AI-driven marketing: Google vs. Microsoft

Chris Middleton
Chris Middleton
  •  
  • 5
  • 28
  •  
  •  
  •  
  •  
    33
    Shares

In a previous article, I explained how devices such as Amazon’s Echo hub, Apple’s AirPods (combined with Siri), low-cost domestic robots, and other smart devices will use AI and machine learning to transform marketing rapidly and in unexpected ways.

One of the key transformations will be in switching many marketers’ focus from human to machine customers, competing for the attention of the vast eCommerce and fulfilment platforms that sit behind each of these apparently innocuous devices.

In this way – as I explained – a customer sitting at home will be able to say to her Echo or similar smart device, “Book me two tickets to New York, flying out Saturday, with four nights at a Midtown Manhattan hotel,” and the device will do everything else. The customer won’t have to lift a finger.

Household goods, food, travel, and much more besides, could soon be ordered in this way, and AI will help the machines – and the platforms and suppliers that sit behind them – to learn and even anticipate their customers’ preferences.

For marketers, that’s a real challenge as it cuts them out of the deal, unless their work has been done much earlier in the process, by placing products and services at the top of the list.

The same AI vs. marketing principles apply to those other tech titans, Google and Microsoft, as much as they do to Apple and Amazon (whose AWS platform sits beneath thousands of online businesses, and whose Alexa AI system has been opened up to partners), and to CRM giants such as Salesforce.com, whose Einstein AI platform is released in October.

Google’s own Assistant AI technology is being rolled into a variety of applications, including its Home hub, which is comparable in design and purpose to the Echo. The difference, of course, is the wider Google platform, the AI applications within which will extend to its new Allo messaging service, its Now machine-learning system, to Android phones, and to search.

Google claims that 20 per cent of US searches are already triggered by voice, rather than text.

In short, AI is rapidly becoming deeply embedded into Google itself – the algorithms behind which marketers already spend fortunes on trying to please and second-guess. Entire industries have sprung up around promising to get products and services onto page one of any search.

Google’s mission is to make information easier to find. But there are two problems. First, customers’ obsession with page one reveals that – in the real world – the majority of those billions of webpages out there are of no interest to most people, because Google has made us too lazy to search for them.

We want everything now, which leaves us looking at the vast landscape of human knowledge through a pinhole – a pinhole through which millions of advertisers are also trying to syphon their ‘Look at me!” messages. That seems unsustainable.

And second, like any massive object in the IT universe, Google’s gravity bends everything else around it. Organisations worldwide now change data – change facts – in order to make content easier for Google to find, which is distorting the fabric of information itself.

The result is an obsession with SEO, SEO-friendly terms, clickbait, and trending topics: machine-readable noise, rather than information that’s useful to human beings. Today, much of what makes it through Google’s page-one pinhole is noise.

Follow this process through to its logical conclusion, and it seems inevitable that Google’s ultimate purpose is to be the network and AI behind an almost entirely automated, machine-based economy. Its core revenue stream is advertising, after all; it’s no more in the information business, really, than any other advertising agency.

That this process will continue is inevitable, with the advent of a Google that’s powered by machine learning and voice recognition, as more and more companies compete to not only second-guess Google’s search algorithms but also its AI.

Which brings us to Microsoft.

The Redmond-based giant is in a fascinating position. The company that once held over 95 per cent of the desktop computing market in the 90s and early Noughties is nowhere in mobile, ceding dominance to Google and Apple – in the consumer space, at least.

For the millennial generation, Microsoft seems completely irrelevant.

But it still holds enormous influence in the enterprise computing market, where it has done a good job of transforming itself into a business cloud apps/services/collaboration provider. The fact that its pre-integrated products and services sit behind many business processes gives it a significant advantage.

CEO Satya Nadella has started to outline Microsoft’s AI vision.

In a recent keynote, Nadella explained how he intends to infuse every part of Microsoft’s portfolio with AI. As Microsoft is a key player in customer relationship management (CRM) and marketing automation, it’s clear that AI-enhanced marketing is going to be huge in 2017 and beyond.

But first there’s the fallout from its Tay experiment to deal with. You may recall, Microsoft’s chatbot learned racism, homophobia, and other hate speech from internet trolls earlier this year.

Nadella recently called these incidents “attacks”, but that’s nonsense: Tay simply failed to understand the nuances of human communication in a society where people are free to ask whatever questions they want – of a robot, at least – subject to the UK, US, and European laws that govern human beings.

Microsoft’s Chinese chatbot, Xiaoice, faced no such problems online, giving rise to an extraordinary possibility: AI and free speech may not be easy bedfellows. And this, of course, remains the challenge for AI: understanding the subtleties of human communication, which invariably have cultural as well as linguistic dimensions.

But one thing is not in doubt: marketers now have to deal with an AI-powered world, not to mention AI-powered machine customers.

Hey, Siri! What should marketers do next?

Written by: Chris Middleton

Leave a Reply

Your email address will not be published. Required fields are marked *