In 2017, marketers will be using artificial intelligence and machine learning as the norm, not the exception. While not every marketer will be making direct connections to AIs such as Google Deep Mind or IBM Watson, more and more of our toolsets will have connections into at least one machine learning or artificial intelligence system.
Right now, artificial intelligence is essentially a toddler. Commercially available, affordable AI is only a few years old, and yet what it can do has already transformed marketing.
Consider the two primary category of machine learning as part of artificial intelligence: supervised and unsupervised learning. Imagine a toddler learning how to classify things, like different colors of blocks. In supervised learning, we tell our child what the color red is, and they learn to find all the blocks which are red. In unsupervised learning, we let our child play and sort blocks on their own; they learn to understand and group the red blocks eventually as a part of overall classification.
That’s where artificial intelligence is today; yet even these toddler-like behaviors are already providing enormous benefits to marketers. Look at your Instagram feed. How many photos a day do we collectively post to Instagram? In 2016, Excelacom estimated we post 38,194 photos per minute, or 3.3 billion photos a day. Could you or I sort through that many photos? Never. For every one photo we looked at per second, 635 more photos would go unseen. Machines are capable of viewing that many photos – and far more – per second, and thus can classify and learn from our photos in ways we simply cannot. If you wanted to know what photo types or subjects to post to Instagram for maximum engagement, only machine learning tools could give us that insight in a timely manner.
A toddler grows at a reasonably linear pace. The 3 year old is more capable than the 2 year old, but is likely not twice as capable or 10 times as capable. The 7 year old is not exponentially more capable than the 6 year old. Contrast that with the growth of machine learning. When IBM Watson won Jeopardy 6 years ago, it was capable of a one hundred calculation computation per second. Last year at World of Watson, IBM announced that Watson is capable of a one million calculation computation per second. Watson’s capabilities grow at a factor of 10x per year. Yet Watson is, for all intents and purposes, still a toddler, a precocious child, in developmental terms. Imagine what it will be like when it turns 10?
Imagine if your marketing were 10x faster, cheaper, or higher quality this year than it was last year. Would your company find value in that? It’s hard to imagine a company that wouldn’t.
In 2017, artificial intelligence and machine learning will continue this 10x trend. The 10x marketer is largely a myth, but the 10x machine is very much a reality.
AI Eats Everything – Including Marketing
So, what should we do as marketers to prepare for AI eating everything? We must dive in now, while the barriers to entry are low. Jump in. Experiment. Learn to develop against the APIs of the major machine learning companies. Right now, we have four basic choices for building AI-driven marketing tools:
All four companies offer strong capabilities, unique offerings, and low costs – especially for marketing purposes. If you’re not diving in headfirst, rest assured at least one of your competitors probably is. Build something. Automate the low-hanging fruit.
Marketing’s future is automated. Be sure you’re the one in charge of the automation by starting now.
7 Marketing Trends in 2017 Series:
- 7 Marketing Trends in 2017, Part 1: Introduction
- 7 Marketing Trends in 2017, Part 2: The Screenless Revolution
- 7 Marketing Trends in 2017, Part 3: Peak Social
- 7 Marketing Trends in 2017, Part 4: The Digital Attention Gap
- 7 Marketing Trends in 2017, Part 5: The Velvet Rope Revolution
- 7 Marketing Trends in 2017, Part 6: AI Eats Everything
- 7 Marketing Trends in 2017, Part 7: The Death of Old SEO
- 7 Marketing Trends in 2017, Part 8: Account-Based Marketing for Everyone
You might also enjoy:
- The Power of Analogy in Marketing Communications
- You Ask, I Answer: Best Language for Marketing Data Science, R or Python?
- Transforming People, Process, and Technology
- Unsolicited "Embargoed" Press Releases Are Absurd
- How to Think About High Bounce Rates in Google Analytics
Want to read more like this from Christopher Penn? Get updates here:
Get your copy of AI For Marketers