What is Data Analysis?
We live in the Data era: everything is data.
Whether you give a “like” to your friend’s Facebook Picture or just ask something to Google, you are giving information about what you like, or what worries you the most.
And that data is available for those who want to consult them.
The Data Analysis is the proper use and interpretation of any data in order to obtain valuable and meaningful conclusions.
In this section, we’ll talk about how properly analyzing and interpreting all the information you’ve got on Internet so it can be useful for you.
- The basic Data sources.
- How and when you should use them.
- The basic statistics tools.
- In order not to obtain wrong conclusions.
- Focusing on simple but powerful concepts.
- How these analyses can be useful for your Business.
- Reading properly the Market you’re targeting.
This is a new discipline (maybe not so new) and whoever starts “mastering” it will be one step ahead of its competitors.
Why worry about Internet data and its analysis?
Because this information is accurate and sincere.
When we ask Google, we don’t lie at all: we just ask about what we want.
Seth Stephens-Davidowitz is a data analyst that published a wonderful book about what he’s found on Google: “Everybody lies“.
By analyzing our Google Queries, he predicted:
- Trump’s election on 2016.
- How accurate Freud was.
- Whether your son is going to be fan of a certain sports team or not.
He could do these, and much other things by analyzing what we search on Internet.
We highly recommend you to read his book. (Of course, we are not sponsored).
All this information represents a “Gold mine” for any Business or academic research.
How? Let’s analyze a simple example.
Should you start a Website or a Youtube Channel?
Imagine that you are hesitating about starting a Website or a Youtube Channel.
Of course, you could start both of them but, since you work so hard you know you only would have time for one option.
What should you do?
If you analyzed Google Trends, you would immediately have an answer:
People is more and more interested about Youtube rather than on a classic Website:
Check how the interest about “starting a Website” is decreasing (blue line) while the interest about “starting a Youtube channel” is increasing more and more.
You could conclude:
The interest about YouTube is increasing and has not yet reached the Popularity websites had 4 years ago (when they reached a maximum interest of 100).
So: there is still a piece of cake for anyone on YouTube.
Of course, for this example, we could have analyzed other different Queries, but we just wanted to share with you this “data” approach and how accessible it is.
* If you are interested about Google Trends, check our amazing Page (inside this section): “How to use Google Trends“. There we give you lots of interesting examples.
When should you develop an Internet Data Analysis?
Whenever you have to “read” the Market, you should think about developing an Internet Data Analysis.
We have insisted several times that only by having an accurate Market interpretation you can make proper decisions.
- Especially on our “Market Research and Big Data” page.
Analyzing the biggest and most sincere Database in the World – Google – should not be something optional but mandatory.
On Internet, you can find:
- Customer unattended preferences.
- Unexpected Trends.
- Different Trends’ correlations.
You have a limitless information source that is waiting for you to use and interpret it.
However, you must be very careful when handling data sources:
Main mistakes when doing Data Analysis
A serious mistake when analyzing data, may lead you to failure more quickly.
Within this section, we’ll talk about what should you avoid doing while analyzing different data sources.
Now, we’ll give you an example about what may happen if trust your numbers too much:
McDonalds Data Analysis mistake
For the 1984 Olympic games, McDonalds designed a groundbreaking Marketing campaign.
For each medal U.S obtained, they would give one food item to they customers for free.
Customers only had to scratch a card that would tell them the Olympic discipline.
Then, depending on the medal obtained, they could obtain:
- A Big Mac if U.S obtained a Gold medal.
- French fries, if U.S obtained Silver medal.
- A Coke, if U.S obtained Bronze medal.
They designed this campaign based on the U.S success in previous Olympic games: They calculated the probabilities and the estimated costs and benefits it would imply.
However… they committed a terrible mistake.
Due to political reasons, lots of Eastern countries boycotted this Olympic games.
And these countries were some of the best ones in several Olympic disciplines.
Without these countries participating in the games, U.S had been able to win much more medals than expected.
And what about McDonalds campaign?
Well… it went terribly wrong regarding its profitability. They lost tons of money.
- On the other hand… it is one of the most famous Marketing campaigns ever, so… maybe it was not that bad idea.
* In the Simpsons, they remembered this with a “Krusty Burger” disastrous campaign. Season 4, Episode 10.
As you can appreciate in this example, having the best numbers, calculations and scenarios doesn’t guarantee a successful outcome.
If you miss something important, you’re doomed.
Most common mistakes when developing Data Analysis
Among all the mistakes you can commit when handling data, the most common are:
- Not taking enough data into account.
- Confusing Correlation with Causation.
- Using averages, when you should use median values.
- Forecasting Markets not having considered enough variables.
- Not testing your Data Analysis.
We’ll explain them all in detail in future publications but we’ll give you a valuable advice that will prevent you from making big mistakes:
- Use your common sense.
If you find a correlation between the price of onions and a certain stock market value, please, don’t invest your life’s savings in that value.
What should you expect from a proper Internet Data Analysis?
Nothing but a better understanding about something that usually makes sense.
Sometimes, you can find an information “gold nugget”: an interesting correlation, an unexpected causation… but never a miraculous way of predicting the stock market.
Now, we’ll give you a good example about a successful Business decision based on Data Analysis.
Walmart Pop Tarts Data Analysis example
* We discovered this example in the Seth Stephens-Davidowitz book “Everybody Lies” we mentioned before.
- Again, we encourage you to read it. It’s amazing.
As you may know, some U.S.A cities suffer from hurricanes.
As soon as the Government warns the population, everybody starts preparing for whatever may happen.
This is something normal.
What is not that “normal” is what Walmart discovered some years ago:
- One of the items people buy the most when they make a “Hurricane preparation” purchase is… Pop Tarts.
It sounds weird, but think twice:
- It is a food item easy to cook: you only need a toaster.
- Everybody likes it.
- You need no water.
- It reminds you good moments.
- Children love Pop Tarts.
- It contains lots of calories.
- It is easy to store.
As we mentioned before: it makes sense.
We don’t know whether people purchasing Pop Tarts think consciously about all these attributes, but consciously or subconsciously, they are making a reasonable choice.
What did Walmart when they realized about this fact?
Simple: as soon as there is a Hurricane warning, they place tons of Pop Tarts where everybody can see them.
This example shows how sometimes you can forecast some of your customers needs successfully by properly analyzing your data.
- Walmart had to analyze what their customers buy, when, the frequency…
- Then they had to find patterns of consumption.
- And since they have hundreds of millions of customers each year, it surely had not been as easy as it seems.
The example also shows that, even if you find a “gold nugget” in your analysis, it surely will make sense.
- Pop tarts make sense once you analyze it.
In this section we’ll try to help you finding your “pop tarts“.
Here, together, we’ll squeeze Internet’s Data until the pips squeak.