Marketing on steroids, thanks to Artificial Intelligence – and how your company can start planning for AI (part 1/2)

Artificial Intelligence (AI), along with the Internet of Things (IoT) and Marketing Automation, is one of the technologies that marketers plan to use most in the coming two years.^[1]AI can provide customers with a superior experience by allowing marketers to engage in a hyper-personalised and individual way.

Machines can do incredible things and often be better than humans: think about IBM’s Deep Blue computer beating Kasparov at chess in 1997 or Google Deepmind’s AlphaGo’s victory over go star Lee Sedol in 2016. But despite what Elon Musk is saying, AI is not the Terminator-like machine that will one day dominate humanity and replace your (marketing) team.

So, stop worrying and start getting excited.

The objective of this Point of View is to cut through the hype and see how marketers and AI can work hand-in-hand. In part 1, we will start by listing a few areas of marketing where AI is already applied today. In part 2, we’ll give you some tips on how to get your company ready for AI.

Artificial Intelligence Marketing

Artificial Intelligence Marketing

As there is no way to write an exhaustive list of AI-powered marketing practices (this Point of View has not been written by a robot…), we will stick to four areas:

Marketing automation
Marketing automation lets computers do a variety of manual and repetitive tasks, it helps marketers save time, target prospects effectively and optimise different stages of the customer journey. As marketers, we already love marketing automation. But AI can take it to another level.

According to a report^{^[2]} based on an analysis of over 235 million emails, the open rate for AI-powered emails was 228% better than that of the regular ones.

How is that possible? Well, machine learning can run through countless of points of customer data and establish what the most effective times to make contact are, what words in subject lines are likely to speak to the audience – and much more. Let’s take segmentation for example. In most cases, segmentation is done manually based on broad buckets such as personas. For “Women aged 30 to 40”, we can easily use a different subject line than for the “Men aged 50 to 60”. But what if you could personalise each and every email based on your customers’ interests, current location or even preferred time of receiving your communication? Wouldn’t that take the customer experience to another level?
Content marketing
Algorithms can map thousands of data points, such as a user’s website experience, browsing data, time spent on a page and topics of interest with the aim of creating and recommending hyper-contextual content most relevant to that person.

WordSmith, a natural language-generation platform, produced 1.5 billion pieces of content in 2016 and is expected to grow further in popularity in the coming years. WordSmith is capable of generating insightful content based on raw data (see how it works here).

Spotify’s Discover Weekly is a good example of how AI can keep customers engaged and continuing to subscribe with personalised recommendations. Each week, Spotify’s 150 million users receive a curated list of 30 songs that they have never listened to before but will probably love. Most of the time, it is so relevant that you have the impression that the Spotify team has handpicked the songs just for you; what a great feeling!
Predictive modelling
Predictive modelling turns large amounts of data into tangible insights and predictions towards future actions. Propensity models can be trained with data you already have and cover a large range of areas in marketing, such as churn prediction, conversion probability and lead scoring^{^[3]}. For the latter, AI can learn to score leads according to defined criteria so the sales team knows whether those leads are worth devoting time to.

Vidora.com puts AI in the hands of everyone within a company, from marketers to C-levels, by providing them with a simple platform for asking questions and using the answers to automate and optimise business^{^[4]}. Questions such as “Find all users 75%+ likely to churn & identify what new features should be sent to each over email” can be asked to Vidora. The algorithm will then run through the data and provide the requested list.
Customer service interactions (chatbots)
Chatbots, fuelled by AI, are dramatically changing customer service interactions. They seek to provide answers 24/7 to a demanding audience without them having to hunt down the information themselves. Chatbots can interpret consumers’ queries and complete orders for them.

You might think that chatbots are extremely difficult to develop and only large corporations with massive budgets and a big team of geeks will be able to develop them. Think again: at The House of Marketing, we have developed a chatbot for a client where each jobseeker is asked a few questions before the bot shows a list of vacancies the jobseeker might be interested in. In the very near future, we will see chatbots moving from text to speech, providing a whole new way of interacting with companies. What a time to be alive!

Now that you have seen some of the applications AI has to offer, you might want to start it off right away. Go to the second part of this Point of View and read about practical tips on how to get started with AI.