Contents
Copyright 2021, Christopher S. Penn and Trust Insights, Inc. Third Edition.
All rights reserved.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the publisher, except in the case of brief quotations embodied in critical reviews and certain other noncommercial uses permitted by copyright law.
Ordering Information:
Quantity sales. Special discounts are available on quantity purchases by corporations, associations, and others.
Affiliates. Affiliate and network marketing opportunities are available for this book.
Contact Information:
Christopher Penn at cspenn@trustinsights.ai or on the contact page on www.TrustInsights.ai
Written and Published in the United States of America.
AI For Marketers
An Introduction and Primer - Third Edition
Christopher S. Penn
Preface to the Third Edition
The first edition of this book sprang to life largely due to the insistence of Ann Handley, Chief Content Officer of MarketingProfs. Back in 2017, she asked for a simple guide, a set of blog posts, that would help her understand what all the hubbub about artificial intelligence was. I wrote a series of 9 posts on the topic, and when the series was done, I bundled them up into an eBook, added some more examples, and called that AI for Marketers.
The first edition did its job of providing some of the basics and practical applications of AI for marketing purposes, but by no means was it a proper book. In fact, for those of you who purchased it, you can still see artifacts in the text from its blog lineage.
But it was woefully incomplete. In 2019, I rewrote it from the ground up, and the Second Edition was born. More complete. More accurate. More thorough.
In the two years since, Ive presented the contents of the book on stage hundreds of times, and fielding thousands of questions about the application of AI in marketing. Those questions, along with the breakneck pace of innovation in the field, are what has led to the Third Edition. To be honest, two years between editions is probably too long, but for whatever reason at the end of 2019, I didnt feel like a Third Edition was right for the time.
It is now. So much has changed since 2019, from groundbreaking new models and technologies to shocking and frightening abuses of the technology. We have so much more to learn and explore together, so thank you for being here for this Third Edition.
Table of Contents
Introduction ?
Why Do Marketers Struggle with AI? ?
What is Artificial Intelligence Anyway? ?
Algorithms ?
Types of Machine Learning ?
Interlude: Data ?
Practical Applications of AI ?
The Journey to AI ?
Preparing Your Company for AI ?
Preparing Your Career for AI ?
Interlude: Unwanted Bias in AI ?
Down the Rabbit Hole ?
The Future Is Already Here ?
Learn More Here ?
Why is AI in Marketing Important?
What does every CMO, every CEO, every Board of Director ask of us, ask of marketing?
Better.
Faster.
Cheaper.
These are timeless demands.
Make marketing more effective and impactful.
Make marketing reach the customer and help close the sale sooner.
Make marketing cost less.
How are we doing, as a profession?
Instead of better , arguably marketings quality is worsening. Why? Were faced with more data than ever. In 2020, as a species, IDC and Seagate estimated that humans generated about 50 zettabytes of data.
A zettabyte is an unimaginably large number. Its the same amount of space on a billion laptop computers. Heres an easy way to think about a zettabyte. The average show on Netflix takes up about 1 gigabyte of data for an hours entertainment. If you were to embark on the most ambitious binge-watching ever, and you started in the Eocene era, 55 million years ago, with no breaks, youd just be getting to the end of your first zettabyte now. If you were to attempt to watch 50 zettabytes of data? Youre talking about starting a billion years before the Earth itself existed, 5.7 billion years ago.
Facebook alone, according to Visual Capitalist and Raconteur, generated 4 petabytes of data per day in 2020. A petabyte, as a reminder, is 1,000 terabytes. That fancy new laptop you just bought? A petabyte would be the disk space of 1,000 of those. Can you imagine filling 1,000 laptops a day of data?
Can you imagine trying to analyze the data on those laptops, as a marketer? Whether we like Facebook or not, thats still an insane amount of data that people create every day - and thats just one platform.
Even something like traditional news - which has suffered mightily in the last couple of years - still churns out around 200,000 news stories per day. Thats 139 news stories per second. Imagine trying to be heard, trying to stand out, when faced with a firehose of news that big.
Data volume isnt the only problem we face as marketers. Weve got a pretty serious data quality problem, too. How many of us have this email address and its many variations in our CRM systems?
Id be surprised if you didnt have that in your database. We have a massive data quality problem in marketing, and thats a major reason why marketing isnt better.
How big is this problem?
According to the January 2020 CMO Survey, when 500 CMOs were asked what percentage of time analytics were used in making decisions, respondents answered a shockingly low 37.7% of the time .
Take a moment to think about that. 37.7% is just slightly more than one in three. So, put another way, roughly two-thirds of CMOs do NOT use analytics for decision making.
What are they using instead? Experience? Gut instinct? Wild guessing? Some combination of the three?
Its no wonder marketing has a quality problem, and no wonder why we are greeted with skepticism when we ask for more budget, more resources, more capabilities. We simply cant show that what were doing is working, or that were learning and adapting from our mistakes.
So, betters out. What about faster ? Is our marketing faster? Are we helping our businesses reach audiences faster, engaging prospects faster, closing sales faster?
The speed at which data is flying at us isnt slowing down. Quite the opposite: data is coming at us faster than ever before - and not just us, but our audiences as well. In a single day in 2021, its estimated well generate 320 billion emails, or 3.7 million emails per second.
Per. Second.
Want a visual of that? Imagine one email was a leaf on a tree. The average large oak or maple tree contains about a quarter million leaves. Now, imagine 15 large trees dropping all their leaves on you at once, every second. Thats what 3.7 million emails per second might look like.
Consumers are also spending more, faster online. Its trite to say it, but the last year truly has been unprecedented in living memory. A pandemic that forced everyone indoors for a substantial part of the year also exploded data creation. Broadband analytics firm OpenVault projected monthly household usage of 500 GB per month in North America, an increase of 47% year over year.
Were you prepared? Was your data infrastructure prepared?
Consider how much data is involved in each sale, in each interaction, in each touchpoint with the customer. Now consider how human most marketing departments are. At one company I worked at in the past, there were dozens of 20-somethings manually Googling for news stories of the day, trying to keep up with the tsunami emerging from the firehose. No human team - no dozens, not hundreds, not even thousands of people - can keep up with the scale and speed of marketing data were generating today.
