Ever since “data gathering” became a common term, the problem of how to gather increasing data gave way to another, harder problem.
As the quantity of data grew, interpretation of the data became progressively difficult. Suddenly, data analysts were faced with such an overwhelming amount of data on anything conceivable that it was difficult, bordering on impossible, to isolate the meaningful bits and pieces.
The more sources you need to consider in your research, the slower and less efficient this process becomes. There are many tools researchers use to overcome this issue, ranging from rudimentary solutions, such as using Google Sheets, to using custom-made software tools to import data from various sources and compare them.
Data reporting in Google Analytics
Google Analytics is one of the sources of data that gathers a large amount of data. All this data can be immensely useful to CRO researcher. Provided, of course, you can wrap your head around this multitude of information. So far, this problem in Google Analytics has been solved by using Dashboards or Google Sheets. Both of these tools allow for easy sharing and updating of reports to different users.
One of the most frequently used solutions for data reporting is Dashboards, integrated in Google Analytics interface. This solution allows users to quickly and easily add the reports they need fast access to, and share them effortlessly. Dashboards are entirely customizable and even feature automated email delivery on specified intervals.
Dashboards suffer from two limitations, though.
- You can only put 12 different reports into one dashboard. If you need more than that, you need to create another dashboard.
- Also, there is no option to allow the user to have any control over the data the reports contain.
Google Sheets, on the other hand, allows for unlimited numbers of reports, easy visualization with built-in charts, and real-time updates of information.But using Google Sheets requires spending time and effort to create and format the reports.
And once again, the user of the report has little control over the content of the report.
Why is this point important?
The purpose of organizing data into reports is to enable users to, well, use those reports.
The idea is to allow your team (however technically proficient they may be), to easily understand and interpret the data so they can use it to produce results. If you’re analyzing an e-commerce website for the purpose of increasing sales, for example, you need to be able to show your salespeople several sets of data (like performance with different consumer segments, products, campaigns, acquisition sources, etc.)
The trouble is that you can’t always anticipate what information will spark a breakthrough (or what missing piece of information could inhibit one). For example, you may send them report on acquisition by source and medium, but some crucial piece of information is contained in data for acquisition by referral.
While dashboards allow for customizing which reports to include, if you need something different, you’ll waste valuable time trying to get it (communicating with the report creator, trying to explain exactly what piece of missing data you’re looking for, etc.)
There has been one alternative to these solutions: Business intelligence software, such as Tableau, for example. Tableau (and other similar tools) is a full fledged business intelligence tool that allows for integrating multiple data sources – not only Google Analytics, but a multitude of web and server-based sources.
But then there’s Google Data Studio – which was just announced in May 2016, and even more recently made unlimited reports available for free.
So how does Data Studio compare?
Why Google Data Studio?
At a glance, Google Data Studio seems to address all of the shortcomings of Google Analytics Dashboards. While Dashboards will remain go to solution for quick and easy sharing of most important data, Data Studio adds versatility to your reports and makes them useful.
Google Data Studio gives you many options for how to share your data: Through email, or giving users direct access to Google Studio (along with “view” and “edit” privileges). This makes sharing the reports with users easier than ever. That way, users can find what they need on their own, rather than having to ask the GA ‘gatekeeper’ for anything the dashboard report leaves out.
Users can format reports easily with a built-in editor, and segment and filter their own data, without any need to contact the report creator.
Data Studio also allows for easy data importing from Google Analytics, other Google services, and data sources maintained by Google. And this data integration is a very important feature. So, before we go into actual using Data Studio, let’s get familiar with some basic concepts of data integration.
What is data integration?
Data integration means collecting the data from multiple available sources and bringing them all together in one place. This makes it much faster to compare different datasets and uncover correlations and interactions between different sets of data.
ETL – extract, transform, load
Extract, Transform, Load (ETL), is the process of accessing a data source and pulling out the data we want – extracting and transforming the data from the source form into the format we need, and finally, loading the data into our own database structure.
The extraction part is, technically, the most demanding. There are so many different formats and types of databases, many of which are maintained with software from rival companies that don’t ‘play well’ together. For these reasons, data extraction tends to cost a surprising amount of money to do.
Transforming data means taking the extracted data, and making it applicable to your purposes. For example, if your data is designated as 1 in its native database, but you need it to be designated as A in yours, you’ll need to “transform” it.
Loading the data is the process of adding the new data to your existing database and making it available for analysis and visualization. It can be done by replacing existing data or merging it. Complex system will even track the history of the changes.
Why CROs need data integration
Before we go into testing the actual piece of software, let’s just briefly elaborate on the importance of data integration for CRO.
In the digital world, everything is available in some form of data.
Any meaningful analysis requires access to this data, in the fastest and easiest ways possible.
When you start the process to optimize for conversion, you set off to gather as much data as you can so you can analyze that data to recognize patterns of visitor behavior. Your success depends not only on the availability of data (a problem that is relatively easy to solve), but on your ability to draw correct conclusions from it.
The better this data is organized and visualized, the easier this higher-level task becomes.
