I'm a marketer studying Machine Learning at MIT - here’s why

Why every Marketer needs to study narrow AI - or fall behind - Image courtesy of Massachusetts Institute of Technology

Why every Marketer needs to study narrow AI - or fall behind - Image courtesy of Massachusetts Institute of Technology

Machine Learning and “narrow” AI in general, are the most important subjects for any marketer and C -Suite exec to understand, they are bringing about seismic changes to life and business in general as well as in marketing. Yet in the course i’m doing on Machine Learning at Massachusetts Institute of technology (MIT), there are very few marketers - why?

I clearly remember the first day I read about Machine Learning, specifically a Whitepaper on Neural Networks. It was like someone had taken time to explain how I think and laid it out on a page. A process of absorbing data (in the case of marketing, cultural and consumer behaviour data) through my senses, then re-sorting and re-analysing that data and creating unique, visual systems, pictures and patterns to make sense of them - a mixture of science and art - to end up with a Point Of View on a brand and a foundation for a brand and marketing strategy.

I still use massive pieces of paper, sometimes running for meters, to sort through and create patterns in my thinking, so when I discovered that Neural Networks can do all of this in an equation based, automated format, it meant I could take ideas and thoughts I had, on trends and cultural insights, ask the Machines the questions, and see if the patterns I was feeling / seeing were right or not.

Even better than that, the algorithms can go off and find connections and “clusters” of alignment, that we as humans have no chance of finding, interpreting the world around us via maths in a far better way than we ever can - at least “in the narrow”, and present them in easy to understand, beautiful pictures.

Suddenly, I had powerful tools that I could control to analyse almost anything and anyone. Polling & focus groups never worked for me, they just don’t cut it, but Machine Learning mixed with Social Media and other data deliver highly effective tools to interpret culture and the motivations behind consumer behaviour, to test content, create dozens of customer clusters, develop hundreds of varied pieces of content, enabling me to “test in the narrow” and then use the results to lend support and inform the strategies and innovations I was recommending. This had a transformational impact on the response from brands and their confidence in the work, and right at the centre of this, every time, was Machine Learning.

In 2018 I had the privilege of working with many global blue chip brands, and had a pitch win rate of over 90%. These wins were a team effort and all had great creative, strong and visionary CEOs and great account handling teams, without which they would not have been a success. However, the feedback from the clients when the pitches were won, always included their excitement about the opportunities Machine Learning had for their business, their fascination with the new insights presented, again because of Machine Learning, and it has to be said, slightly disturbed about the disruption Machine Learning was going to cause in their Industry, which I normally present in a 3 year window.

Neural Networks - re branded as “Deep Learning”, only came to prominence in 2012 when Geoffrey Hinton created an iteration that was revolutionary and enabled the supercharging of this “narrow AI” and the products and services that have been developed using this system since then are helping us move toward an age where the Virtual, Digital and Biological combine - the 4th Industrial Revolution.

For now, we have “Narrow AI”, where the algorithms can do one specific task infinitely better than any human, so long as there is enough data, and a clearly defined goal or “Target”, and it is changing everything in marketing. At every point of the strategic marketing mix, insight, planning, channel choices, content creation, content strategy, paid adverting, influencer campaigns, outdoor, experiential, retail, all of them have Narrow AI tools that if applied will beat any competitor not using them.

Theres a fascinating study that shows that “at the current rate of churn 50% of the S&P will be replaced in ten years” - that is a staggering statistic, and the reason, mainly, is Machine Learning. But it doesn’t have to be that way, if companies understood what was going on and retrained and revolutionised their business models to understand how ML was affecting their industries, they could survive then thrive again.

Everybody has heard of driverless cars, but here are some other industries about to be changed by Machine Learning (ML) and narrow AI.

Social Virtual Reality - this will revolutionise hospitality and socialising and take it online.

Volumetric cameras - these will completely change the way we shop for clothes, homewares and anything else we surround ourselves with.

Virtual assistants and Virtual Influencers powered by ML Natural language processing (basically the ability to chat to you) - these will change our relationships and how / who we receive life and product advice from. Check out www.virtualinfluencer.agency

Voice - Text searching will all but disappear and those owning the voice channels will dominate their product categories.

I would argue its important for all strategic marketers to get onto and be ahead of this curve, understanding exactly what Narrow AI can and cant do for their brands and how its going to affect their vertical, whether that involves a new social platform, a data scraping or customer insight tool, a content resonance tool, sentiment analysis or a host of other ML services.

There was a an interesting article in The Drum** “IBM predicts AI will create a new breed of marketers”. It is an excellent article and predicts the emergence of the 'martecheter,' a more tech-savvy marketer - hey ive got a new name!

“IBM believes AI and machine learning will make hyper-personalization a reality as the proliferation of data and the streamlining of marketing stacks will allow marketers to deliver personalized content at a massive scale”.

They are right.

More and more my role is to advise whats possible with tech to help clients understand their consumer and be more relevant in culture, helping brands commission bespoke tools and apps to deliver what they need for their marketing strategies and business in general - its not a one size fits all market place any more, brands need to create their own versions of tech, curated to their needs and their own data.

But back to my course, im one of 500 lucky students to take part in the first professional learning course on Machine Learning run by MIT, learning from the worlds best, Devavrat Shah no less, the Director of Statistics and Data Science Center at MIT - he’s a legend. Educational technology is also going to have a transformational and democratising force on our world, it will create many opportunities for those who would not normally be able to access the education through lack of privilege or money.

They even run a free “Machine Learning” course right now anyone can access. Also by way of example I took a course in how to programme Python for £25.00 - online - it was excellent. And there are jobs going that are well paid, for people who have skills in there areas - they don’t even need much experience - the playing field has been levelled, the opportunities are there and tech has facilitated the process - so there is no excuse for marketers not to up-skill in these areas.

But back to my course. It takes us through;

  1. Understanding data

    How to gain insights from data using effective visualisation tools

  2. Making predictions

    How to Identify the right prediction models to predict business outcomes

  3. Making decisions under uncertainty

    How to apply effective decision-making frameworks to make data-driven decisions

  4. Determine causal inference

    Evaluate decisions to determine whether the outcomes resulted from my actions

Pretty relevant to marketing right?

Its interesting that in my course less than 5% are from the UK and less than 5% are in marketing - thats a bit crazy… They should be running the course again so if you are remotely interested then i urge you to look at it. This is the first time in history someone like me, a marketer who has always had a deep love of Tech, creativity and the future, is able to study with someone like Professor Devavrat, and I encourage all my colleagues to have a look at this and other Machine Learning courses, it is only going to grow in importance and it is essential for any marketer to maintain relevance and value.

  • * https://www.innosight.com/insight/creative-destruction/

  • ** https://www.thedrum.com/news/2019/01/02/ibm-predicts-ai-will-create-new-breed-marketers