The role of the data analyst not even to mention the data scientist is more important than ever!

In this article we will look at the differences between correlations and causality between data sets. Also I will show you some hilarious correlations between data found by Tyler Vigen.

What is Correlation in Data analytics?

A simplified definition of a data correlation is when two metrics are strongly linked to each other and follow up on each other in a linear line.

A correlation can be positive or negative.

When a correlation is positive there is an increase of value.

If the the correlation is negative it will show a decrease of value:

When the variables show minimum difference and the line that goes through all the dots in the chart has almost no difference we are talking about a very strong correlation. The highest correlation number is 1.0 which means all dots are aligned perfectly.

Let’s use a sample with this. I own a bicycle shop and I want to see if there is a correlation between temperature during the months and my sales amount.

You can already see that there is pretty strong correlation here. In fact the correlation is : 0.95775.

Learn how to calculate your correlation with this correlation calculator.

What is causality in Data analytics?

If you want to figure out what is causing for example a peak in your sales there could be a lot of factors. But it also means that some correlations are not proven to cause affecting others.

Some basic causation rules:

• One thing could cause another
• The other might cause the first
• Or maybe they are connected through something else
• And finally… it could just be a “stupid” chance.

In a lot of cases our mind likes it to find correlations that are actually not there. There are tons of samples on the internet where correlations are 100% clear, but simply don’t make any sense.

Tyler Vigen is famous for his book on Spurious Correlations where he analyzed tons of metrics and found some hilarious “correlations” between them.

When you look at these, you can see that both have strong correlations but just (I hope ) coincidental.

Summary

The main takeaways from this article where we explain the difference between data correlation and causality:

• Dig deep analyzing and finding “real” correlations
• Not all correlations are strong enough to make decisions on, make sure to calculate the correlation coefficient.
• Use common sense in a lot of cases to exclude causality
• Make use of easy to use data visualization software to help you blend data and find quality correlations quickly. Give Datadeck a try

The post Data Analytics: Correlation vs. Causality appeared first on .

Introduction
The following explains the methodology of this transition in the Internet era. By studying the secrets of world-class companies such as Google, Apple, Amazon, and Netflix, the author summarized the functional integration of business model design methods that were unknown to all companies in the Internet era, and proposed seven implementation rules.

The Core Principle
Integration
First, the company needs to integrate the core product groups with functional integration and become a product group with a clear brand goal, so that the brand can be prosperous and form a continuous attention among users.
Use one of the most practical and simple features to get users first and foremost, and think about second-generation products on the basis of having a large user base.

Case: Amazon
Amazon’s flagship “Everyone can find any items that you want to buy online”. When people became accustomed to picking books on Amazon, Amazon introduced the Kindle, so that users can use Kindle to read e-books anytime and anywhere.
Tip: Find a book. — This very useful feature has become a function of Amazon’s product integration point.

Second Principle
The scene is king
The traditional mandatory advertising marketing model is outdated, and it is an “interruption” for users to use the product. After the emergence of social media and mobile devices, people are relying more on advertising and interaction in new interaction scenarios. Each company’s marketing department must be changed from being a disturbing master to a scene master, making full use of information, social networking, payment, and other scenes to invent new ways of communication.

Case: Ford Motor
In the Internet age, it is more convenient to rent a car or taxi. Many of the top 500 auto companies have found that car sales have been declining year by year. Many young people do not buy cars. This makes many car companies very anxious, even out to resist uber such a taxi software. Ford found a way: invested in a car rental company, Zipcar, and provided Zipcar with 1,000 Ford cars for rent. Zipcar’s company data is very beautiful, with 760,000 members at the time. Sales of Ford Motors soared, and Zipcarfe was quickly acquired by another company. Ford Motor Company received a huge return.

Third Principle
Catch users with synergy
Make your products work together, interdependent, and mutually supportive. Buffett has a well-known concept called the economic moat. Any company with such a moat will have a pricing power on the brand, and competitors will never be able to break it.

Case: Apple Inc.
Apple was the first company to make a personal computer, but at the end of the last century, Apple’s market share has dropped to 3%. After Jobs returned, he created a closed-loop product ecosystem. In 2001, Apple introduced a series of free applications such as iTunes and iDVD. In order to match iTunes’s music features, the iPod was introduced. Because the iPod uses the iTunes program, you can download music from your computer and transfer it to the Pod. Users develop the habit of sharing ecosystem. When Apple introduced the iPhone, the user bought the mobile phone for convenience. As long as there is an Apple ID, the data in the PC, iPad, and iPhone can be communicated to each other in the iCloud.

