AI Changing the Day of Inbound Marketers
Artificial intelligence has been around for a while and it is definitely here to stay. Everyday experiences with Apple Siri, Google Assistant, Amazon Alexa are all powered by AI. In 2016, 2.5 trillion pictures were taken. If you want to search for a certain picture you took one night while eating Pizza, AI of Google will help show a filtered result of all photos of you having pizza in them.In short, we have to open accept the fact that AI is here to aid human intelligence and make our lives better. Now let’s focus on marketing - AI influences analytics, social media communications, outbound marketing and even content marketing. A typical inbound marketer’s day begins with Analytics and ends with it with a huge chunk on Content Marketing in the middle with Social Media sprinkled all over.
AI in Analytics: Right now a big chunk of the bell curve for analytics adoption across companies is at the diagnostic stage with descriptive also taking a significant amount. A few early adopters have merged machine learning with analytics and have reached the stage of predictive analytics, but prescriptive analytics is the part where the beauty of AI in analytics comes to full bloom.
AI in Content Marketing: Content marketers create content and make them reach the deserved and desired audience. AI today is improving both these aspects of content marketing.
• Content Creation: Ever since an AI was able to write a complete sci-fi screenplay by itself, the possibilities of content creation through AI have been a dream for marketers and developers. A few news agencies have even tried this with services such as wordsmith. It seems the possibilities of automating content creation are seemingly endless.
• Content Curation: The marketing of content is much to do with personalization and aggregation of content, with the intent and interest of the reader in mind. Though such capabilities already exist, it is limited either to the channels of marketing or to larger platforms such as Facebook, Quora, Medium and alike. Wouldn’t it be great if such abilities were available for your blogs and other big rock content pieces?
AI in Social Media: Scraping data, building buyer personas, enriching CRMs and macro-sentiment analysis has been around for
some time now. Though all of this has some ML and AI aspects tied to them, I think AI in social media has a more dramatic role to play.
Have you watched the famous sci-fi TV series, Person of Interest? In the series, a master computer analyses all the data (visual, voice, written) captured through all the CCTVs, mics and websites; public and personal, to come up with a social security number of person who can be a perpetrator or a victim of a violent crime in the next 24 hours. This is what AI can do for your brand. It can read (written), decipher (video and audio) and analyze every voice about your brand and raise priority flags for you to attend.
AI in Lookalike Audience Modeling: Look a like modeling focuses on aggregating first, second, and third-party data to find and target groups of users by consolidating user profile data. Lookalike modeling allows you to automatically find new target segments based on the significant characteristics that build your existing customer segment. For example, an e-commerce merchant’s learning algorithm finds that segment of recently converted customers who purchased an Iphone has a significant overlap in demo graphical and socio-economical characteristics with individuals who visit the websites such as product review sites such as Engadget.com. The new segment can now be targeted through PPC campaigns on such portals to acquire more customers who are most likely to give you the maximum ROI.
AI for Real-Time Personalization: Majority of personalization done today is influenced by of a set of manually curated rules that look for specific contextual data points such as a user’s geographical co-ordinates, customer status, gender or matrimonial status to send varying messages based on the marketer. Algorithmic personalization aims to use ML to dynamically personalize a website in real time based on a user’s browsing pattern. In a travel app context, this can include offering a dynamic discount on air tickets if a user is likely to exit the website based on pre-defined probability models. This in build on the concept of reinforcement learning and its purpose is to optimize actions against a fixed outcome, such as e.g. the revenue expected from this user. If an expected outcome is reached (e.g. a customer books a ticket) the AI system backtracks the steps and applies a reward. In the opposite case, where a user exits the website, it ignores the personalization actions for future sessions. The approach explores multiple possible actions so it can adapt to contextual changes such as behavior during a long weekend. This is balance of exploitation - using past data to find personalization actions that performed as expected - and exploration - continuously trying new routes and observing their outcomes.
The future of AI in Marketing: Including the already mentioned ways AI will impact marketing, such as content creation and curation, prescriptive analytics and social media management, there are a few more ways I think AI will noticeably impact marketing in the future:
• Handling the monotony
• Bots will come and conquer the conversations
• Algorithms will be commonplace, and data will still be the differentiator
Let us know what you think about the use of AI and the potential impact it can have on how marketers like you and me operate. Join the conversation and may the force (and intelligence) be with you.
Majority of personalization done today is influenced by of a set of manually curated rules that look for specific contextual data points to send varying messages based on the marketer
Have you watched the famous sci-fi TV series, Person of Interest? In the series, a master computer analyses all the data (visual, voice, written) captured through all the CCTVs, mics and websites; public and personal, to come up with a social security number of person who can be a perpetrator or a victim of a violent crime in the next 24 hours. This is what AI can do for your brand. It can read (written), decipher (video and audio) and analyze every voice about your brand and raise priority flags for you to attend.
AI in Lookalike Audience Modeling: Look a like modeling focuses on aggregating first, second, and third-party data to find and target groups of users by consolidating user profile data. Lookalike modeling allows you to automatically find new target segments based on the significant characteristics that build your existing customer segment. For example, an e-commerce merchant’s learning algorithm finds that segment of recently converted customers who purchased an Iphone has a significant overlap in demo graphical and socio-economical characteristics with individuals who visit the websites such as product review sites such as Engadget.com. The new segment can now be targeted through PPC campaigns on such portals to acquire more customers who are most likely to give you the maximum ROI.
AI for Real-Time Personalization: Majority of personalization done today is influenced by of a set of manually curated rules that look for specific contextual data points such as a user’s geographical co-ordinates, customer status, gender or matrimonial status to send varying messages based on the marketer. Algorithmic personalization aims to use ML to dynamically personalize a website in real time based on a user’s browsing pattern. In a travel app context, this can include offering a dynamic discount on air tickets if a user is likely to exit the website based on pre-defined probability models. This in build on the concept of reinforcement learning and its purpose is to optimize actions against a fixed outcome, such as e.g. the revenue expected from this user. If an expected outcome is reached (e.g. a customer books a ticket) the AI system backtracks the steps and applies a reward. In the opposite case, where a user exits the website, it ignores the personalization actions for future sessions. The approach explores multiple possible actions so it can adapt to contextual changes such as behavior during a long weekend. This is balance of exploitation - using past data to find personalization actions that performed as expected - and exploration - continuously trying new routes and observing their outcomes.
The future of AI in Marketing: Including the already mentioned ways AI will impact marketing, such as content creation and curation, prescriptive analytics and social media management, there are a few more ways I think AI will noticeably impact marketing in the future:
• Handling the monotony
• Bots will come and conquer the conversations
• Algorithms will be commonplace, and data will still be the differentiator
Let us know what you think about the use of AI and the potential impact it can have on how marketers like you and me operate. Join the conversation and may the force (and intelligence) be with you.
