I’m no stranger to data analytics. And everything I believed to be true about data analytics was based upon the premise that you absolutely had to reshape data to get it to perform the way you needed it to.
Yet our first attempt at business intelligence (BI) at GC Services—which used Microsoft Power View and followed this traditional data reshaping path—failed miserably. Three months after a tedious, six-month deployment led by an expensive external consultant, we were stuck at zero percent adoption. People wouldn’t use it, so they suffered from the same lack of business insight that prompted the complicated project in the first place.
That’s why, when an Incorta representative insisted he had a new way to analyze data, one that didn’t require Extract Transform Load (ETL) or data reshaping—and had the case studies and performance studies to back it up—my supervisor, data analytics manager, and I were intrigued. Intrigued because bypassing ETL would benefit us in so many ways. But very skeptical the technology could possibly work in the manner Incorta claimed and really work the way we needed it to.
So we agreed to put Incorta to the test.
We decided to evaluate Incorta, Tableau, and Microsoft Power BI against each other in a “vendor bake-off” of sorts via a real-world, proof of concept (POC) environment.
Our POC parameters were very rigorous, giving each vendor only three weeks to deliver a functional analytics solution that:
- didn’t require ETL to be used or data to be reshaped;
- performed well on even the largest table joins;
- could access and combine SQL Server, Microsoft Excel, and text files;
- supported near real-time (every 15 minutes) updates of data;
- quickly aggregated data;
- drilled into data at multiple levels; and
- enabled support for and self service by non-technical power users.
We knew we demanded a lot, but according to everything Incorta had told us, they could do it. And what we found was absolutely eye-opening.
Tableau Didn’t Work without Data Reshaping—and then Didn’t Work with It!
To function at all, Tableau had to reshape data. There was absolutely no way around it. But even with data reshaping, Tableau only worked on very small data sets. As soon as we reached one million rows, Tableau wouldn’t return results—it just spun and spun and spun, and never came back.
Power BI’s POC Was Shut Down after Flunking the No-ETL—and the Relationship—Test
Power BI paired us with the same consultants that had built our earlier, failed BI solution. We didn’t hear back from them for two weeks, and, when they did finally contact us—with only one week left in the three-week POC—they questioned our requirements and told us “you really need to have an ETL.” When they continued to insist ETL was required despite our clear requirements, we shut down their POC altogether.
Incorta Does What It Says It can Do—and Much, Much More
And then there was Incorta. First of all, Incorta’s amazing team worked side-by-side with me throughout the entire POC process to make sure everything went well. And did it ever! Incorta didn’t even need the whole three weeks. Within only three days—and without using any ETL—we had set up Incorta on our Cloud Azure instance; loaded and analyzed more than 20 tables and 450 million rows of data (in about 2-1/2 minutes!) from three complex data sources, including SQL; and put together the exact same, near real-time dashboards our previous solution couldn’t deliver after six months of grueling work.
IT. WAS. AMAZING.
I’ve never been that productive that fast on any other analytics platform. And, until Incorta, I never would’ve believed it was even possible.
So we picked Incorta as our new analytics solution.
The ongoing success of our Incorta implementation proves we made the right choice. With Incorta, our leaders understand and can optimize key metrics—for example, helping managers understand how many calls reach the right person each day, so they can staff call centers appropriately in order to maximize productivity while preventing overstaffing—and quickly identify and fix data quality issues. And any user can easily create the performant reports they need on their own while achieving sub-second response times on even the most complex queries, without needing any help from IT.
A particularly big win for me was the fact that we can update and customize business schemas in Incorta in only minutes. In our world, different clients use different grammar and vocabulary. By giving clients access to the same data source but presenting each client with customized business schema using the specific fields, formulas, and grammar they know, their users are much happier and more productive because they can self-serve their own data needs. They don’t need a translator anymore.
Most importantly, our Incorta-powered analytics success encourages everyone at GC Services to think beyond the concept of basic measures and ask better questions—harder questions—because we know we can answer them now.
I think I’m going to like this new, no-ETL world. Thanks, Incorta!
Want to learn more? Register for the upcoming How GC Services Broke Free from ETL in Only 3 Days to Gain Unprecedented Insight into Call Center Operations webinar and read the full case study.