In the article “What You Can Learn From People’s AI Marketing Successes,” we covered some of the brilliant ways that online grocery shopping websites are using AI to vastly increase sales, brand loyalty, and so forth. It was a pretty in-depth look at all the ways machine learning is used, and how the conflicting demands for further sales, new sales, branding, and paid sponsorship are all melded together to create a seamless promotional experience. It is probably one of the best uses of a sophisticated marketing AI system outside of gambling websites. Yet, if AI is doing such a good job for online grocery websites, how are humans supposed to compete?
Table of Contents
How Is Human Expertise Supposed To Compete With Such Perfect AI Marketing?
There are two answers. The first answer is that the online grocery stores we featured may have AI running at their core, but it is it brilliant application that is the key to their success. In other words, where a bad workman blames his tools, a successful workman shouldn’t blame his tools either. Those billion-dollar companies have the best AI tools at their disposal, but the fact they are using them so well and so perfectly is because of the marketing team behind them. It is the human deployment and human optimization of these tools that made them great.
The second answer is that the online grocery stores use AI so brilliantly that they have become the exception. To put it another way, they are the ones using a Ferrari to win an F1 race, whereas the rest of us are lucky if we can rent a Ferrari for a school reunion. Many marketing teams couldn’t compete with what superstores are doing, but if it were a boxing match, the superstore is the 12 foot 60 stone monster, whereas everybody else is the Pink Panther.
In short, AI is only a tool, and it still requires an expert team of marketing experts in order to exploit it correctly. The companies that use AI the best are so far out of everybody else’s league that they themselves don’t actually matter.
Doesn’t Vertanet Use Artificial Intelligence?
Yes, and this article is not decrying AI, the point of this article is to show that AI on its own is no more useful than a bowl of carrot soup when it comes to marketing. It is how AI is used that defines how effective it is.
Many bungee managers have tried to expand our use of AI, but the fact is that our biggest drivers of success come from our marketing expertise, and not from the tools we use.
We use Artificial intelligence. Our team doesn’t try to beat it, our team uses it as a marketing tool. It makes information gathering and information dissemination more efficient, it doesn’t make our decisions for us.
Computer Chess Masters Will Never Be Beaten By Humans
Humans will never beat the most advanced chess-playing AI because we have simply reached a point in history where machine learning beats humans at chess every time. In fact, we now have supercomputers trying to outwit other supercomputers at chess because humans vs machine is now a lost cause when it comes to chess. This fact worries some people. The fact that a human will never win against a supercomputer at chess is worrisome… until you remember that chess is a structured game where luck doesn’t play a role. Randomness doesn’t play a role either.
Chess is a luck-free random-free, balanced and structured game. These are perfect boundaries for a computer to become perfect. But, there is very little else in the real world that matches Chess’s no-luck, non-random, fair and perfect framework. It puts me in mind of a line from the Movie “The Matrix” (1999) where Neo is able to beat the machines because they are forced to follow the rules where humans are not.
Some Examples Where AI Loses to a Seasoned Marketing Team
The best way to show that “Some” humans beat the most sophisticated marketing AI system is through examples. That is to say that even if you created the perfect thinking AI that was able to take hold of a marketing campaign and run it independently, it would still lose when pitted against a seasoned marketing team.
To push the point home, here are some examples showing how even the most sophisticated marketing AI system is going to fail when pitted against humans.
Example – Even The Finest Propensity Model Needs Human Intervention
Since the first Iron Man movie in 2008, the superhero genre exploded in Hollywood, and since then we have seen superhero movie after superhero movie and by 2021, the world is sick of superheroes, superpowers, and giant sky beams.
Humans know that the tide is turning against superhero movies, but based on all data, machine learning suggests more superhero movies are needed. Both historic and recent superhero movie success makes AI think more is required because machines cannot judge a weary atmosphere, it cannot understand why people have superhero fatigue, and it doesn’t have the capacity to experiment in a way that mines completely new data that is relevant to the future success of movies. Artificial intelligence is limited by its own lack of creativity, as well as its lack of understanding of the human spirit.
Example – Predictive Analysis
There is a great book by Nassim Nicholas Taleb called “The Black Swan,” in which he tries to prove that all predictive models, equations, and prognosticative AI systems are useless when it comes to predicting how the stock market will behave. He posits that we only use such predictive models and programs to make ourselves feel better about our investments and that a randomiser tool could make better profits than all predictive AI programs that work to predict stock market behaviour.
This is another area where humans will beat AI every time, and it extends into marketing. Machine learning is limited to what it learns from the past. It can make very clever leaps in terms of predictions, but there is always a railroad in play (there are always lines the AI is going to follow like a train). The best explanation is from Nassim’s book.
A turkey is the predictive AI program. Mr Turkey sees his food bowl become 3% fuller every week. Based on previous data since the turkey was born, he predicts that the food-increase trend will level out or increase over the coming years, and he is hopeful for his future. Mr Turkey could have never predicted that tomorrow is Thanks Giving.
