Raj Venkatesan and Jim Lecinski. The AI Marketing Canvas. Stanford UP, 2021.

This is not the usual kind of book we review, but after all, English Plus did begin as an online Software business. We wrote computer programs for SAT/ACT review and English grammar. Our programming was current for the nineties when Artificial Intelligence (AI) was in its infancy and my first PC had 640 kilobytes RAM, which was the maximum for commercially available personal computers. Now an inexpensive cellphone will have four megabytes of RAM or over six times the computing power of what I wrote the original Verbal Vanquish on.

Artificial Intelligence, especially the kind used for marketing as in The AI Canvas for Marketing, uses much more RAM and enormous degrees more of storage than the 360 Kb floppy disks I used back then. The critical component for AI is speed. Now computers can do some things so much faster than they used to, that they can compare things and see patterns they are programmed to detect, and especially they can manage huge amounts of data in almost no time

The book starts by giving some well-known examples of AI and its speed; for example, the IBM Watson computer that won at Jeopardy and beat a Go champion. This speed and ability to sort through data has made for a breakthrough in marketing.

The so-called Marketing Canvas consists of five steps or stages for a company to introduce, use, and take advantage of artificial intelligence. In fact, someone who is already familiar with AI programs could probably take a photocopy of the chart on page 93 and make a go of things.

First is the Foundation. What are you trying to market? Whom are you marketing to? How can you use AI to help your business? This often means a change in the way things are done.

Second is Experimentation. One thing this book emphasizes is that you do not have to be successful all the time. Because AI is quick, and ought to be implemented quickly, not every method will work for your business, but that is OK. Stick with what works. You learn.

For people in your company or department who are skeptical the authors say, “Challenge your internal staff to work with these [AI] specialists to apply algorithms to your clean data…Then track your results against the human-led way.” (120)

Third is Expansion. Once your group or business of whatever size learns and gets used to AI, expand it. What other uses are there? How can you make use of other approaches. For example, you may start with cookies or trackers, but find you can get more data from loyalty programs. Also there may be commercial databases that have information that is good for your business or even your nonprofit.

Fourth is Transformation. As AI begins to work, there should actually be an effective personalized approach, even if machines are doing most of the research and working for you. This includes a discussion of machine learning.

Fifth is Monetization. This, of course, requires human ingenuity and a good project or cause, but if it works, your income should increase. Even here, the authors note that success is determined by positive mathematical probabilities, not absolutes.

The authors use some hypothetical examples like a small business called Raj’s Bakery, but mostly they show how real businesses like Starbucks, CarMax, and many others have catered to customers with AI.

Ultimately, this is a how-to book. It shows and tells. There are charts, yes, and also good questions to ask as you enter into each stage. If it lacks anything, there is little on what AI programs are out there to use. It does mention ai.google which itself has a slew of possibilities. It also notes that some successful AI companies have been bought out by large corporations that have used them successfully. For example, McDonalds bought Dynamic Yield. It notes that McDonalds uses phone apps and web sites for data, but it also gets electronic data from its drive-through order screens. Some drive-through businesses note license plates rather than names to track customer preferences.

This tells me how things have changed since we were marketing our software. We were proud to note that we had no cookies or trackers on our website. We respected people’s privacy. We still do. How old fashioned! A few users do voluntarily give us email addresses, but it has been years since we have sent out an email.

Even though I no longer concern myself much with marketing since our English Plus software is largely legacy, The AI Canvas for Marketing was fascinating to read. It also helped me understand how AI can be used in other ways.

We know, especially in the last year, how social media have been censoring things. I stopped using Google as my main search engine years ago when I was looking up something innocuous about former President George W. Bush (I think I was looking for the year he was born), and the first things that popped up were all about conspiracy theories (Bush and the Illuminati) and extreme politics (Bush: War Criminal). That told me more about Google than Bush.

Two years ago a friend of mine told me Kamala Harris was going to be Joe Biden’s running mate before it was announced. How did he know? Candidate Biden had said he would choose a woman of color. My friend Googled Senator Harris, Donna Brazile, and other prominent women of color in the Democrat Party. When he clicked on the top links for Harris, all the articles and web sites said the same thing in the same words. She was being scrubbed. He knew she would be the pick.

This also explains how even the founder of Wikipedia has complained that virtually everything even on his “open” encyclopedia has certain political bias. No, it is not AI for marketing, but AI for propaganda. When Facebook or Twitter tells someone that their posting does not conform to their standard algorithm, it is not a cop-out. The algorithm has likely been programmed to flag things like “liberty” or “truth about Covid” for example. (NB: I personally am not hung up about wearing masks or getting vaccinated, but such programming stifles discussion.)

The AI Marketing Canvas boasts that the canvas helps the consumer. In one sense it does. If the AI knows you always order the same kind of coffee, mocha latte can pop up first on your phone screen. That is convenient. But if it is used to conceal facts or promote half-truths, that is censorship. Jesus said the truth will set you free (John 8:32). If something that is not threatening or pornographic needs to be censored, someone does not want you to know something. As John Milton wrote in his seminal work on freedom of speech and an press:

Let her [Truth] and Falsehood grapple; who ever knew Truth put to the worse, in a free and open encounter? (Areopagitica 1644)