In case you haven’t heard, the robots are coming after your job. In fact, according to Elon Musk, they may even bring about the end of human civilisation itself. In all this gloom and doom, it’s understandable that you might be somewhat wary of artificial intelligence (AI) or machine learning.
But what you may not realise is that AI and machine learning are already powering many of the interactions you have today. Think the Product Recommendations on Amazon, or the “We just added a show you might like” emails from Netflix, or AI assistants like Apple’s Siri or Google Assistant. Meanwhile, machine learning is vastly improving facial recognition capabilities for brands like Facebook and Google, and helping Uber work out how long rides will take to arrive.
And it will become even more ubiquitous, with Google recently announcing that it’s moving from a “mobile-first” company to an “AI-first” company – a signal that should make us all sit up and take notice. In other words, AI is here to stay, and the rest of us need to either get on board or risk getting left behind.
Rather than feeling threatened, marketers need to embrace the possibilities afforded by AI. As Google’s VP of Marketing Martin Chow says, AI and machine learning could get us much closer to that elusive marketing Holy Grail: “relevance at scale”. With AI and machine learning comes the potential to create truly personalised, one-on-one experiences for hundreds or even thousands of customers with ease.
And while research firm AlphaBeta predicts that 28% of marketing officers’ jobs are susceptible to automation, that still leaves 72% of the fun stuff: building customer relationships, analysing customer and marketing trends, and developing marketing strategies. In other words, the odds of you being replaced by a machine are pretty low.
In this post, we’ll look at some examples that show where AI and machine learning in marketing is currently, where it’s going, and how you can take advantage of it for your own business.
But first, let’s quickly cover the basics.
What is artificial intelligence and machine learning?
Artificial intelligence is a machine’s ability to think and act “intelligently”; that is, to solve a problem or carry out a task in the way that a human might.
Machine learning is a subset of artificial intelligence, and refers to a machine’s ability to learn by looking at data, without explicitly being programmed with sets of rules.
For example, facial recognition software learns by being fed datasets of labelled pictures, and trying to guess which contain faces and which don’t, until it can eventually identify faces with a high degree of accuracy.
Eventually, with enough data, the software can even learn to recognise who is in the picture (this is how Facebook’s tagging prompts work).
Where is AI and machine learning in marketing at now?
As you can see by the definition, artificial intelligence is a wide discipline that can have infinite applications. Here are just some of the ways AI is being applied in marketing:
AI-powered personalised digital experiences
By analysing hundreds of data points about a single user – location, demographics, device, pages/content viewed, behaviour, and so on – AI can create a personalised digital experience for that user, surfacing relevant content and using push notifications to deliver the right message at the right time on websites and social media.
You’ve probably experienced this yourself, in the way Facebook curates your News Feed based on what you’ve read, or the way Amazon recommends other books based on your past purchases, or the way Spotify recommends and creates playlists based on your listening habits.
In fact, consumers are now so used to this type of digital experience that they have come to expect it from other brands. According to the 2017 Real-Time Personalization Survey, 88% of marketers surveyed said their prospects and customers expect a personalised experience – and marketers are delivering, with 33% of those marketers reporting that they use machine-learning algorithms to deliver personalised web experiences. Of those not using it, 32% said they would start using it in the next year.
One such AI-powered personalisation platform is Blueshift, which can use learned data about an individual to tailor the customer journey, using a predictive point system to determine when a user starts that journey, and offering personalised content along the way.
Expect to see personalisation become more pervasive and more powerful, as AI becomes sophisticated enough to integrate data and create a holistic customer journey across all touchpoints.
AI-powered chatbots and assistants
We’re undeniably in the era of mobile messaging. According to Business Insider, the big four messaging apps (WhatsApp, Facebook Messenger, WeChat and Viber) have more monthly active users than the big four social networks (Facebook, Twitter, Instagram and Google+), with over 3.5 billion users globally. It makes perfect sense, therefore, that marketers would try to harness the power of conversational commerce to connect with customers via this technology.
Chatbots are able to mimic human speech and converse with customers to answer queries and provide assistance with various transactions, making them the perfect way to create engaging brand experiences at scale. In some cases, they may even outperform human customer service representatives, as they can access millions of data points and aggregate that data to spot patterns and make predictions – capabilities that humans would certainly struggle with.
Facebook especially are gunning to be the go-to platform for branded chatbots, with plenty of tools to help businesses build bots quickly (and without having to learn how to code). (To learn more about Facebook Messenger bots, check out our post: ‘To bot or not: can Facebook Messenger bots work for B2B?’)
There are also virtual sales assistants, such as Conversica, which are able to reach out to leads with highly personable emails, interpret responses, ask questions to further qualify the lead, schedule appointments with sales reps and send summaries of its interactions to those reps. One company reported an ROI of $20 in new contracts for every dollar spent on the system.
AI-powered pay-per-click (PPC) advertising
If you’ve ever used Google AdWords, you’ll have seen AI in action. Google AdWords has an automated bidding system, which allows advertisers to bid for the lowest possible cost-per-click for keywords. But, for the most part, PPC ad campaigns are managed in-house or by agencies.
But now we’re starting to see the emergence of AI-powered systems that can manage an entire PPC campaign.
Albert, for example, is an AI platform that can target audiences, optimise ads and bid autonomously, with minimal human input.
