By Imad Sarrouf, Head of Publishers, DMS
The word “AI” (AI) can be an umbrella term that covers a variety of machines that learn, with the help of human or on their own entirely. In this way, AI technologies perform certain cognitive tasks like humans or even better. AI-powered machines can read and understand text, see and identify images, detect sounds and understand them, move around obstacles physically, and sense their external environment like temperature.
For example, Gmail and Google Docs now use AI to read what you are typing and then understand it well enough to recommend what to type next with Smart Compose. Facebook uses AI to identify who is in your photos, then recommends whom to tag. Tesla Self-driving cars use AI to identify obstacles and drive (hopefully) in a safe and effective way. Siri on your iPhone uses AI to understand your voice commands and create responses that make sense. Smart home technology, like Nest, use AI to sense changes in their visual field and then take action based on what they sense.
Over the last few years, AI has shifted from being a branch of computer science to an everyday tech that almost all consumers carry in their pocket on daily basis, such as a smartphone with voice input Siri, bixby or Google Assistant. The AI technologies has been gaining significant interest in the advertising and media industry in the past few years. Although the applications for AI in the advertising and media business remain in infancy stage, there is huge potential for the technology to form the next generation advertising and media sector. Machine learning includes a significant effect on the advertising ecosystem already. The best example is real -time bidding (RTB), which is buying or selling advertising space online in real time. Through AI, self-learning algorithms are used for running online campaigns that allow advertisers to critically identify the potential user. This further helps in deploying personalized ads to each user and in taking the desired action.
The demand for AI in media and advertising industry has been increasing due to greater demand for rich digital user experience. Media and advertising companies are focusing on optimization of ad campaigns, email marketing and website content customization by ongoing learning from user behavior and actions. Furthermore, programmatic advertising could be fueling the growth of the global AI market in this industry. Automated process of buying and selling ad inventory through AI is connecting advertisers to publishers effectively.
The trend toward effective voice based search tech has shown benefits on the marketplace growth. Companies are trying to bring next gen input method within their services and product through AI. The widespread adoption of smartphones has let media companies to add AI technologies as an essential part of their product development lifecycle. AI offers a new competitive edge to the media companies through contextual relevancy allowing you to connect the right users to the proper content at the right time. Therefore, the global AI market in media and advertising carries a huge scope for rapid growth.
Globally, the US and Europe witness greater demand for AI solutions in the media and advertising sector, due to robust technology and IT infrastructure. The market in Asia-Pacific is expected to witness the quickest growth rate in the same forecast period, due to huge growth opportunities in media industry and emergence and rapid adoption of predictive technologies by major media and advertising companies in those markets.
Though, AI allows big media and ad agencies to serve cognitive ads and integrate voice based assistance within their campaigns, there continues to be scope for small and medium agencies to totally adopt AI as their prime functional and operational business element.
Companies are buying AI technology to better understand users proactively. With greater adoption of AI in those industries, consumers are likely to better communicate and engage with content through speaking apps and chat bots in near future.
A few of the major players in the global AI market in advertising and media are IBM, Microsoft, Google, NVIDIA, Intel, Sentient Technologies, and Numenta.
The many use cases for AI in advertising
AI is crucial to the infrastructure that underlies digital ad products on many platforms, though users might not see it or interact with it always. Modern programmatic platforms use AI to often control real-time ad buying, selling, and placement.
Ad Exchanges all use AI to regulate the selling and buying of advertisements in real-time. Which includes programmatic exchanges, third-party networks, and advertising on social media platforms. AI dictates the way the campaign budget is spend also, who sees the ads, and how effective the entire campaign is.
Performance and spend optimization
Performance optimization is among the key use cases for AI in advertising. Machine learning algorithms are accustomed to analyze how your advertisements perform across specific platforms, and then offer suggestions about how to improve performance. In some cases, these platforms could use AI to intelligently automate actions that you know you need to be taking based on best practices, saving you significant time.
