The Business Case for AI for Marketing

Paul Roetzer, founder of PR 20/20 and creator of Marketing Artificial Intelligence Institute, and MarketMuse Co-founder and Chief Product Officer Jeff Coyle discuss the business case for AI for marketing. After the webinar, Paul participated in an ask-me-anything session in our Slack Community, The Content Strategy Collective (join here). Here are the webinar notes followed by a transcript of the AMA.

The Webinar

What is AI?

The most practical and logical definition of AI comes from Demis Hassabis, CEO of DeepMind who writes that it is “the science of making machines smart.”

Types of AI

AI is an umbrella term for a collection of tools and algorithms that make machines smart. Machine learning makes decision without explicit programming, taking in data and making predictions about future outcomes. Deep learning is a subset of machine learning that uses neural networks (layers of algorithms).

Marketing AI Use Cases

Try AI Score for Marketers where you can explore and rate dozens of AI use cases, and get personalized recommendations for AI-powered vendors.

Top 5 Marketing AI Use Cases

The top five marketing AI use cases are:

1. Analyzing existing online content for gaps and opportunities.
2. Choosing keywords and topic clusters for content optimization.
3. Constructing buyer personas.
4. Creating data-driven content.
5. Discover insights into top-performing content and campaigns.

Questions to Ask When Evaluating Marketing Technology

The five fundamental questions to ask when evaluating marketing technology are:

1. Can you explain how their AI works in simple terms?
2. How is it smarter than what you’re doing?
3. How will it save you time and money?
4. Can an engineer walk you through it?
5. How will it increase the likelihood of success?

How to Build Your Business Use Case

Build your use case by starting with a pilot with the goal of establish whether you save time or money, automated the process, and achieved business goals faster. Start small. Measure and evaluate your pilot. Then scale your investment once satisfied with the initial results.

The AMA

What resources or thought leaders do you regularly read?

I’ll go with some favorite books first:

• Prediction Machines: The Simple Economics of Artificial Intelligence (Ajay Agrawal, Joshua Gans and Avi Goldfarb)
• The Algorithmic Leader (Mike Walsh)
• Human + Machine: Reimagining Work in the Age of AI (Paul R. Daugherty and H. James Wilson)

In terms of whom I follow, here’s my Twitter AI list.

What use cases have you seen get tackled well and advance the most?

The most common across businesses of all sizes/industries usually fall into email, advertising, content and conversational. But we also see lots of activity (i.e. funding and interest) in sales uses.

Are there Personalization solutions that also integrate A/B/MVT — or does it just make A/B/MVT obsolete?

There is still value for A/B/MVT, but I don’t think humans should be the ones figuring out what to test and adjusting the creative based on the data. That’s a very logical place for machines to take the majority of the work. If it’s repetitive and data driven, machines can do it better than humans (as a general rule).

Do you see that Enterprise and SMBs approach AI differently?

Enterprises definitely have more data and resources, so they can be more aggressive in how they approach AI. However, that doesn’t always mean more success. Meanwhile, SMBs need to be more strategic in terms of the resources they commit, so it’s often best to start with those quick-win pilot projects we discussed.

What are your thoughts about areas of content marketing AI that you think are under-used?

All of them! The adoption rate of AI is low across the industry, in large part because the majority of marketers (our AI Score data shows 55%) classify themselves as Beginner level in their understanding of AI. I see content strategy and content creation as two of the primary use cases marketers should be considering though.

 How do you see AI changing content creation?

Due to advances in natural language generation (NLG), I see AI playing a larger role in the actual writing/creation of content. In recent years, NLG in this space has largely been formulaic writing (i.e. give the machine a dataset, teach it how to write via a template, then the machine fills in the blanks at scale to create the content). We’re moving into the next phase of AI-powered content.

Do you have any recommendations on any tools that might be ready for NLG content writing?

I’d suggest having a conversation with Jeff Coyle and his team at MarketMuse. Jeff is teaching a course on The State of Natural Language Generation for our AI Academy, so I know it’s a major focus of his.

How can I determine if a technology is really powered by AI or not?

That’s a very common issue. Unfortunately, there’s a lot of vaporware in the industry. Claims of AI and machine learning that, when you dig a bit, turn out to be less than advertised. I would just suggest that you ask them to explain, in basic terms, what forms of AI are being used (e.g. natural language processing, machine learning, etc) and how that makes the use case you are interested in smarter (i.e. more efficient, more successful). If the salesperson can’t explain it (which is very likely) then ask to speak to an engineer who can.

What are your 2020 predictions in companies adopting AI for their marketing strategy?

The adoption curve is slow. But, I think the current economic climate is going to accelerate interest in and adoption of AI at a higher rate. It honestly comes down to two things:

1. Tech vendors building real AI tech that makes their software better and helps marketers improve their performance. Some big marketing software companies have been very slow to commit to an AI roadmap.
2. Marketers being proactive in seeking an understanding of AI so that they can accelerate adoption in their companies.

Are there any AI content marketing use cases that are your favorite no-brainers?

Headlines. What headline to write. Also, recommendations to improve writing (e.g. Grammarly) that goes beyond basic spellcheck, and truly becomes a virtual editor.

What do you find to be the most valuable approach if you are starting a blog from scratch and how can AI help with this?

Ooh, challenging question! I’m not sure there’s any replacement for finding a niche and creating something of value to a target audience, which are prerequisites for building a blog from scratch. But, AI could certainly be applied to figure out what to write, how to optimize it and potentially even help with the creation of the content. So you could potentially write more, in less time, with greater impact.