For example, you may be optimizing a large e-commerce store. But an e-commerce store is never just an e-commerce store. It’s probably also a mobile site, or possibly even an app.
In addition, to be able to correctly estimate the available data and put it into perspective, a CRO researcher may want to compare the website performance against another set of data, maybe even the total population of the country or target markets, number of internet connections, data on competitors and other benchmark data that can serve as predictors of future (or potential) performance.
If all this information is typically gathered by multiple tools, and to access and analyze it, you need to spend valuable time integrating and comparing the data. That is time much more profitably spent in actual analysis.
So far, the realm of data integration and aggregation has been dominated by expensive tools – tools that often were overkill for small and medium companies. There are, of course, open source solutions that can be used for free. But they suffer from their own share of problems.
Recently, Google released Google Analytics 360, a bundle consisting of already existing products and a few new ones. Google bills it as an integrated solution for all your data analysis needs. And from the looks of it, they certainly do seem to cover a lot of ground.
Data Studio is an integral part of the bundle, which is currently in Beta.
Hands on with Google Data Studio
Anyone with access to a Google Account can access Google Data Studio and start working almost immediately. As usual, there is a short registration procedure.
The main screen of Data studio consists of the menu on the left hand side where you can create new reports or select a data source from which to import data and a list of reports created so far.
Connecting Data Studio to data sources
Before we start reporting, we need to create a data source connection that enables us to draw data for visualization.
Adding sources is done through a connector – connectors represent the interface between the data source and Data Studio. You just need to open the ‘Data Sources’ tab on the left hand menu (the second option on the menu), select the data source of your choice and follow the instructions to authorize access to that source.
Google Data Studio currently supports the above data sources. Since it’s only in Beta, we can expect more will be added eventually. However, even with this limited selection, Google Data Studio can provide a great way to visualize different data.
To begin, we added the Google Analytics data source. Data Studio immediately showed all the dimensions and metrics available from that source in a table, called schema.
Using this, you can add anything to your report. The “Field” column defines the character (dimension in Google Analytics terms) of the measurement. It denotes what the metric is actually measuring.
The “Type” column states the unit of measurement, or what type of metric is used to measure. It can be a number for numerical metrics, or text for descriptive ones, and even geographical latitude and longitude for locations.
The “Aggregation” column defines how the metric will be aggregated. If you use a metric that is reported as number, than you can select it to be reported as an absolute number, sum, average or automatic. Some metrics, like longitude and latitude, cannot be aggregated.
Confused? Just select an auto setting – it will try to recognize the best method to use.
Once you have defined the schema you want to use, it is time to do actual work.
Creating a report
When you click on “create a new report,” you get an empty space that you can fill with tables, charts, pie charts and even geographical maps. You can create your own templates and Data Studio will guide you through the entire process.
The interface makes aligning the charts and images very easy with helpful lines that provide a visual reference for alignment.
Here, we connected Data Studio with the Analytics account of the Google Merchandise Store. Once you add different visualizations, your final report could look like this:
You can also add filters and controls, so users can adjust the charts to view only the data that interests them, instead of all of the aggregated data.
For example, you can add a chart depicting the total number of sessions on a website. If you add a country filter, you can include or exclude sessions from individual countries you are interested in. This is how it looks:
The filters enable you to do the same for any dimension you want. You can filter the data by device, users, or any other dimension from Google Analytics.
To make your reports truly outstanding, you should follow some guidelines.
Google’s Analytics Advocate, Daniel Waisberg, created an incredibly useful step-by-step guide, that includes 5 best practices for creating reports:
Filter controls give power to the users
Headers and page dividers are great for organization and consistency
Chart diversity makes the report more engagingColor styling helps guiding the eyes
The Report purpose informs the design
Following these guidelines makes your report more lively and easier to interpret.
So far data integration with Google Analytics has been relegated to using either custom made solutions, open source software, or improvising. One of the most frequently used methods was Dashboards, a feature integrated in Google Analytics. Another option, for more complex reporting was using the Google Sheet Analytics plugin, that allowed bringing in the data from analytics directly into sheets. Visualization and integration was done by creating charts and other suitable visual content manually.
These methods have their limits, and the setup process is very time consuming.
But, with Google Data Studio, you’ve got the same results in mere minutes.
That said, Google Data Studio isn’t perfect.
The primary issue is that its data source integration is limited, at least directly. You will not be able to integrate the data coming from most social networks, for example. In fact, the connectors that are available right now are mostly Google oriented with only two non-proprietary sources available – MySQL and PostgreSQL
You can always use Google Sheets to import data from other data sources manually or automatically and then import these data to Data Studio. This is a roundabout way, however. As Data Studio is still in the Beta phase of development, I would expect that this capability will be added later on.
Whether Data Studio will eventually graduate into a full fledged solution (which will most likely not be free) with a limited free version, similar to Google Analytics, and be accepted as another contender in the marketplace of business intelligence tools, such as Tableau (the most widely used Business intelligence system we’ve encountered), Microsoft Power BI, BOARD and others remains to be seen.