There’s also Frank, another AI-powered tool, that uses machine learning to find the best channels for your target audience, and buys ad placements in real-time auctions.
Machine-learning algorithms that predict user churn
Preventing customer churn is essential to a business’s bottom line, especially if you have a subscription-based business. Machine learning algorithms and predictive analytics can analyse data to work out when customers are in the most danger of churning, and take appropriate action, whether that’s offering assistance, providing relevant content, or enticing them with an offer or a discount.
Urban Airship, for example, analyses mobile customer behaviour to identify the most loyal customers, as well as those with a high probability of churning, allowing marketers to take action to target those users.
There’s also Vidora, which helps to identify dropping customer engagement, as indicated by behaviours like lower usage time, and sends relevant emails, push notifications and offers to keep those users engaged.
What is the future of AI in marketing?
Manual A/B testing will be a thing of the past
Rather than pitting one ad against another in a “fight to the death” bout of A/B testing, Google AdWords has recently introduced optimised ad rotation.
Here, you have a group of ads, and Google uses AI to consider many variables, such as the user’s query, device and location and the historical performance of the ad, to determine which of the ads in your group will perform the best.
In some situations, it might be ad A; in others, ad B – and this way you don’t simply lose the clicks that B would have otherwise generated.
Improved data visualisation and analysis
One of the biggest advantages of AI is that it can help marketers handle ever-increasing amounts of data, and perform complex computations well beyond human capabilities. But one problem that still plagues marketers is the fact that data is still very often trapped in silos. Trying to integrate the data across platforms can therefore pose significant challenges.
AI will be the key to not just unlocking the data from the various tools and platforms that marketers use, but also using this data to create a more holistic picture, and identify patterns that were previously indiscernible.
Automated data visualisation will also become more user friendly, giving marketers an even deeper understanding of their data, so they can be even more confident in their ability to make decisions that affect the business.
AI-influenced consumer behaviour e.g. voice and visual search
As the capabilities and popularity of AI assistants grows, they will start to have a marked effect on the way in which people search for information. For example, rather than typing “Indian restaurant Surry Hills” into Google, they’ll instead ask their virtual assistant, “Hey Google/Alexa/Siri, what are some good Indian restaurants in Surry Hills?”
We’ll also start to see more widespread use of visual search. This feature is already available on Pinterest – by simply zooming in on the object of interest, you can search for similar results. So, for instance, if you like the look of a lamp in someone’s living room, you can do a visual search to try to find out where you can get one for yourself.
As voice and visual searches become more widely adopted, this will have marked effects on SEO, and brands will have to rethink how they are “found” in this new search landscape. In fact, Gartner predicts that by 2021, brands whose websites support visual and voice search will increase their digital commerce revenue by 30%.
How can I take advantage of AI and machine learning in marketing?
Despite what you might think, it’s not just large companies with teams of developers at their beck and call who can afford to take advantage of AI and machine learning in their marketing efforts. There are things you can do now without incurring too many costs.
And even if you don’t feel that your business is quite ready for AI yet, you can think about how to get it ready – the martech landscape is growing in leaps and bounds, so while some AI technologies may be out of reach for now, you can bet they won’t be for long.
Here are a few things you can do today:
Keep up-to-date with marketing trends
Between 2016 and 2017, the martech landscape in the US increased by about 40% – that stat gives you some idea of how quickly things are evolving in this space.
In order to keep up with the latest and greatest in AI marketing technology, it’s a good idea to stay on top of marketing trends, whether that’s by subscribing to blogs (like ours), going to events or joining some relevant LinkedIn groups.
Pay particular attention to what the big players are doing, and what’s producing results, as well as any emerging martech startups that may make the technology more accessible to SMEs.
Explore your marketing automation platform
If you’re using a marketing automation platform, odds are there are some AI features tucked away somewhere. HubSpot, for example, uses AI to power Predictive Lead Scoring and machine learning to power Content Strategy. Does your platform have AI features you’re not taking full advantage of?
Experiment with chatbots
Chatbots are a relatively easy way to dip your toe into the world of AI, particularly on Facebook Messenger, as Facebook provides plenty of development tools to help you build your bot.
They can be a great way to carry out basic procedures and provide answers to relatively simple customer service queries, but be wary about supplanting your customer service altogether.
Consider how to centralise your data
The businesses who get the most competitive advantage out of AI will be those that can aggregate their data, and use AI to draw inferences between varying aspects of their business that otherwise would not have been possible. If your data is currently in several silos, consider how to combine it onto a single platform.
Brave new world
With the advent of AI comes lots of exciting new opportunities for marketers, in terms of automating repetitive and monotonous daily tasks and creating seamless personalised buyer journeys to a degree never before possible, affording you and your team time to focus on the strategic and creative aspects of marketing.
Those who embrace this brave new world, and who are prepared for the changes to come, will be the ones who reap the rewards.
There are many aspects of inbound marketing that AI can assist with – but strategy isn’t one of them. If you want to see how your inbound marketing strategy measures up, then give our inbound marketing self-assessment a go. It only takes 5 minutes!
Brand chemistry is a strategic inbound marketing agency that goes the extra mile to deliver results for our b2b clients. Our inbound marketing specialists are HubSpot certified and use the latest techniques to provide our clients with a steady stream of relevant new leads.