Artificial intelligence may not be as “smart” as we want yet, but it will be; use algorithms to guide smarter content-driven personalization
For all the promise of big data and artificial intelligence, the current reality is that most smart technologies are just simple, logic-based systems guided, directed, and controlled by humans. In other words, they don’t produce many actionable, significant insights beyond pattern recognition.
The problem, now, is threefold:
► Hype has eclipsed reality. Machines aren’t yet advanced enough to own every aspect of your digital marketing operation — but they can augment/enhance many areas.
► False expectations. Many digital marketers erroneously believe there is some magic data set that will unlock the flood gates and unleash tons of online shoppers leading to massive sales — the clouds will part, and marketing manna will drop from data heaven.
► The promise of marketing — personalized interactions for each customer — is still going unfulfilled. It’s been (and still is) impossible to scale 1:1 communications across your database — that is, until artificial intelligence marketing came to fruition
AI technologies are becoming more and more widespread among marketers today (Source: Building Trust And Confidence: AI Marketing Readiness In Retail And eCommerce, Emarsys and Forrester, 2017)
Is AI Hype — Or Are We Just Marketing Badly?
Most marketing systems/platforms are glorified prediction systems that run on large amounts of existing customer behavioral data. Good stuff, but not actually that predictive now is it?
I’m certainly not an AI expert, but have been actively learning about the space as it relates to marketing for the past year. I’m learning by active experimentation and observation — the same approach I took a decade ago to understand influencer marketing, which allowed me to create the Walmart ElevenMoms program.
The biggest barrier I see for AI marketing is the hype surrounding its application and how it’s being used today.
Related Content: 4 Myths About AI Marketing & How to Combat Each
Most marketing application of the technology is limited to conversion tactics and results in relentless amounts of digital spam filling every part of a “target customer’s” digital path. We need to change that up.
“Most #marketing application of #AI is limited to conversion tactics resulting in too much digital spam” says @Katadhin CLICK TO TWEET
Too much spam, not enough personalization
Just because your predictive algorithm knows I’m in a certain location and matches the profile of other people who’ve bought a certain product, doesn’t mean that because you intelligently place an ad for me to trip over that I’m necessarily going to buy anything.
If this were true, you’d see engagement and conversion rates across the industry going up, not down.
The sad fact is most digital media today relies on an ever-increased amount of reach — not true personalization — to deliver results. The real result: more spam in the form of every digital interruption imaginable… and a poor experience for users.
Facebook now puts ads in the middle of videos, which many people simply skip, along with the rest of the video. I ask marketers all the time how they’d like this in their feeds; and, of course, they don’t — yet continue to foist it upon their customers. This isn’t 1:1 marketing, and certainly leaves the promise of marketing (true personalization) left unfulfilled with consumers just resenting us marketers.
Digital disruption is spam, period. There’s got to be a better way.
Use AI Marketing to Enhance Customer Communications
By using customer data to understand WHO each customer really is, we can start building better targeting, or rather as my business partner Ted Rubin prefers to refer to it as “match-making,” models to connect with customers. However, we need to also pair this approach with better content and place it where people are actually receptive.
Digital media consumption isn’t passive, it’s active. People DO want to consume certain types of media, and they pick and choose when, where, and how. The sweet spot is meeting them there with relevance. But too many marketers ignore this idea, and treat their digital marketing like TV, spewing forth messaging they care about but that their audience doesn’t.
“Digital media consumption isn’t passive, it’s active…. too many marketers ignore this idea & treat their #digital media like TV, spewing fwd messages they care about but that their audience doesn’t” says @Katadhin CLICK TO TWEET
Missing opportunities to leverage AI correctly
Take social media for instance. Any AI-driven social media tool has a simple job: to keep you tuned in. It does this primarily by understanding what content accomplishes that task. Content that doesn’t make the grade gets punished by never showing up in anyone’s feed (unless it’s paid). Even when paid, it automatically increases in costs as people avoid, block, and just reject it as crappy content — especially in this archaic, interruptive ad model.
The most tragic part is that marketing reporting, analytics, and insights have never been stronger. Every piece of digital content you’re using is perfectly measurable. You can literally see whether individual ads are relevant to consumers by looking at what your data is telling you. Unfortunately, though, brand equity is being destroyed like never before.
But, the opportunities to create 1-to-1 brand relationships exists in a way never dreamed of before, thanks to the abundance of data being generated paired with artificial intelligence marketing.
AI will help usher in a marketing renaissance
As mentioned, AI’s most common and useful application today is pattern recognition. Big data certainly helps fuel and improve this use; but marketing teams shouldn’t worry about AI taking their jobs any time soon.
Instead, look to AI as an augmenting technology — one that will, in time, bring about a new era of marketing.
But in the near term, marketing teams will find themselves most commonly using AI to relieve themselves of painstaking, repetitive tasks like segmentation and data mining. Instead returning to tasks they signed up for like creating content that resonates with consumers and helping train the machine to understand how to deliver content that matters.
Use Case: Fake News
Interestingly, this approach is already used commercially as evidenced by the fake news phenomenon.
Quite simply, anchor content in the form of videos, blog posts, and the like, that is created by various parties, syndicated by bots, and “authenticated” by real people — creating a trust layer that sparked the algorithms to share with like-minded people.
This is actually a rudimentary approach to AI execution but it worked rather flawlessly and effectively. Marketers can take note, as this model is relatively easy to replicate.
The simplicity of the current state of AI is that it seeks to simulate dominant human pathways. By combining the customer data you have collected, the content you’re creating for narrowly-defined segments, and the logic of your AI-enabled marketing platform, AI can help you not only achieve hyper-efficiency (with your team and your input/output), but also move you away from a “spray-and-pray” reach model that spams everyone in sight, including customers.
Carefully evaluating the organic path to purchase will produce a variety of predictable AI-driven outcomes from web and e-commerce traffic, to longer-term loyalty and brand preference. ◾
This article originally ran in the whitepaper “AI & Marketing” by Access AI, and has been slightly reworked for the Emarsys blog.
Handpicked Related Content:
John Andrews is a serial entrepreneur and a career marketer in the consumer packaged goods industry. John was a senior manager of emerging media at Walmart, where he created the ElevenMoms program in 2009. He was also the founder and CEO (2009-2013) of Collective Bias, which drove business development with top 50 consumer brands and was named one of Forbes ‘Most Promising Companies’ three years in a row.
John currently works as CEO of Collaborative Content App Photofy and operates The Katadhin Company, a marketing consultancy, and Prevailing Path, a shopper platform designed to identify the most efficient path to purchase across digital channels, which he co-founded with Ted Rubin.