Fourth Principle
Re-conceive value creation
Establish a disruptive mindset and find new product directions by continually disrupting and redefining products. Most entrepreneurs or corporate executives are often experts in the industry, and they are self-conscious about market conditions and user habits. This has become a problem. They say that they are transitioning, but they cannot get out of their own experience and do not dare to do anything. The author gives three ways of thinking to open up subversive thinking: from the product’s perspective, from the brand’s point of view, from the perspective of commercial value.
Tip: Rethink the product as a service. — Do not think from the perspective of increasing sales, but instead guide the transformation path from the perspective of making the user more convenient.

Case: Tesco’s service-oriented APP
The world’s second largest retail company, Tesco, positioned the company’s APP as a service. For example, many elderly people buy something very heavy and it is inconvenient to take home. They can then scan the bar code order with Tesco’s APP in the supermarket and the goods will be sent to his home. If someone likes to visit the supermarket, or wants to see real goods, you can pick the product you want to buy on Tesco’s website. When the person arrives at the supermarket, APP will guide him to the location of the product. This saves a lot of inconvenience.
Tip: Re-imagine brands from the perspective of added value.
How to Advance Insurance & Compromise with Customers
Progressive insurance companies have installed a hardware called Snapshot under the dashboard of each driver. As long as the users keep updating data, they promise to reduce the premium after a certain period of time. Advance insurance not only obtains the data sent by the users anytime, anywhere, but also reminds customers with poor driving habits at any time, reducing the chance of accidents. Users also get cheaper at no cost.
Tip: Reimagine the future of the company. — In many cases, entrepreneurs are not reluctant to make a transition because of not want to, but they have not opened their minds. When Steve Jobs conceived of the mac book, he said that if the computer can be made as thin as a book, it would be fine. Tesco’s CEO is also very foresighted. His vision is to transform Tesco into a company that orders online and collect in store, and soon realized Tesco’s functional transformation. To change the thinking, cross-border to think about the future of the company, often the mind is opened.

Fifth Principle
Redesign value transmission
Every company should have a core brand goal. When this goal is established, all strategic designs and product designs cannot deviate from this track. It must be reflected through product design and user experience. It is better not to increase the function. The simpler the better. Single product features and brand goals are more conducive to spread, users are also the easiest to understand the nature of the product. Moreover, if it is a small startup company, it is often not easy for an opponent to discover your threat to him. Many subversive innovations often do not appear in the field of competition but cross-border areas.

Case: Apple Inc.
Apple is a company that makes personal computers. Mobile phone companies such as Nokia and Motorola all regard Apple as an IBM competitor. Until Apple suddenly launched the iPhone, they just woke up, but it’s too late.

Sixth Principle
Reposition value acquisition
How do you create a product’s functional ecology? The approach is data-oriented. The decision of adopting new technologies and developing new products is driven by data.

Case: Nike+
1. Do subtraction first and concentrate on acquiring data. The first generation of Nike+ only had iPods and websites, and Nike focused all its attention on the user’s real-time updates of their running data. After the success of the Nike+iPod, Nike removed the iTunes playlist on the site because it was an ineffective tool for Nike. During the Beijing Olympics, because many Chinese did not use the iPod at the time, Nike developed the second generation of the product before the Olympics. Nike+Sportband, as a new entrance, completely turned off the music.

2. After the core users are basically stable, add additions, expand new user groups, and make new product iterations. For users who couldn’t afford the iPhone, Nike introduced the Nike+Fuelband, a cheap bracelet. Moreover, office workers who do not participate in sports and fitness can also use the bracelet to record data.

Seventh Principle
Think functional integration as the most important thing
Functional integration is a long and slow road. At the turn of the clock, Nike took 7 years to develop the first generation, and forward insurance also took nearly 10 years to develop Snapshot.
For so long, many companies have fallen into various pits. For example, if you are doing it, you will sell user data, or become a platform for paid-for-sharing data, and do not concentrate on iterating of their product.
There are also companies that dismantle the functionally-coupled product divisions because of conflicts of interest between the company’s internal departments. Some companies have too much ambition to move toward another extreme. They want to transform the company, so shutdown most departments, and concentrate on getting new product out. Because the trial-and-error cycle is too long, it causes serious losses and even went bankrupt.
Business leaders should create a supporting department and a budget department for the functional integration department to ensure that the development of the ecosystem is always at the most important strategic position.

Golden Rules
Almost all companies will initially create a website on branding and service-related information when they are involved in digitalization. Today, the information scene on the Internet is more reflected in online search. This has enabled Google to jump from an obscure brand to the world’s most valuable media brand.

1. Google: every day, “to process billions of search messages” we sell targeted advertising in ways that are better suited to the information scene than to bother users.

2. Rethinking a product as a service requires you to fully consider the user experience. In terms of completeness, functional integration is to take “integrity as its root”.
Now, users expect to get clarity and consistency from the brands they contact. Whenever they get a good number from a brand When experiencing, they will increase their expectations of all other brands interacting with each other.

