In our article at the start of the year predicting marketing trends we expected to see during 2018, we made a brief mention of the growing use of artificial intelligence in commercial applications for both sales and marketing purposes. Here we take a closer look at how it is becoming ever more important.
Artificial intelligence (aka AI) might conjure up visions of machines that look, sound and think like us humans, which (as far as we know) are still confined to the pages of science fiction. In the context of practical real-world applications today, AI is where a device perceives its environment and takes actions that maximise its chance of success at some pre-programmed goal, without the need for further programming. In other words, it learns and adapts as it goes – hence an alternative term for AI that might be better to keep in mind is ‘machine learning’.
Even this more basic version of AI might sound far away from how marketing is currently being delivered in 2018, but it is already being used in a rapidly increasing number of ways. According to research by Salesforce, just over half (51%) of marketers already currently use AI in some shape or form, and an additional 27% are expected to incorporate the technology by 2019. You don’t necessarily need to understand how it works, but it is important to understand its general applications to maximise the potential for its future use in your business.
We see three aspects of marketing where AI already plays an important and growing role:
1. driving consumer traffic to websites
2. providing tailored content to customers and prospects according to their needs, wants and interests
3. engaging directly with consumers to help them make better buying decisions
The line between these three aspects is sometimes blurred – for example, sending out content tailored to consumers will itself generate additional incoming traffic to a website – but with that caveat in mind, let’s have a look at how AI is affecting those three activities.
Generating website traffic
If you want to see a truly startling statistic, visit www.internetlivestats.com/google-search-statistics, which keeps track in real time of how many Google searches are made each day. The number changes far too quickly to read, because Google responds to about 1,000,000 searches every 15 seconds. It would obviously be outside the scope of humans to deliver superior quality results to all these searches, even if a sizeable proportion of the world’s population was dedicated to the task. Google has always relied on computers to do the work, but whereas in its early years it relied on programmes based on simple rules and logic, now it relies heavily on AI.
For consumers, the impact of AI is that they get better results to their Google searches. It’s already hard to recall how hit-and-miss product searches used to be a few years ago, unless you knew the product’s exact name. Now, not only will a few descriptive terms produce better results, but those results will include videos showing the product in operation, online places where it can be bought, and maps showing local stores with it in stock. Even if you make a spelling mistake within your search, Google’s algorithm is clever enough to work out what you meant and deliver the right results more often than not.
Google’s AI programme is called RankBrain. It uses natural language processing (NLP) – i.e. the language both to help identify what an internet searcher is really after and match to content across the whole of the web. The result is that keywords consciously added to websites have become less important in attracting traffic to websites in comparison with excellent quality naturally-written content – which is why we spend so much time harping on about the benefit of creating fresh blogs and other forms of content on a regular and ongoing basis. These days, a review comment posted on a site could be the basis of Google including that site at the top of their research results, rather than anything specific that the website owner might have done. This is why we urge our clients to have website content written with their customers, not Google search bots, in mind (see 5). In effect, it’s because the search bots themselves now ‘think’ much more like a regular human does.
Directing tailored content
Traditional advertisements and promotions rely on a rather scattergun approach. Some logic can be applied to attempt to match them with their likely audience (for example, TV adverts for stairlifts are broadcast during cheaper slots during the day on certain channels, when older viewers of a certain demographic are likely to be interested in them form a larger proportion of viewers). As more and more people have turned to the internet to make their purchases, the reach of marketing has not only been extended – to anyone in the world with an internet connection – but also deepened, in that adverts and other promotions can be far more closely targeted. This being said, we should not dismiss TV advertising completely – it still works extremely well for some companies, often as part of a much larger integrated digital and non-digital marketing plan.
AI plays a large part in digital searches, by comparing the online browsing activity of millions – even billions – of people and then predicting what adverts everyone will be most likely to respond to. As with web searches, the scale of this activity would be beyond human capability.
