Is Data All that Great?

"Is Data All That Great?"

Blog Post Number 7

I know I know... my whole blog I've been talking about how great data is and all of its practical uses. However, I wanted to spin things around today and play devils advocate in order to take a new perspective. Data is numbers but sometimes numbers are wrong, sometimes numbers don't tell a story, or provide reason. So lets take a look into some examples I found of how data can be doing more harm than good perhaps.

Wall Street is famous for its "quants," high paid mathematicians who build complex models to predict market movements and design trading strategies. These are really smart people who are betting millions and millions of dollars. However, sometimes their models fail. The key difference between those models and many of the those being peddled around these days is that Wall Street Traders lose money when their data model go wrong.

Something has gone seriously wrong here in the world of big data. When machines replace human judgment, we should hold them to a high standard. We should know how the data was collected, how conclusions are arrived at and whether they actually improve things. And when numbers lie, we should stop listening to them.

"Another example, imagine we're running a business that hires 100 people a year and we want to build a predictive model that would tell us what colleges we should focus our recruiting efforts on. A seemingly reasonable approach would be to examine where we've recruited people in the past and how they performed. Then we could focus recruiting from the best performing schools." On the surface, that seems to make sense, but if you take a closer look it is sometimes flawed. 

So data… take it or leave it, but its always good to question and never just assume. Data is really useful and I do believe it is a great asset to many companies. However, questioning and learning along with asking why is important when assessing and viewing data rather than just assuming.

https://www.inc.com/greg-satell/is-big-data-doing-more-harm-then-good.html

Drink Up

Blog Post Number 6- Big Data Case Study

“Drink up”

Starbucks and data... lets get to the bottom of the cup on this one and determine who your favorite coffee shop brews up their tricks and where they get their data from. 

Where to place Starbucks locations

This explains why Starbucks stores are clustered near each other in most of the cities across the world. However, this seems contrary to common-sense expectation as it can negatively impact the profitability of these densely located stores. But, Starbucks is a clever market player in determining an optimal store location using data analytics. This leads you to find this coffee shop in some of the most prime and strategic locations all across the world. They are placed in high-traffic, high visibility locations near downtown, suburban retailers, work-spaces and university campuses.

Starbucks is using data for insights

•   Consumer demographics

•   Population density

•   Average income levels

•   Traffic patterns

•   Public transport hubs

Types of businesses in the location under consideration.

2. Deciding Starbucks menu offerings

If you love Starbucks and drink it all the time, you should know your beloved coffee shop has been taking notes of what you purchase. Not only does Atlas help it to bag an optimal store location but even assists in customizing menu offerings. Analyzing consumer data, Starbucks drafts its new line of products to supplement the habits captured from its own stores. Atlas also provided data to determine areas with the highest alcohol consumption. Starbucks smartly picked up its stores to serve alcohol as a part of a special menu called “Starbucks Evenings”. Launched in 2010 in Seattle, it has since been expanded to other stores in different cities.

In another example, this caffeine purveyor used Atlas to predict the arrival of heat waves in the city of Memphis and launched a local Frappucino promo to beat the heat. These data-driven menu enhancements enable Starbucks to reach out a large customer pool.

3. Starbucks loyalty program

Starbucks has one of the most sophisticated loyalty programs in the world making it a marketing success.

The program has over 10 million loyalty program members in the US and around 24,000 stores worldwide. There are a ton of takeaways of data from this loyalty program, a few examples are below.

How Starbucks uses this data?

•   Identifies you by linking ‘What’, ‘Where’ and ‘When’ of the products you buy.

•   Tracks down your product purchasing behavior.

•   Delivers targeted advertising and planned discounts directly to your mobile devices.

Using analytics and Business Intelligence (BI), Starbucks has successfully transformed 300 pages long reports into 11 KPIs for each store across the globe.

Starbucks is clearly a data driven company, and uses data to their advantage in order to learn the most about their customers. With piles of information about Starbucks customers they are never at a loss for knowing what people want. Even the Unicorn Frap a free publicity phenomenon essentially was created because they knew people would like it, it wasn't a guess. Pairing with the taste and color desire, they made it a limited time only drink in order to get people racing to a Starbucks, before they are gone forever. So whether its the Unicorn Frap or whatever Starbucks has up their sleeve next, they are data master mines and are never guessing but rather predicting and launching what they know will be successful. Starbucks will keep people coming into their locations day in and day out due to their high personalization of every experience they give to customers. It is clear that Starbucks brews every cup of coffee with strong data science and analytics beans.

https://www.linkedin.com/pulse/starbucks-roasting-data-brewing-analytics-nigrah-bamb

Talk Data to me

Blog Post Number 5 – Monetization of data analytics

 “Talk Data to Me”

Time is money right? Well.... so is data so lets stop wasting time and get to it. Data monetization is becoming more and more frequent in today’s day and age. To sum it up quick, a company basically sells subscriptions to unique data it has created. Which sounds like a great idea, if a company is sitting on huge amounts of under- utilized data and wants to increase its value, why not sell it? In a digital economy there are many ways to monetize data. A company can improve internal business processes and decisions, wrap information around care products and services and sell information offerings to new and existing markets. All three of these approaches offer ways for a company to distinguish themselves in a market place. It is interesting after learning about data monetization to think of data as a product, I have never looked at it this way and it changes the way I’ve thought about data. To better understand data monetization, I think it is helpful to break it down and apply it to some relatable companies to show how they monetize data to get a better understanding of some of these approaches.

In my post last week, I mentioned how Spotify collects data in order to know there customers as best as they can. Now we can look at these two apps from a monetization approach. The music apps Pandora and Spotify have both found profitable ways to mine available sources of data regarding music and musical tastes. They’ve also created layers on top of their data that allow users to: search for the data they want receive suggestions about similar data that may interest them and use social media tools to share their data with friends.

At the heart of Pandora’s success is a machine algorithm that gets “smarter” over time. Every time a user clicks on a song and says they like it, the algorithm adds points to the metadata associated with that song like its artist, genre, beats per minute. The app then uses that metadata to search through larger repositories of music to find every song or artist that matches, or at least comes close to the one the customer already likes. Spotify, has a different twist on music and how it collects its data. Like Pandora, it offers a huge collection that ultimately has something for everyone. In addition, its social media feed allows users to easily share with friends whatever music tracks they’re currently listening to, favorite playlists and more. I personally think its cool to see a real time feed of what people are listening to, since no other social media app really includes music like that. New Apps like Bumble and possibly Tinder, even allow you to sink your Spotify to it, so you can match with other people who may have the same music interests as you. I think this is a cool way to collect data and combine a data app with a music app. Data is being collected in two forms this way, along with cross marketing.

Both apps tap more traditional revenue streams as well, including selling advertisements that are seen by those consumers who subscribe for free. Another way Spotify is monetizing its data is to allow third-party developers to create apps that can be hosted within the Spotify desktop player, providing such capabilities as synchronized lyrics, music reviews and others. In essence, they’re creating an environment around their data that functions as a marketing platform on which other companies will pay to appear.

Data monetization can be seen as a modern day gold rush. Everyone is scrambling to gather information about people in order to aid in their companies strategic measures. Companies are dying to get their hands on data and its almost like every time you go on a website it seems you have to enter your information just to view the page. Sometimes I find it can get a little out of hand if you ask me. Spotify and Pandora show how data monetization has tremendous potential for development in this field.

http://dataclairvoyance.com/blog/what-you-can-learn-from-a-few-data-monetization-examples/