3. Just like the sun and water, technology and data are interdependent and they jointly promote the growth of the ecosystem. The integration of the digital world and the real world can only be achieved through proper management of the functional integration ecosystem.

4. The Internet and function integration brings other equally profound changes and choices to today’s enterprises. People tend to ignore the threat to their status quo in the short term and get stuck on their core competitiveness, even these core products are gradually replaced by new ones in the market. As more and more companies begin to benefit from functional integration, it becomes increasingly difficult to distinguish which industry these companies are actually engaged in.

For more good articles, please visit testblog.datadeck.com
Datadeck provides a dashboard and reporting solution that includes charts, graphs, scorecards, maps, and ad-hoc reports. Everything is driven from the dashboards.
Datadeck is not a metric design tool per se, but it is a powerful tool for visualizing your measures. It also allows others (stakeholders) to access your metrics on their own.
Datadeck works well with Excel. You can easily import data into Datadeck from Excel or other common tools. Datadeck’s power is in the ability to quickly and easily try different graphical representations of your data.

The post 7 Principles for Digital Transformation appeared first on .

When you’re an HR manager named Larry, you manually build HR reports in Excel each week.

And when you manually build HR reports in Excel each week, you waste hours of your life. When you waste hours of your life, you feel guilty. When you feel guilty, you visit a hypnotist. When you visit a hypnotist, you wake up at a voodoo shroom party. When you wake up at a voodoo shroom party, everyone thinks they’re a peacock. When everyone thinks they’re a peacock, the feds kick down the door.

And when the feds kick down the door, you end up on the run with a man who calls himself the Mad Hatter.

The moral of the story? Don’t end up on the run with a man who calls himself the Mad Hatter. Stop losing hours of your life each week building and re-building HR reports. Check out how to speed up the process in this article.

Larry’s Excel Problem

First, let’s look more into Larry’s dilemma by observing his reporting process. Every time a new candidate came up to interview for Larry’s company, the hiring manager overseeing that position would input a new row for that candidate into a shared spreadsheet and overtime, input additional data regarding the progress of hiring for that candidate.

Pertinent Information for each candidate would be:

• The name of the position and the department
• The responsible hiring manager
• The dates of the initial contact, the first interview and the 2nd interview
• The results of each round, whether or not an offer was given and whether or not it was accepted.

Example Spreadsheet (columns A — H)

Example Spreadsheet Continued (columns I — P)

Every Monday, Larry would then take information from that spreadsheet and create another spreadsheet report that summarized the weekly progress of hiring for each position so his team could detect any slowdown or issues in the hiring process.

For each position, the report would include how many candidates were contacted, how many were given a 1st round interview, how many were given 2nd round interviews, how many were given offers and how many accepted offers. Also attached to each position would be other information such as who the hiring manager was, when the position was first opened and what department it belonged to.

Example summary report

Now Larry’s company was a fast-growing startup and they were aiming to fill at least another 30 positions within a few months, many of them developers. Depending on the number of candidates, a report like this could take between half an hour to over a few hours to make each time.

This was because to get the stats for each open position in his report, he would have to find the row for every candidate that interviewed for the position and add up the number of initial contacts, first time interviews, 2nd time interviews, given offers and accepted offers.

In addition，he would have to tack on the other information for that position as listed above. The manual nature of the report meant that errors often occurred and had to be fixed later. As if that wasn’t bad enough, Larry would have to go through this all over again the next week.

The Solution, Datadeck.

What Larry needed was a tool that could cut through the manual finding and adding and even better, update the reports automatically without him having to do it again. Datadeck could make this possible. After uploading the original shared spreadsheet (the hiring managers were updating) to Datadeck, he could literally build a table that showed the pertinent stats for each job position in under a minute.

All of the column headers in the spreadsheet were available as selectable data points in Datadeck’s chart builder upon uploading the spreadsheet. All he had to do to create a row for each position on a table was drag the job position header into the category section of the chart builder.

He could then easily drag the column headers: 1st contact, 1st interview, 2nd interview, offers given and offers accepted into the metric section of the chart builder and miraculously, the respective stats for each column would show up for each job position.

The same is true for the other pieces of information for each position.

Pretty much the entire manual process could be shortened to less than 10 clicks. Even better, he would never have to do this again. He could either upload the updated spreadsheet file the following week or even better, have his team edit a Google Sheet that was connected to Datadeck. This would automatically populate his report with the latest data and make it ready for browsing the next week.

Are you ready to reduce your HR reporting time and focus more on getting high quality hires and improving the number of pizzas consumed per team member in the coming year? Then start using Datadeck to automate your HR reports and don’t end up on the run with a man who calls himself the Mad Hatter.

The post How Excel Reports Turn HR Managers Into Fugitives appeared first on .

However, we can give you the next best thing, one that we prefer. How about the quick and dirty 1–2–3’s to help your team use data quickly and efficiently?

