broken nameplate necklace
PHOTO: Sarah-Rose

We've all heard the phrase “what a difference a year makes.” It’s true for many things, including the state of technology. Just last year, I wrote an article titled “Brands Still Haven’t Tapped AI’s Full Promise,” where I highlighted the pitfalls and successes brands have had with artificial intelligence (AI). In the 18 months since, significant advancements in AI have been made on multiple fronts. From chatbots to robotic process automation (RPA) for distribution and logistics, AI is progressing at a rapid rate. 

While we hear countless stories of how AI is aiding brands on the back end — from Amazon’s distribution automation to AT&T’s and Walmart’s improved logistical process automation to E&Y’s HR process automation — my question today is, what about AI on the front end? How is AI and the automation it provides benefiting brands on the front end, specifically with personalization initiatives? 

I’d like to approach this through the lens of three distinct areas, all of which directly impact the concept of personalization: 1. how AI-based data improves brand’s ability to scale, 2. how AI-based analytics improves context and 3. how AI-based insight improves customer activation. We’ll conclude by looking at how combining these three areas can result in improvements to voice of customer (VOC) programs and personalization initiatives.

AI-Infused Data Improves Scale

Data is clearly a powerful thing. When introducing AI to data, it can expand traditional data sources, increase volumes and significantly turn up the dial on the detail within the data. Large data volumes (>1 million rows) used to be extremely difficult to account for.  Systems would screech to a halt, analytical jobs would never complete, and frustration abounded. 

With AI, the game is changing. Today, with AI, we can collect more types of data — device data, network data, in-home automation data and of course vehicle data. We can handle greater volumes of data — through the use of automation we can distribute and store it more efficiently. And finally, we can see much more detail in our data — models that used to be impossible to run or could only run against limited data sets are now being executed with ease against hundreds and thousands of variables. As we all know, with more variety, velocity and volume to our data comes better, more accurate personalization results.

Related Article: The Final Steps in Data Mobilization: Analytics and AI

AI-Based Analytics Improves Context

Personalization works when it's based on contextual relevancy. Interacting with context (knowing your customer and their needs) and delivering something relevant as a result (the correct offer or interaction) depends directly on your data sources and the analytics applied to that data. 

AI-based analytics takes many forms. On the machine learning side, models such as neural network or decision tree-based models provide better fit than classical generalized linear models. Better fit translates into more accurate results and improved context. With regards to cognitive computing, using natural language processing or sentiment analysis to extract and then create pertinent structured data from unstructured sources provides numerous advantages for brands. It allows them to append customer data profiles with data which previously was not used for context and insight. It allows them to predict and forecast customer behaviors, and it provides insights into customer sentiment and emotions to react accordingly. Prior to the use of these techniques, call center or social data sat in siloed systems and was rarely used for large-scale personalization efforts. 

Related Article: How GDPR and AI Turned Unified Data Into a Business Imperative

AI-Based Insight Improves Customer Activation

Customer activation is a fancy way of saying “the ability to increase customer engagement.” The more knowledge you have about a customer on the individual level, knowledge related to their behaviors and needs, the more you should be able to influence activation or engagement. Machine learning based techniques like natural language processing and sentiment analysis allow us to extract key customer data, such as key phrases from a voice chat, an instruction to an AI-powered assistant or a discussion on a social platform, and use that data to then derive insight. This insight can then be used to increase customer activation based on the more comprehensive understanding it gives us of our customers and their needs. We aren’t just using standard data that a customer provides upon initial engagement, registration or application. We are augmenting that data with data from other sources, sources we couldn’t collect before these AI technologies were introduced.

Related Article: 5 Drivers of Personalized Experiences: A Walk Through the AI Food Chain

All 3 Together Signal a Seismic Shift in the Works

So, what’s the result of all of this? The result is that a slow but seismic shift is occurring. 

Customer expectations are slowly changing. Just a few short years ago, it was OK to receive a generic offer. It was OK to hear “no, we don’t carry that any longer, and won’t have it in the foreseeable future,” from a customer service rep. It was OK to receive that same credit card or insurance offer in the mail. It was all attributed it to brands “not knowing us.”

That’s no longer the case today. The data is there. The analytics are there. The voice of the customer insights and the level of customer understanding are stronger than they've ever been. The result is consumers like you and I now expect Amazon-levels of personalization. The digital leaders of today have created this expectation that other brands are chasing. The entire customer experience, which hinges on the ability to personalize to a certain level, depending on the industry, will be the leading competitive differentiator that a brand will possess.