August 24, 2015 | Written by: Trevor Sanford
Categorized: Customer Experience
Share this post:
Whether you’re a marketing veteran or a fresh graduate straight out of university one thing you will undoubtedly be tasked with is writing great copy / content. Its a task all of us need to be skilled in, but if I was to ask for a show of hands now of how many marketers thought of themselves as copywriters, I suspect only a few hands would go up.
Why is writing good copy and content so important? I’m sure I don’t need to answer this, as we all know right? Content marketing, Calls to Action, engagement and conversion all should feature in your answer.
Marketers are finding more and more need to generate copy, be it for blogs, websites, social media, Ads or the old favourite the email. I can attest its not always easy to write great compelling copy, but then to write so that its compelling for your target audience can be even harder. Are you sending the same email copy to your 50+ audience as your are to your under 30s? Are your blog posts targeting young mothers written in the same style as those aimed at sports fans? Probably not (I hope!). But here is one of the problems that many marketers face, identifying how to successfully write for a given target audience.
Sure we have plenty of metrics about the interactions with our copy, opens, time spent on page, buttons clicked etc. we can even score for readability using something like the Flesch-Kincaid Grade. What would be super-cool is to be able to match words and sentences to our audience individually and be able to score success, for example Trevor responds better to “Get great deals just for you” whereas we can appeal better to Stephen with “Stephen, come find your personal deals”
Where is this all leading I hear you ask… Well, I recently encountered a new showcase tool launched by my colleagues at IBM Watson™ on the Bluemix platform. The tool is called The Tone Analyzer and what it does got me thinking – More on that shortly! First let me introduce what it does, Tone Analyzer uses linguistic analysis to detect emotional tones, social propensities, and writing styles in written communication. Then it offers suggestions to help the writer improve their intended language tones. I hope you can start to see idea developing here.
My first thought while having a quick play on the tool was how can I use this to create great copy or content that really speaks to my audience. Lets have a closer look at what the IBM Watson™ Tone Analyzer service actually does is it helps individuals understand, rework, and revise the linguistic tones of their writing. The service uses linguistic analysis to detect and interpret emotional, social, and writing cues located within the text and offers rhetorical suggestions to improve the intended tones. The emotional tone focuses on positive or negative feelings disclosed by the writing to derive a particular disposition. The social tone focuses on aspects of the writer’s personality to determine the author’s social nature and the writing tone focuses on the language in the writing itself to show reasoning.
Above are the results of a sample text I entered, at a basic level we could now use this to create copy for Marketing which is high on Cheerfulness and Agreeableness. Maybe for Sales we could add Confidences as well, and Customer services communications would want to score high in Openness and Agreeableness. Taking it a step further individual products and services could each have their own requirements.
What if we also start to analyze all the inbound communications from our customers or use social data like Twitter to start creating scores for each individual. Now we can really start to produce super personalised communications based on personality that will talk directly to the customer using their own emotionally compatible language.
As the natural language skills of Watson keep improving could we see marketing automation reach new levels where by we not only automate the given channel responses but also automate the actual copy within the responses tailored to the segment of 1.