“Technology Married with Liberal Arts” : Modern Day Marketing with Data & AI
The topic on Artificial intelligence has got immense attraction and probably is the most hyped up technological revolution in this Era. The entire industry is now talking about data science , data engineering and basically everything revolving around data & predictions. “data is the new oil !” , even though this has become such a cliche term ,I personally think it’s not that cliche and it definitely is credible.
You would see many technical jargon’s such as Machine learning, cognitive computing, neural networks, deep learning, natural language processing, vision recognition, speech recognition etc. But what does all this mean ? does it seem all confusing ?
Let’s take a few steps back and see how this bubble all started. The research on Ai was not something that started a decade ago, in fact it began in the 1950’s. Researches continued to develop mathematical algorithms for statistical calculations and then for machine learning purposes. But most of these were unsuccessful and not in a mature state to be implemented in real world applications. During the late 1970’s computer scientist such as geoffrey hinton researched on how the human brain works and tried to simulate it using computing algorithms & deep learning techniques. Hinton and the other scientist along with him got most of the concepts and models correct but still it was not at a mature state to be used in real world business applications. But things started to kickoff by 2006 where companies like google , Microsoft & IBM started to see the light at the end of the tunnel. And a few decades later we see the growth of Ai and how it has been used in almost all of consumer & enterprise business applications in modern day. What was the reason for this exponential growth ? It was mainly because of 2 key factors.
1: Amount of Data
2: Compute power
“data becoming the new oil” doesn’t seem much of a cliche now isn’t it ? And to add on to this compute power has grown significantly as well. From CPU to GPU & CUDA architectures to FPGA’s
To sum things up just have a look at this short video which gives a really good overview on How Ai kicked off to what it is now in the modern day
How Modern day marketing has evolved.
As we all know communication channels & mediums of marketing have changed from time to time. From television screens, radio , emails, sms , social media and now social media influence marketing. The reach and scale of these channels are enormous and this has created opportunity for brands to attract new customers but it has also created a lot distortion in the market. Each and every brand is now battling each other to take the spotlight by telling their unique stories to attract customers. But this is not easy since there are too many options with very little differentiation in the market. This becomes an extremely tough challenge for brands to grow their business.
The social media influencer market has grown tremendously during the past couple of years.
Challenges faced by Marketers
My personal opinion is that the challenges have always been the same for brands & marketers, but what has changed is that customers consume information in various digital forms, hence brands need to embrace technology into their marketing strategies. According to my personal opinion some of the core challenges are.
- · Creating differentiation & uniqueness
- · Turning customers into loyal fans
- · Tracking actual ROI spent on Digital Marketing
- · Turning brands to culture.
To address these challenges brands can
- · Make relevant content
- · Tell stories that would make your heart melt
- · Make it personal & emotional
- · Make customer service & marketing one strategy
So how do we do this when everyone else is also doing it ? How do we do it differently to find our own roots and uniqueness ?
DATA DATA DATA !
Social media is a perfect place for building insightful Machine learning (ML) models because of the continues flow of data. To build a good ML model you need lots of data.
Examples such as Classification ML models to classify customers by identifying personal interests & Regression can be used for predictions. Reinforced learning can be used to predict personalized content and offers according the consumers behaviors & digital foot print. These are some of the most common and popular approaches that many researches have been talking about.
As shown in the above diagram when we are looking at creating a machine learning model for predictions we cannot only look at the data on digital channels such as social media , websites we also need to map them to a correlation with internal data which the businesses already has about customers such as CRM, ERP, Banking application etc.
There are also deep learning models (Watson personality insights https://www.ibm.com/watson/services/personality-insights/) that could classify the personality and consumer behavior of customer as well. Marketers can use these insights for building more personalized 1:1 segmentation for campaigns. Most of the time we see brands not really looking at personalized marketing but more of a bulk campaign approach example : sms & email campaigns. With this approach brands cannot really track the ROI or success of the campaigns. Ai models can help to target the right customers and the right time for more successful penetration in the market.
Example use cases
How shazam uses consumer data to predict the next big hit
How googles new Ai & AR powered map’s app could create a new medium for brands
Google has recently launched its beta mode on its all new Ai+AR powered google maps. I think that this would create a new space for brands to pop up offers and content in google maps AR. The brands can have their own mascots walking around the streets and attracting customer. This creates a new world for brands to communicate with their customers, consumers can engage with these campaigns and data can be collected for creating unforeseen insights as well.
Brand also need to create interactive story telling. The traditional one way communication doesn’t attract the new generation. They want things that would interact with gamification & gaming.
Getting the right social media influencer to your brand
There are many social media influencers out there, but selecting the proper personality to your brand can be challenging. Using Machine learning & data techniques could help brands to identify & predict the best influencer. Data & analysis such as
- · Audience demographics
- · Analyzing the historical content posted by the influencer
- · Analyzing the sentiments of the comments by the audience
- · Analyzing the type of personality the influencer represents
- Analyzing the areas of expertise such as food reviews, travel, adventure, geeks , can be used for Creating models that could predict the right social media influencers for the right campaign at the right time.
How brands can use Ai to communicate with differently abled people
Have brands ever thought of using Ai to communicate their stories to disable people ? we need to understand that these are our customer too.
Using technologies such as speech to text, text to speech & vision recognition could create exciting engagements with these customers.