woman whispering in a child's ear
PHOTO: Sai De Silva

As an analytics professional, I get asked about key performance indicators (KPI) all the time. But if someone were to ask me what metric is coming on the horizon, I’d have to point to stats regarding word association.

The rise of chatbots, voice assistants and voice-enabled search on smartphones is introducing nuanced results for words that appear in search queries and for responses in the software that serves customers. Being aware of how your company acts on that information can not only drive your advanced analytics but also influence how customers perceive your brand.

Related Article: How Voice Search Speaks to Your Customers' Needs

SEO: The Original Word Association Challenge

The idea is not new. In fact, the marketing world has faced word association challenges for some time. Search engine optimization (SEO) was the original word association challenge, with website developers using it to manage keyword density and examine word placement within HTML page elements. 

The purpose of SEO was to ensure a website would attract the intended audience, and a key element of an effective SEO strategy was to recognize that some words or phrases can be useful descriptors in one kind of association yet have a negative connotation in another. For example, “strawberry recipes” would be a negative keyword phrase for a bakery seeking to sell strawberry muffins. So a marketer would filter a paid search campaign to ensure the bakery’s ad would be shown to people who were looking for something to eat rather than something to bake.

A New Communication Dynamic

Fast forward to 2018. After almost a decade of mainstream social media usage, dynamic word associations have moved beyond the realm of SEO. That’s due to the ever-increasing number of channels and platforms people use to communicate every day. People are communicating more frequently on social media. They are also using more smartphone and tablet features to communicate and are increasingly engaging with chatbots — all while on the go throughout the day. The fact that people can access these channels and platforms wherever they are allows them to share real-time response to real-world events.

This engagement introduces an online communication dynamic in which customer attitudes, environmental conditions, culture and demographics begin to shape the words constantly associated with a brand. Consumer feelings about a product or service are now visible on media that is not optimized on a webpage, making more associations possible.

Related Article: Voice Search Optimization 101: What Marketers Need to Know

A New Word Association Challenge

Analyzing and interpreting all of the new word associations spawned by this new wave of engagement and communication requires an assessment of which words come together regularly. Right now, marketers have different ways to do so. The simplest, most straightforward method is setting a Hootsuite or Tweetdeck column dedicated to a subject hashtag or using a social media management dashboard like Sprout Social or Trackur.

A more complex approach involves advanced statistical models. A statistical model using text as data can place words into a corpus, filter that input to remove unimportant phrases, and then use machine learning technology to sift through and identify clusters of words based on tone, or even look for different words for the same product, like pop and soda.

One aspect that must be addressed in modeling involves social norms and cultural perspective. The fact that people are speaking more with their devices on a variety of topics, beliefs and perspectives influences why certain words are chosen. This requires analysts to understand how certain words are used and then make decisions on what word associations are valuable to an analysis. This approach can also minimize bias, which can introduce unfavorable discriminatory concerns alongside a brand message if left unchecked.

Overall, conducting a text analysis to understand word associations improves marketer understanding of consumers because of the volume of ideas people are expressing. Optimizing how a business responds to customer comments about products, services and the customer experience is the future SEO for brand value.

Related Article: A Test: Which Vendor Wins at Sentiment Analysis?