Let’s say someone is thinking about going on a safari holiday to East Africa and researching possible providers online. Naturally, they will easily come across information about, and adverts for, providers of such holidays. But AI will also see to it that they start seeing adverts and information about insect repellents, sunscreen, lightweight luggage and clothing, phrasebooks, camera equipment, and so on. Not only that, the mix of all these items will be different for each consumer, because each will have a different browsing history indicating which of these items are likely to be of most interest to them.
The systems behind what pops up on to your screen, and when, are called ‘recommendation engines’, and use AI in increasingly sophisticated ways. For example, Netflix not only knows which movies its subscribers have watched, but also how many times they have been watched, whether they were viewed in single or multiple sessions, and so on. This then determines what other films are suggested for viewing, with each playlist being different for each user.
Sometimes, the uncanny way in which adverts directly relevant to what’s in our minds appear can feel a little spooky, but the fact is that consumers generally find that recommendation engines enhance their online experiences. Instead of having to laboriously research topics unaided, the search engines throw up shortcuts to information likely to be useful. Part of the reason why so many people shop on Amazon is that they are offered plausible alternative suggestions to the items they are searching for. Spotify will offer music it thinks its users will like based on AI analysis of tastes and habits. Although both Amazon and Netflix are vast enterprises where AI is essential to promote their enormous range of products, even small companies can benefit from AI through the programmatic adverts offered by the likes of Google and Facebook. These will tailor material and present it to a carefully selected audience – all handled by machines, of course – to help get the best return on your ad spend.
Engaging with consumers
This third impact of AI on marketing is probably closest to the science fiction idea of machines that think and act like humans. What AI offers here is something that directly takes the place of a human, to help consumers obtain information and make better buying decisions. A growing number of ‘chatbots’ are handling one to one online conversations with consumers. For example, Chatobook is a chatbot available to restaurants which integrates with Facebook Messenger to take reservations, show menus and promotions, answer questions and collect feedback, all entirely without human input. Chatbots like this can help small businesses to upscale without having to take on large numbers of staff, and potentially without sacrificing a huge degree of customer service (although care must be taken with regards to this last point to make sure that internal processes are in place to escalate and deal with any customer issues that cannot be resolved by chatbots as efficiently and effectively as possible).
AI is also behind the trend towards interactions with the internet being through voice commands and responses. Electronic personal assistants, like Amazon’s Alexa and Echo are already installed in many homes, and it makes complete sense that speech, our favourite means of communicating with one another, should also become our normal means of communicating with our machines. Getting to the point where machines can understand speech, though, has been a real challenge to AI software developers. Humans speak different languages, in different local idioms and accents, and in voices that vary in tone according to our age, gender, physique, etc. While voice recognition is a long way from perfect, it’s an astonishing achievement that it exists at all. What are the implications for business though?
Well, Amazon refers to Alexa having ‘skills’, but these rely on what are, in reality, apps that have been built based on actions being controlled by voice command. So, if you’re thinking about what app could enhance your business, think too about making it voice-responsive. Also, perhaps consider how you name new products: one reason Alexa has that name is because it is the one-word command that starts each user interaction – and it contains the relatively rare ‘x’ sound. This helps to make it stand out from other words; in the same way, products will be more easily found by the likes of Alexa if they have names which are distinctive.
AI is here to stay
Some may not like – even fear – the impact of artificial intelligence, but that has been the history of every recent technology. The recent demise of several high street chains is a timely reminder of how businesses that fail to keep up with technology are likely to struggle to survive. AI will have a huge bearing on how marketing is done in the future, so it’s better to accept it now, and use it to your advantage as soon as possible.
Keep up to date
The world of marketing is changing all the time. To keep up to date with all the latest news is almost impossible, but you can keep on top of things by following us on social media – Twitter, Facebook and LinkedIn. If you would like to have a conversation with us about the use of artificial intelligence or anything else to do with marketing, please contact Stephen Brown on 020 7795 8175 or email email@example.com – you can also visit our website at www.abacusmarketing.co.uk to find out more and to view our work with other clients. We are more than happy to meet up for a two-hour chemistry meeting at a venue of your choice without cost or obligation to discuss your business challenges and marketing requirements in more detail.