I mean don’t you want to make it easy for every Jim, Sarah and Nazeem in your company to find insights in a way that’s fun and groovy, instead of always relying on the evil data sorcerer that sits in the nether reaches of the office? Behold, Datadeck’s quick and dirty intro to using data, made for dum dums by dum dums!

Let’s first get some misconceptions out of the way with some affirmatives.

1.Data is freaking sexy

(Check out the curves on that.. parabola..)

They say that data analysis is the sexiest job of the 21st century and 64% of data scientists agree with it wholeheartedly. I mean, just check out our other post about using data to get more hot dates!

Even more, Datadeck is out to give data a hip new makeover. In fact our secret desire is to replace “Friday night netflix and chill” with “Saturday night bar charts in my PJs” ? ?.

So get the image of a data nerd with his pokemon cards out of your system. Because those living the real data-driven lifestyle prefer Pokemon Go!

2.Youz already knowz Big data!

Nothing to learn here. Big data was just a way the Kardashians of the data industry wanted to sound fashionable and jingly. All it really means is, you guessed it, big, data.. The large corporations out there harped on the term, attaching it to every shiny server, cloud or software they wanted to sell. You now know as much about Big Data as you ever could, give yourself a pat on the back!

3. Data Scientist isn’t the hottest data job. Yours is!

That’s right, it might have been previously that data was only accessible to the initiated but with new and easier tools, it’s becoming easy for everyone from the CEO to the beer pong guy to look at data and gain insights. The idea that you need to beg some guy in a black box like you’re in a confessional to get your analysis or reports is over. EVERYONE on your team should be data wizards and have a blast doing it!

Now Some Real Tips For the Dum Dums

So are you feeling sexy and stoked to start killing it with data on your team? Then I guess you’re ready to take the short cut! Let’s jump to the essentials that will let your whole team use data effectively without having to suffer all the downtime and pain of your forefathers.

1. Choose goals and metrics

This first part isn’t really a shortcut but an obvious prerequisite. Choose the goals your team wants to improve. Is it more sales, leads or customer satisfaction? Once you’ve done that, select the metrics that if improved, would directly improve the performance of your goals.

For example, if you’re trying to improve lead generation, important metrics might be things like conversion rates on your website or click through rates for your Ads. You’re basically going to want to track them alongside your goals.

2. Report on them meaningfully

Next, decide which angles you’re interested in looking at your metrics and goals from. For example, do you want to look at conversion rates by different geographies, products or devices? This helps you see if perhaps conversions are weak for a certain device, so you can focus on improving the way you engage people on that device.

But if you’re using analytics tools and looking at the standard reports that come with them, chances are, the data isn’t shown in the exact way you want. For example your tool may only give you click through rate data for each campaign, but maybe you want to see it instead for each keyword.

Or maybe you want to compare the same data side by side over different periods. Luckily tools like Datadeck can connect with the original data you’re looking at and let you pick and choose how you want to view them.

3. Make reports easy to understand

Nobody wants to look at a bunch of numbers all day. If you put all the metrics you chose above into a spreadsheet, both you and your team members are going to quickly give up with glazed eyes from row and column overload. Instead, show the data using eye popping visual charts that help tell a story with your data.

These will make it easier for your eyes to notice special circumstances in your data and act on them. You can easily visualize your data by connecting them to Datadeck and building charts from them.

4. Get rid of the manual work

Visualizing your data in a nice report that shows everything you want is nice. But having to do it over and over again is going to kill the team. If you try building your reports in an Excel file, not only is it going to take a lot of time on the first attempt, you’ll have to repeat the process again the next day or week when new data comes in. Instead, use a tool that not only lets you build reports quickly, but continues to pull new data into your reports whenever there’s new data. You can easily do so by automating your reports with Datadeck.

5. Gain freedom from IT

Ever worked at those companies where whenever you wanted to get a report you’d have to go up to IT, explain what stats you’re interested in and then wait 3 hours for them to build it? Or perhaps you worked somewhere where all of the data people wanted to use was spread all over the place. Some of it was in different analytics tools, while others were owned by other departments or people, making it a pain in the neck for whoever wanted to build reports from those sources.

Like we mentioned previously, Data is now meant for EVERYONE. With new centralization and self-service tools like Datadeck, anybody in the company can bypass all of this nonsense and literally in moments, self build reports from whatever piece of data they wanted.

So go for it and start pimping on your data!

That wasn’t so bad, was it? From one dum dum to another, believe it or not, these tips are really all you need to hit the ground running. Not only that, you can literally do all of the above with just one tool. Datadeck! Get your team on it right now. Connect your different data sources to Datadeck. Then bam! Anybody in the company can build reports that visualize the data however they want, in minutes! Do it now, it’s free!

The post Intro To Data Analysis For Dum Dums appeared first on .