Thursday, November 9, 2017

Big Data is the key to Artificial Intelligence

Have you ever thought about all the data your business is capturing on a hourly, daily or weekly basis? It is probably incomprehensible in light of the channels and volume of information captured 24/7.

The overall, high-level purpose of mining all this structured and unstructured data from your CRM, sales, marketing and advertising channels and most recently IoT devices is to garner insights into your customers, competitors and potential market trends.

It is not humanly possible to categorize and find insights from these oceans of data quickly enough so that the information is relevant.

Big Data and Artificial Intelligence AnalysisWith all that data, the teams of data analysts that companies rely on today to interpret the data simply can’t keep pace with the volume.

The real challenge is merging all the analysis together to get a 360-degree contextual picture of your customers, potential purchases and market trends.

Apple's Steve Jobs once said during an interview,
“I remember reading an article when I was about twelve years old. I think it might have been Scientific American, where they measured the efficiency of locomotion for all these species on planet earth. How many kilocalories did they expend to get from point A to point B? And the condor won, came in at the top of the list, surpassed everything else. And humans came in about a third of the way down the list, which was not such a great showing for the crown of creation. But somebody there had the imagination to test the efficiency of a human riding a bicycle. A human riding a bicycle blew away the condor all the way off the top of the list. And it made a really big impression on me that we humans are tool builders. And that we can fashion tools that amplify these inherent abilities that we have to spectacular magnitudes. And so for me, a computer has always been a bicycle of the mind. Something that takes us far beyond our inherent abilities.”
Artificial Intelligence (AI) is the new bicycle bridging the virtual world with the physical and big data is the fuel and lifeblood of AI.

Big Data and Artificial Intelligence
With recent advancements in computer processing, data storage, and better machine-learning algorithms it is possible to ingest and analyze more data than ever before. At the same time, there is a connectivity boom as more and more devices and apps connect to the Internet producing even more data.

With these advances, is it now possible to feed your big data into an AI engine and let machine learning mine the precious insights, predictions and next course of action. We can teach machines through supervised learning now, instead of programming them and they will then learn on their own through trial and error. That’s why having large amounts of data is more important than ever. The more data AI has, the more accurate it will become.

Data is now more valuable than oil

The Economist says the world's most valuable resource is no longer oil, but data.
"As devices from watches to cars connect to the internet, the volume is increasing: some estimate that a self-driving car will generate 100 gigabytes 
per second. Meanwhile, artificial-intelligence (AI) techniques such as machine learning extract more value from data. Algorithms can predict when a customer is ready to buy, a jet-engine needs servicing or a person is at risk of a disease. Industrial giants such as GE and Siemens now sell themselves as data firms."
The International Data Corporation (IDC) predicts that 44 zettabytes will be generated by 2020 (A zettabyte is 1 trillion gigabytes).
Trends that will shape Big Data and AI in 2017

TechRepublic, a resource for IT decision makers, says there are five major big data trends to watch in 2017.
  1. AI and machine learning will increase the need for for big data analytics
  2. Self-service big data tools even for beginners are hitting the web
  3. Analytics is struggling to keep up even with big data warehouses like Hadoop and Spark
  4. Data cleansing will become a prominent industry as AI is only as effective as the data it ingests.
  5. Democratization of data - server-less, micro architectures will allow data to be accessed, analyzed and managed without servers from anywhere by anyone.
AI is ubiquitous and growing

No matter what you do, AI will eventually touch every aspect of your life. AI, machine learning and deep learning are making big impacts on business and your personal life from simple chatbots to self-driving cars.

Many people use these terms interchangeably, but they are different.

  • AI is defined as the capability of a machine to imitate intelligent human behavior.

Examples are computer chess and most chatbots where the AI is programmed to ONLY play chess or answer a specific subset of questions like customer support issues or a back-to-school sale.

  • Machine Learning (ML) is a subset of AI and designed to analyze large subsets of data and learn from it. ML allows computers to learn without programming to complete a task.

ML understands speech and can make predictions based on the data it analyzes.

  • Deep Learning (DL) is a subset of ML and uses neural networks to learn the characteristics of something like face recognition.

Google's DeepMind AlphGo used Dl to beat 18-time Go world champion Lee Sedol in 2016. AlphaGo studied 30 million human moves in Go and learned by playing against itself.

Google Translations can now teach itself to translate languages it doesn't know using its DL Google Neural Machine Translation (GNMT) system. The new DL improves translation quality, and enables “Zero-Shot Translation” — translation between language pairs never seen explicitly by the system.

