The Foundation of AI and Advertising
To make advertising more effective, AI (Artificial Intelligence) now makes it possible to target to the nearest possible unit of user preference.
Research shows that 50% of all global companies will use one form of automation or the other by the year 2021. The advertising industry is not left out. Every forward-thinking executive wants to cut costs and maximize profit. This singular reason is why many are in business.
Google and AI Advertising
To make profits rising in the age of intelligence, we must appreciate how Artificial Intelligence technologies and applications are making this possible. Google makes 86% of its revenue from search advertising. Of its $110 billion profit in 2017, less than 13% came from its other subsidiaries (Gmail, YouTube, Maps, and so on).
The multibillion-dollar company remains predominantly an Advertising company. Google is one of those companies, that therefore, does not undermine how Artificial intelligence can make part of its work easier. Google doesn’t only aim to make this easier, it also wants to continue to dominate the art of doing it better than all its competitors.
The company maintains a monopoly because its AI algorithms for search advertising are exponentially better than Microsoft’s Bing, or Yahoo. How does Google run page ranking when you run a search query, or which YouTube Video to play next, as well as the choice of advertisement to play in the middle of a music video a user is obsessed with?
Google Predictive Algorithms
Google employs a series of predictive algorithms, cloud-powered web crawlers for indexing pages and multiple behavioral analysis tools to determine user preferences. If you have made a Google search today, it means they do the job well.
Once you make a search, empowered by this same AI, the company then tailors the ads to you and you can find them “following” on every other website that carries Google ads using your meta-data. That’s how Google makes customer recall ads, making you see this over and over again. That’s the power of digital.
From 1998 when Google started, we have faster computers today, more storage, and better inference engines. To make advertising more effective, we can target to the nearest possible unit of user preference.
What movies they love, who their favorite celebrities are or who they went to high school with. Facebook’s ‘Find a Friend’ algorithm helped hundreds of millions of people restore connections but also helped the company create the most effective documentation of how human social connections work.
Geotagging and AI Advertising
Geotagging, or Geotagging, is the process of adding geographical identification metadata to various media such as a geotagged photograph or video, websites, SMS messages, QR Codes or RSS feeds and is a form of geospatial metadata.
Over a billion people are linked by geography, past schools, love for ice cream or kickboxing. The company’s artificial intelligence understands what connects all these people and how merchants and campaigners can reach them easily.
Without writing a line of code, you can use intelligent analytics, to see how many people your advertisement reached and whether they behaved in a certain way (clicked the link to your product – a term known as a conversion, or simply scrolled away – what is often referred to as a ‘bounce’). Companies intending to create recall must tailor content to be both interesting and trendy, with a definite Call-To-Action (CTA).
AI Integration in Traditional Advertising
There are several other applications of artificial intelligence in advertising. With internet data exploding, companies can save plenty of time in research by using AI tools to read through thousands of textbooks, past editions of newspapers or magazines to find distinct keywords, images of people or videos. We can save more time and focus on creating unforgettable experiences.
Billboards used for traditional advertising can now carry Arduino powered computers chips and tell us how many people looked directly at our ad on a storefront; whether they have been here before and how long they stayed when looking.
AI (Artificial Intelligence) API
Image recognition modules released by Microsoft in its Azure computer vision API, will scan handwritten English text from the seventies (70s), and convert to computer generated text in Spanish or Mandarin. The same module can identify keywords, familiar images of people and who they may look like or be linked to in the current age.
The ability of computer vision APIs from Microsoft (and indeed other companies) to distinctly analyze sentiments like happiness, sadness or indifference, in real time, while a customer yet stands in front of a traditional screen ad (positioned in a public place) is unlocking real-time insights.
Machine Learning Applications
A company in the United Kingdom called “We-see-through” uses machine learning (an application of AI) to unlock this type of insights for large companies running traditional advertisements in the United Kingdom.
They can take regular videos from store cameras, to determine how long it took for a client to make a buying decision while holding a liquid washing soap from A&B Limited. Though this area of application is met by regulation roadblocks in different countries, the practice has proven that it isn’t so difficult these days to basically put a “this store is under surveillance” notice outside.
This is why several countries are drawing up regulations on how data is being gathered, and for what purpose it is being mined. Using machine learning, based on historical data, we can predict future buying behaviour, and advise a brand on whether they should take their advertising online or offline.
Based on celebrity fan pages, we can determine which celebrity should be on our next series of marketing videos. With the help of machine learning, engineers can predict if a video will go viral by studying the science of virality. The application is as far as we can think.
Gaining Speed and Reducing Time to Market
As a small business, you can start from applying simple tools that have already been offered by Microsoft, Google, Facebook and Amazon web services. One that intrigues me is “Design Ideas” by Microsoft PowerPoint.
While this is generally a presentation tool, advertising professionals can generate quick slides for clients, by just typing what the presentation is about, and click on the “Design Ideas” tab when connected to the internet. Using Artificial intelligence provided by PowerPoint, you can present results for a campaign in languages you do not speak; in simple, crisp ways.
The Google Data Studio and Microsoft Power BI can help you generate amazing insight from user data about customer churn, or which product will do better in the next fiscal year and which one needs more advertising budgets.
Facebook advertising dashboard is user-friendly, it will allow you to target ads to the nearest preferences. Microsoft Word’s ‘read aloud’ feature can help you when proofreading an ‘ad’ copy before you post. Grammarly uses machine learning to help you construct grammar better and write enjoyable content. In addition, QuillBot paraphrasing tool uses its state-of-the-art machine learning paraphraser in order to function as an article rewriter. Aiming to enhance the state of natural language processing by making AI agents sound more human.
From small companies who just want to use AI to make their communications and advertising departments better, to large companies that are willing to launch research on why a product sells better on Amazon, there are tools to use. From a few hundreds of dollars to millions of dollars, custom made applications can be built to make your processes of creation and evaluation of advertising way better than it is today.
To Attain Recall
In the future (which includes today), the focus of advertising will be to help people to remember your brand long after you are gone despite the information overload from traditional and online sources.
As businesses compete for customer “Ear-Time”, customers are expected to process more information than they can handle, the advertisement must, therefore, be sticky, having a capacity for recall. If we combine the various methods above, from predicting with data and AI what a user might like, to attaining speed by reducing time to market, we are definitely on the path of doing effective advertising in 2030.
The use of hashtags in advertising for social is a step forward from the keyword. To keep people finding the same ads (recurrence) you must use a series of string to tie them up across the digital nation-space called the internet. This is why hashtags are getting more popular not only for Instagram and Snapchat but for LinkedIn advertising.
Ever heard of mixed reality? As people get more visual, millennials particularly have a knack for look and feel. Augmented reality is the future of advertising, not virtual reality. There is an influx of digital headsets with the ability to superimpose a virtual reality on our real-life environment.
In that future, billboards will completely become mixed reality boards. When you have your goggles on, you see a different set of ads from when they are off. You can scroll through a billboard, at your own pace, click a link on it and take action. Any company looking to do advertising correctly must use this Omniscience channels to achieve the desired result. You can order a simple virtual reality headset, a 360-degree camera for less than $200. Or you can go big and order the HoloLens 2 from Microsoft for $3500. Whichever way, whatever investments you make show that you are aware that advertising is no longer business as usual.
AI is essential to effective advertising campaigns. Leading to reductions in acquisition costs and an increase in customer lifetime value.
In the coming years more and more, content will be designed, and campaigns executed automatically in real-time based on AI algorithms.