It sounds like a joke, but it highlights the inherent flaw in AI predictive models and demonstrates how there are circumstances a human could anticipate that a machine could never conceive of.
Example – Predictive Pricing For New Campaigns
I personally think that predictive pricing AI is amazing. But, I still believe that humans are better when it comes to promotional pricing.
By all means, allow AI to control most of your promotional pricing, your regular pricing, your upsells and your extension pricing. However, when it comes to big promotions, or new products, or big marketing pushes, a human hand must play the defining role.
Artificial Intelligence will never understand the desires, wants and needs of a human. Take the tumble drier industry as an example. Outside of Japan, there is a dire need for tumble driers that last a long time and do not break. The current offering around most of the world consists of tumble driers that last around three-to-five years and then break. People are crying out for long-lasting machines.
Now, let’s say that you have built a long-lasting machine. Your AI may tell you to sell for a very high price because you will not see repeat sales for another 20 years. However, a human would suggest selling them at an average, perhaps even cheap, price. This is because a human can reason that while they may not receive repeat sales for years, they can actually dominate the drier market to such an extent that repeat sales wouldn’t be needed. They would smash the competition so thoroughly that their business would be a guaranteed success. A human marketing expert could envision this scenario, but a machine couldn’t.
Example – Classic AI Marketing Fails
In almost every classic marketing fail, there is a moment where the embarrassing mistake wouldn’t have happened if only a human had intervened at one of several junctures. In fact, if humans had been monitoring its AI behaviour a little more closely, the “Fail” moments wouldn’t have occurred.
Such as the content marketing AI that doesn’t realise that families use the same computer, so little Timmy was seeing content about adult nappies. Or, the content marketing AI only used Facebook cookies, so the website targeted the young girl who used the communal computer in the old people’s home.
From the marketing AI that targeted the affiliate click farms in Nigeria (because it was producing the most clicks) to the AI that didn’t differentiate between ending a session and signing out, and so kept sending “Are You Leaving?” notifications to people who tabbed away on their web browser. Again, if humans had simply kept a hand on the steering wheel, or had at least monitored what their AI marketing systems were doing, it would have avoided a lot of embarrassment.
“Some” Humans Will Beat the Most Sophisticated Marketing AI System
Not to toot our own horn, but our Vertanet marketing experts can beat even the most powerful machine learning systems being used by superstores and grocery stores. Perhaps our staff couldn’t handle the weight of traffic that AI funnels, but our marketing team could make far better marketing decisions than even the finest AI system.
Do understand that only “Some” people can beat a sophisticated marketing AI system. It does take years of qualifications, knowledge, experience and the cooperation of many people in a team to create and maintain a successful marketing campaign.
Good marketing professionals use AI to its fullest in order to enhance their campaigns. The point is that high-quality marketers could still lead a very successful campaign without using machine learning.
When AI Helps Perpetuate a Bad Idea
Not only are poor-quality marketers unable to beat AI, but they may also actually hide behind AI marketing. In other words, AI marketing tools can make a poor-quality marketing team moderately successful.
Take the example of the woman who wants to transport 30 breeze blocks 50 miles up the road. She carries each block by hand, and it takes her two months to complete the process. Then, somebody invents the van, and she takes each block and delivers them one by one to the destination 50 miles away. She completes her project in one week. She is still using the same inefficient method of taking one block at a time, but the van makes her appear more successful than she really is. That is what poor quality marketers are able to do with AI, they can perpetuate bad ideas simply because their AI tools are making them efficient enough to create the illusion of success.
Do Not Mistake a Tool for a Marketing Professional
The title of this article, “Human Marketing Expertise Will Always Beat The Most Sophisticated Marketing AI System” suggests a human vs machine mentality. That is how many people see AI in the realm of online and offline marketing, and it is quite understandable. The way that online grocery stores and gambling websites use AI makes it seem like a moving, thinking, feeling, living creature, but the fact is that it isn’t. Artificial intelligence, when used in marketing, is just a tool.
You Will Always Need a Top Quality Marketing Team
When a piece of AI technology does a very good job, such as with online grocery shopping, we can be tempted to believe the success it creates is due to the AI itself. Whereas, in reality, it is the humans who directed the AI who are to blame for its success.
Humans do not need to compete with AI because it is humans who are at the heart of AI’s success. Online grocery stores and gambling sites are using AI to such a perfect degree that they operate in their own online dimension. They are so big and perfect that your marketing efforts, be they AI-powered or not, do not need to compete with them.
The point of AI is to make the marketing process more efficient and hopefully more successful. Artificial intelligence isn’t supposed to be used to make decisions for a company. It is the “Human” marketing team that makes the decisions.
AI only beats humans in a world that is 100% fair, structured, consistent, and where things like luck and randomness do not exist. That is why AI will always beat us at chess, but not at marketing.
AI is still a powerful tool to the point where it can even make a poor-quality marketer appear successful. Still, the power of success is born from human hands, AI tools simply make success a little easier. In the end, you will always need a good human marketing team if you want your marketing campaign to succeed.