Cognitive Computing & Magical Marketing

By Linh PMP

“2.5 quintillion bytes of data is created every day

90% data of the world today has been created in the last two years alone

Every minutes, 1.7 megabytes of data is created for every person on the planet – all 7.7 billion of us

80% of data today are dark and will increase to 93% in 2020”

These numbers are from a presentation by Dr. John Kelly, IBM. And he asked “What is the price of not knowing?” Well, would be huge… That’s why human came up with cognitive computing.

Cognitive computing is the simulation of human thought processes in a computerized model. Cognitive computing involves self-learning systems that use data mining, pattern recognition and natural language processing to mimic the way the human brain works.” (

In other words, cognitive computing results in a human-like thinking process with much bigger capacity and higher speed. It, thus, opens a new magical world for marketing activities, both in terms of market research and customer experience, when marketers have an artificial “assistant” who systematically learns more from quintillions bytes of data everyday, investigates into the 80% data in the dark, infers the meaning and initiates actions. Here comes our future…

Screenshot 2016-04-27 09.22.07

“I’ve hired a master in analytics… I mean, a machine…”

A human problem

In a focus group by Pizzie Pizza:

Q: “Do you like this new pizza flavor?”

A1: “Yes” (Trying to be polite)

A2: “Yes” (Don’t care much. Just join the focus group for filling my stomach)

A3: “No” (Actually like the pizza but hate the moderator)

It’s hard to tell what we really feel sometimes. It’s even harder to tell what others really feel without being misled by their pretension or our own bias.

Magic in the future

Pizzie Pizza market researchers will work with a machine master in analytics, namely Brainie, which can help “read the mind” of their customers by scientifically analyzing these customers’ facial & body expression, the words they used, etc. when they try the new pizza flavor, compared to when they enjoy their favorite food, for example.

Why believe?

According to, natural language analysis recently empowered Kia’s use of IBM Watson to choose the social media influencers to work with for their Super Bowl advertisement. Besides, another company also applied Emotient’s artificial intelligent software to measure the facial expressions of the target audience when told they could take one specific fragrances (such software then accurately predicted the preferred fragrance, while the focus group and survey failed to do so). Machines like Brainie will be eventually developed upon such advancement in a more holistic way.

Screenshot 2016-04-27 09.22.20

“She’s sooo understanding… I mean, the machine…”

A human problem

Samantha goes on a wedding destination website, filtering the options by date, duration and budget, then studying more on each option she finds. When narrowing down her choice to 3 options, she decides to call the website contact for more information.

Magic in the future

Samantha will go on a wedding destination website, namely Wedding Talkie, typing on their help box:

“I want a fun wedding this July”

“You and your fiance loves beaches?”, automatically replied from the website, by the machine Talkie


“Sandie, Starfishie and Coconie are the top recommendations for you. You like a fun wedding. What type of fun activities do you like?”

“Something related to the local culture that we can make the wedding really special”

“Nice! Coconie, Sandie and Villagie have really fun culture-related events for you to explore. Please look at the right side in the screen to check them out”

Why believe?

Digital Genius, an artificial intelligence startup, has been developing an automated platform of customer service based on deep learning algorithms and historical service transcripts. One of their projects is the SMS-based AT recipe assistant for Unilever’s Knorr brand. In the future, such artificial intelligence combined with natural language processes will further automate real-time conversation with more speed, accuracy and personalization like Wedding Talkie.

Screenshot 2016-04-27 09.22.33

“Don’t talk to me! Talk to my cool machine!”

A human problem

Van, a friend of my who plans to go to New York next week, just asked: “Linh, what’s the weather like in New York”. “Generally warm, cool sometimes. Bring your spring dresses but don’t forget some thin coats and jeans”. And the conversation about outfits is exactly her utmost purpose of asking about the weather. That’s why she prefers asking me rather than going on google.

Magic in the future

Van will have a cool smart machine assistant next to her, namely SmarterLinh :-), for example.

SmarterLinh, what’s the weather like in New York”.

“Generally warm, cool sometimes. Bring your spring dresses but don’t forget some thin coats. Do you want me to mix and match the outfits in your closet or recommend you a store to buy a new thin coach?”


“Cool! I’ve found a trendy coat at $89 in Digie store. Only 5 of your size left. Creamy color may match with the black jumpsuit you bought last week. 2 days delivery. Do you want me to buy online now or do you want more options?”

How convenient it is! The smart machine learns from both the mass data on weather, outfit match and online stores and the personal data of such particular owner (her size, her normal expense range, her purchase history, etc), and most importantly, makes logical inferences to deliver reasonable customized options.

And the key thing is that, in the case, the stores now sell by attracting the attention of the machine, the gateway that can orient or even make the decision for Van. Thus it will be super important for them to deliver the marketing messages, not only in Van’s language, but also her machine’s language of algorithm. Such a science!

Why believe?

Well now we can kinda talk to Siri, Alexa, Cortana and so on, which is pretty amazing. The “SmarterLinh” above is just a further step to make such talk become more like a conversation of semantic inference and personal recommendation rather than just frequent Q & A.

What’s up to marketers?

For short, keep up with the trend, be comfortable with big data – cognitive science and ready to challenge the current process with the upcoming technological magic.

“I’m just a little bit caught in the middle…

Just enjoy the show…”


P.s. Pizzie, Brainie, Sandie, Starfishie, Coconie, Villagie, Wedding Talkie, Digie and SmarterLinh are all fictional! But just for now ;).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s