Saturday, July 29, 2017

In-store customer location tracking similar to digital website analytics

My wife and I were shopping in a one of our favorite department stores when I noticed a small nondescript sign between the racks that read, "Free WiFi - sign in to get discounts."

(BTW: that's not my wife on the left; that's our dog, Nickie in case you were wondering.)

So I pulled out my iPhone and logged in. A coupon popped up on the screen, "15% off your entire order - today only!"

I was thrilled. My wife was looking at several blouses she liked but she thought the price was too high so I told her to pick one out.

What just happened is a harbinger of in-store interactive, personalized marketing in its infancy. I say infancy because logging into a WiFi network will soon be paramount to taking a horse-drawn wagon cross the country rather than a commercial jet.

What is currently happening in some retaiI stores and what I envision will be in all stores eventually in the not too distant future is the following scenario:

My wife and I walk into a department store. The store's WiFi detects our presence from the store's app or MAC addresses on our phones or smartwatches. We had agreed when we loaded the app to allow detection because the feature would provide discounts. We wander over to the men's jeans department.

A text arrives, "Hello Anthony. Interested in jeans today? Buy two pair and get the third pair half off. I decided not to buy the jeans as good as the deal was at the moment.

We head over to the women's section and my wife is looking at dresses. She picks one off the rack.

"Do you like it?" she asks holding it up to her shoulders.

"Yeah, the color accents your hair. You should get it," I said.

"Maybe, it's a bit expensive."

My wife's phone chimes. She takes it out her pocket and reads the text, "Look at that!"

She shows me the text; it was from the store, "Hello Joann. Looking for a new dress? Take 15% off for being a loyal customer."

She buys two dresses.

At home, I grab my tablet and check ESPN for the latest basketball scores for my favorite teams. A display ad from the store we had just visited pops up with the message, "Just for you. Buy two pair of jeans, get the third on us. Today and tomorrow only."

Spooky, but we had agreed to allow the store to access our devices and to receive messages.

What transpired is just one future scenario in marketing. The WiFi in the store detected when we walked into the store and Near Field Communication (NFC) or Bluetooth technology or iBeacons detected when we were near specific clothing racks.

It was like having a virtual digital sales person standing near us with the power to give discounts to close a sale except the sales person is an algorithm. I purchase all my jeans at this particular store and my purchase history is in the store's database as a frequent buyer of jeans. The algorithm detected my profile and pushed out the text message discount to my phone.

This is tracking of customer behavior similar to website analytics only in the physical realm.

My wife, Joann, has made many more purchases at this store than I have over the years, so her profile is that of a high value, loyal customer hence the 15% discount on the dresses.

All sorts of retailers — including national chains, like Macy's, Nordstrom, American Eagle, Family Dollar, and Cabela’s among others — have been testing these technologies as early as 2013 and using them to optimize store layouts or offer customized coupons, according to The New York Times.

Screenshot of Nomi Technology's In-Store Customer Analytics Dashboard
One company, Nomi Technologies, which provides the technology to track customers in store, recently settled with the Federal Trade Commission (FTC) for allegedly lying about tracking customers in stores, according to a report by the International Business Times.

"Nomi previously defended its use of phone-tracking tech, telling the New York Times in July 2013 that offering retailers the ability to keep track of a shoppers' habits is similar to the way Amazon and other online retailers use cookies to keep track of their customers," reported the website, circa.

According to a video on Nomi's website, their technology can track the number of customers walking into the store, track where they browse and push relevant messages out to their smart phones.

This is tracking of customer behavior similar to digital website analytics only in the physical realm.

While most consumers are Ok with being tracked online with cookies, database profiles of their buying habits and cookie matching used by most e-commerce retailers, there are those who bristle with anger and fear over being tracked physically.

In a March 2014 survey by Chicago-based Opinion Lab and reported by AdWeek, consumers feel this way about in-store tracking:
  • Eight out of 10 consumers don't want to be tracked without giving their explicit consent
  • 64 percent said they should only be tracked if they opt-in or sign up to participate in a program
  • 24 percent believe retailers shouldn't do any in-store tracking at all
  • Promises of a better shopping experience didn't change consumers' minds with 88 percent saying it wouldn’t make any difference
  • Discounts or free products would sway consumers towards acceptance of tracking
  • 81 percent do not trust retailers to keep their data private and secure
The study was based on feedback from 1,042 consumers.

In-store tracking won't go away and what will foster its widespread acceptance are the incentives retailers offer to convince consumers to buy in and reap the fruits of a great discount or free merchandise in exchange for a little less privacy.

 What's your take on in-store tracking? Do you feel it is a violation of your privacy or are you Ok with in-store tracking? Feel free to leave a comment.