In advanced cases, some tools automatically manages ad performance and spend optimization, making decisions entirely on its own about how best to reach the ad campaign KPIs.
Dynamic Creative Optimizer (DCO) is an AI-powered system that may partially or fully create ads, based on what is most effective for the campaign goal. This functionality has already been present in a number of the social media ad platforms and through third-party tools designed for premium publishers, which use intelligent automation to build advertisements and serve it based on the links the brands are promoting or the target audience, location, time of day etc.
Upcoming tools will use smart algorithms to create the ad copy itself. These systems leverage natural language processing (NLP) and natural language generation (NLG), 2 AI-powered technologies, to create ad copy that performs as well or even better than human-written copy—in a fraction of a time and at scale.
Ad targeting matters just as much as the ad copy and creative. Thanks to Data Management Platforms (DMP), we have a healthy set of consumer data touch points with which to target audiences. But manually doing so isn’t efficient. That's where AI can automate the procedure by looking at past audiences and ad performance, weighs this against the campaign KPIs and real-time performance data coming in, and then identifies new audiences likely to engage with the ad.
Human Resources AI Technology
Human resources (HR) is an industry that has been rather slow to adopt AI and ML technology. The one exception has been in the implementation of applicant tracking systems (ATS) that use ML techniques to perform application screening for potential hires. That alone has led an industry of AI-enhanced services to improve potential applicants’ chances of landing an interview and saved HR time by automating CV reviews. The thing is, the increase in ATS adoption is just the tip of the iceberg and much wider implementation of AI and ML technologies in the land of HR is expected in the coming years.
AI-Powered Marketing Tools
As the world is getting into an always-on internet connected reality with the rise of the Internet of Things (IoT), companies are becoming aware with the fact that there are more marketing channels to control than previously. The only reasonable solution is to carefully turn the bulk of the ongoing work to AI-powered marketing systems, using ML to change and evolve marketing efforts over time.
Already, such tools are available in all phases of the marketing industry, from social media management to content marketing and everything in between. That, however, may be the beginning. Businesses, which have already seen how Chatbots and AI Call Centers have influenced marketing decision-making, are actually looking into methods to turn more of their marketing efforts to AI-powered solutions. In some cases, AI platforms was advising the sales rep on what things to offer during negotiation over the phone depending on the buyer tone of voice and input (more discounts, free items etc.). Recording of thousands of customer support conversation was fed to the Machine Learning (ML) systems to allow such assistance to real life cases.
Again, such application needs to have scale to make sense (an Amazon or bank that have thousands working in a call center or have millions of transactions each day).
Financial AI Solutions
AI and ML tech adoption in the world of banking has been so immediate and complete that it produced a completely new business category of Fintech. Specifically, asset managers are going all-in on the tech, as are hedge fund managers, financial advisors and the entire banking sector. JPMorgan Chase sees AI as critical to their future success. Until recently, identifying and pulling in relevant data to train AI models was taking up around 60% of the time of the bank’s growing army of data scientists. JPMorgan Chase pulls in data from around 7,500 external sources along with leveraging its own information.
That was an inefficient usage of an expensive and relatively scarce resource. A relative brand new data platform the bank has developed, called OmniAI, is helping it to get relevant data into its models faster (consider data warehousing as a prerequisite before dwelling into an ML project). OmniAI does not merely let data scientists obtain practical raw data due to their models quickly, but it also automatically verifies that the data being used is in accordance with various regulations (like privacy).
To sum up, some of the expectations accumulated on AI, can amount to smoke and mirrors. The way certain vendors in Silicon Valley and elsewhere, suddenly turned into AI providers is a question mark. Many companies somehow transformed from a data management company or a workflow processing company to an AI company in a single year. How?
Getting tangible value from AI requires a strong staff, leadership, compliance, and support across the group. AI also needs accurate data to work with; otherwise, it is like buying an Aston Martin and have no gas (it will cost you but will not take you anywhere).