The choice of a data visualisation or analytics tool is often more personal in nature than technical. The question that acquires prominence before purchase is how comfortable the analysts are or would be with them. The steepness of their learning curves pushes other questions backward such as whether enhancing the scopes of the tools involves major costs or if they will cater to future data processing needs.

I’ll share an odd experience to explain why I have decided to move some important and oft-used dashboards to Google Data Studio. This is despite the relative immaturity and incompleteness of the system.

I had a bitter experience while purchasing a few licenses for a popular data visualisation and analysis software for my employer at the time about two-and-a-half-years ago.

The license was expensive — way too expensive. In an age when most so-called “enterprise” tools/platforms are making a beeline to oblivion — having their myths busted, user-unfriendliness ridiculed, and hype-triggered halo doused — those tools still managed to grow their businesses because of growing awareness about analytics.

We preferred an online version for its collaborative work capabilities, which is essential. But high costs made us change our mind and we finally settled for the tool’s desktop version. What came along was a real “enterprise-style” shocker — a hefty annual maintenance fee! Failing to pay would bar us from downloading the updates/upgrades of the software.

For heaven’s sake, why?

That was also about the time Google Data Studio debuted within the Google 360 pack. Many people like me couldn’t wait to lay our hands on it — and we were disappointed instantly.

At first look, it was so not “Google.”

But despite the heartbreak, one could hardly ignore the promise it packed inside its rudimentary features. First, because we all know Google’s ability to add great depth to any product it decides to push ahead with. Second, Google itself owns some vital data sources or platforms that matter to publishers, which made us expect Data Studio would integrate seamlessly with those platforms.

Two years and a slew of updates on, it’s now free with no limitation on the number of dashboards and it has gotten a lot better. It’s ready to breathe down on the balance sheets of the haloed stars of Gartner’s “Magic Quadrant for Business Intelligence and Analytics Platforms.”

Media houses are now focused on report templatisation and group viewing by building shareable dashboards so team members look at the same combinations of metrics and dimensions. Data Studio made the task super easy with a Google Drive-style access control system.

The other major differentiator is the date range selector. Charts on a dashboard change simultaneously as the date range changes, while offering a comparison with the previous period.

Data connectors are a pivotal part of any data analysis tool; they let the dashboards feed on the data streams from platforms or services. The connectors put an end to the need to download data on a spreadsheet to analyse — wasn’t that way too uncool?

The data visualisation market leaders have long lists of connectors to boast. While Data Studio doesn’t offer many beyond Google’s own products, it worked around the limitation by allowing a user to connect to Google Sheets, where data contained in each worksheet of an online spreadsheet becomes a distinct “data source.”

That’s interesting, and it also caters to about 80% of a publisher’s data connectivity needs, considering third-party data connector tools are put to use. And there are a few good ones that actually don’t cost much.

The third-party data connectors allow you to pull data at regular intervals from many services and platforms. The most common ones are Google Analytics, Facebook Insight, Facebook Advertising, Twitter, Instagram, Google AdWords, Google Search Console, Google Doubleclick, Bing, Semush, and many more.

Advanced users can similarly pull and store data in online Big Data warehouses like Google Cloud Platform or Amazon Redshift. Data Studio connects to most major data warehouse platforms, making Google’s outlook very clear: Instead of just letting the product be just a good data visualisation tool, the company wants it to evolve into a popular front-end for Big Data stacks.

That’s where it becomes very interesting. Most media businesses can’t do without online data warehouses now.

They are keen to understand the behaviour of a good article or video piece consumed concurrently on a Web site, apps, AMP, and third-party aggregators like Facebook Instant Articles or UC News. They want to view that for other pieces too, together or separately. They want to apply statistical logic to discover consumption patterns. They want to create projections based on those findings. They want data to show them light.

A baffling limitation of Data Studio is the absence of any option to “join” multiple data sources. You can’t stitch two or more pieces of data through common “fields” like you do in most advanced data analysis tools. The limitation does not allow a user to use more than one piece of data in one chart. And that’s an awfully big limitation.

Regarding this situation, I asked around on forums but didn’t get a satisfactory answer. Among the probable reasons is that a “join” requires higher computational resources and free consumers may have been deemed unworthy of that.

It can also be that Google wants its advanced users to get rid of joins and romance columnar data structures instead. They respond to queries much faster, are horizontally scalable, and have easy convertibility to “NoSQL-column store”-type database structures.

However, we all honed our analytical skills on SQL-supported tools, and joins aren’t going away soon. We would love to see the stitching option on Data Studio. Advanced users can connect Data Studio to Google Big Query (which supports join) for a nice work-around though!

Calculated metrics on Data Studio are smooth; a publication data analyst may use them to her heart’s content. The only logical function, “CASE-WHEN-THEN-ELSE,” can also be used to create a calculated metric, but attempts to blend metrics and dimensions in the same function would throw an error. This is another big limitation. But I guess Google will improve on it soon.

That said, there are no tools as powerful as an innovative analyst's mind. The innovations will be on nitro-boosters if two other very cool Google technologies are brought together: App Script and Query functions.

App Script is modeled on JavaScript and is fairly easy to learn. Among many other amazing capabilities, it enables automatic porting of data between spreadsheets and servers. And Query functions makes it possible to run SQL-type queries on data stored on Google Sheets. Many such small wonders can help you overcome your challenges, even a major one like the absence of joins!

Does the diagram give you a clue?

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .