governance and security? Building AI propositions means that the use and storage of data is paramount to success. With all the recent concerns around data privacy and access, we need to reassure clients that we have strict governance and security in place to protect both their and our data. At present, we use blockchain to store the data output, in addition to our own cloud platforms including Azure, GCP and AWS where we apply our own encryption – this ensures protection and transparency. Usually used for crypto currency, blockchain is a very secure digital ledger, which gives all users the same viewing access, providing 100% transparency. With any modifications, blockchain retains the original and the new record and provides the sequence in which changes are made.

AI infrastructure: In-house or external partners? At present, most clients use third party expertise to help build their AI capability- from Amazon to Google, to start-ups, the marketplace is rife with option of partners to tap into. Our clients come to us because we have invested time, money and acquired the right expertise and infrastructure to build bespoke machine learning applications to address specific business challenges. One of Zenith’s most pioneering and successful projects has been our work for Coty, which saw the data sciences team in London build a fragrance recommendation engine in partnership with one of the UK’s largest retailers.

Where do you get the talent? The speed at which our business is changing means we now find capability in places we didn’t expect. Media agencies would not necessarily be top of mind to have hundreds of in-house experts in the data science and machine learning area. But that’s what we have done, competing with the likes of Amazon, Google and Deep Mind for the best talent. As you look to deploy AI solutions within your organisation, do you have the right talent in place to know what “great” looks like?

Clearly, agencies and brands are still figuring out what the world of AI means to their own organisations and the industry. No-one has fully cracked it yet, but it’s great to know that we at Zenith are at the forefront of a revolution. As we design AI solutions to help evolve our clients’ business and how they engage with consumers, this will only continue to develop.

Watch the highlights from the IAA discussion

Related Blog Posts