Xero customers: Sign up free and get AI-powered insights on your P&L in 60 seconds!
Learn more
Back to Blog
Getting to the "why" - Tableau vs Hyper Anna

Getting to the "why" - Tableau vs Hyper Anna

After a report or dashboard is presented to management, the analyst’s work is not done.

Getting to the details that matter is what’s crucial in any business hence there will always be clarifications to understand the data better.

Manager: “This looks great, but can you explain why this happened, or what caused this to happen?”
Analyst: “... I’ll get back to you.”
Manager: “I’ll need the answers by today as we have a meeting with senior management/stakeholders tomorrow morning”

Does this sound familiar to you? Unless you’re a mind reader, and can anticipate all potential curve balls management throws at you, it would be a common occurance to be bombarded with ‘why’ and having a looming deadline above your head to get the answers quickly.

Data driven decisions are becoming popular as it helps the organisation with continuous growth; having actionable business insights helps drive better business decisions without speculation or gut feel.

Getting to the why is always a tedious job, and having to juggle both their existing workload and ad-hoc queries is the bane of any analyst.

At the start of any analytics journey it’s to explain data by establishing relationships and pointing out anomalies records. Automating this piece is a step forward in the right direction.  

The end goal of that journey is to democratise insights. At Hyper Anna, we believe that anyone regardless of skills or background in data analytics, can be confident in knowing that their business insights and trends of their data are at their fingertips.

Tableau vs Hyper Anna

Tableau’s latest release update 2019.3 boasts a new feature ‘Explain Data’ that helps you understand the “why” behind the unexpected values in your data. It provides potential explanation for unexpected data points with a single click.

In order to demonstrate Tableau’s Explain Data functionality, and how Hyper Anna works, I will be using Tableau’s training data, Sample Super Store. That way, the comparison between the two products is objective and fair. A user of this data might want to understand sales performance, and if sales have declined, why?

1. Tableau's "Explained Data"

Tableau’s ‘Explain Data’ helps the user to identify the specific outlier record that skews the specific data point, and provides an alternative calculation if that outlier is not part of the dataset.

Observe that I had to click onto Dec 2018 data point to identify such occurrences.
By clicking on the ‘Explain Data’ icon, Tableau explained that Dec 2018 sales was that it’s higher than expected and this is due to the number of records.

This does not explain much about sales decreasing, so I decided to click onto Nov 2018 data point to understand more on the peak. Tableau identified an anomaly transaction that might have contributed to the high sales in November. This however still doesn’t exactly explain why sales has decreased in Dec 2018.

Tweaking the chart around, I managed to get another view of ‘Explain Data’ for Dec 2018. It explained that a possible reason for the high amount is due to Category - Office Supplies. Hm, I guess that works too.

2. Hyper Anna's "What caused this?"

Anna proactively provides key drivers for each measure (in this case it is Sales) and users will be able to drill down further on what caused them.

Anna highlights that the factor with the largest impact that might have led to the decrease in sales was machines sales.

By drilling down further into what caused machine sales to decrease, Anna identified that Burlington sales in machines dropped the most.

In conclusion

As technology advances, and with more data being generated by the day, the standard task of businesses understanding their customers and their own teams gets increasingly complicated. Businesses will be expected to keep up and one way in doing so is to invest in an AI strategy with the right tools. There are different ways to implement an AI solution that solves the same problem and it eventually boils down to the company’s ecosystem and desired user experience in this AI journey.

I hope from this comparison on different applications of AI solving the same problem, you will be better equipped to evaluate the product that helps you get closer to the ‘why’ and subsequent ‘why’s.

Share on social media: 

Subscribe now

Subscribe for all the latest in data analytics.
We will never share your email address with third parties.