Saying Tableau is just a tool for making dashboards is like saying Superman is just a guy with a cape. Locked inside of Tableau is the ability to completely transform your entire business decision-making process from end to end. The problem is, extracting that level of value from Tableau requires either a massive army of analysts who never sleep or sophisticated automation of critical but low-level, repetitive tasks. I believe that this level of automation, which is not available natively in Tableau today, needs to focus on three key areas: Detection, Prescription, and Notification. Let’s look at how robotic process automation (RPA) can provide this level of sophisticated automation and unlock Tableau’s superpowers by freeing up analysts to get back to what they do best - analyzing data.
First, what is RPA? RPA is the name for software (sometimes called software robots) that is designed to reduce the burden of repetitive, simple tasks on users. RPA software robots are typically programmed to do basic tasks across applications just as human workers would. In this post, we dive into three key areas we are seeing leading companies deploy RPA with Tableau. They are rapidly discovering that this frees up analysts and delivers insights to business decision makers in a format they can take action on.
Most large enterprises now have dozens -- if not hundreds -- of Tableau dashboards. Relying on humans to monitor and detect key changes 24/7 across all of these dashboards is pretty much a non-starter, and from what we have seen, is not really working well for anyone. One CEO we met with last week told us how he now considers Tableau dashboards to be “analytic spam.” Because he has so many of them, he doesn’t know which ones to trust, so he is forced to ignore them all. And even if he knew which ones to look at, he doesn’t have the time to stare at them waiting to detect what meaningful changes will happen on which dashboard and when.
He solved this by setting a “no-dashboards” policy and demanded that people simply send him insights from the dashboards along with suggested actions to take. While this made him more effective and efficient, it essentially shifted the impossible burden of tracking dozens of dashboards and hundreds of metrics onto the analysts and BI managers. That led to his staff inviting us in to demo RelayiQ’s RPA solution for Tableau. Using RelayiQ, we simply agreed upon the thresholds that would be meaningful across the CEO’s core metrics. We then set up automated detection for these metrics right on top of their existing Tableau dashboards.
But RPA for Tableau does not just benefit the CEO. The BI director in the room asked us if he could use RelayiQ’s RPA to detect issues with data quality in Tableau (we get that question everywhere we go). We responded, “Of course!” Their chief analyst was also in the room. She looked a bit uneasy when we outlined how many tasks in Tableau can be automated with RPA. But pretty quickly, she realized that automating detection of important changes would actually increase dashboard adoption/usage and free her up for more strategic analysis.
The team was psyched, but worried that this level of automated detection could take months to set up. We explained that pretty much all the work had been done during the demo and they would be ready to go live by end of day. Another happy customer!
I should also add that we give a granular level of control on when you detect, and how frequently you detect, based on your use case. You can choose to detect only after the dashboard was refreshed or your ETL tool finished. Or maybe you need dynamic detection based on specific conditions and data trends? We give you full control but also offer automated machine learning anomaly detection so that detection times and frequencies can be requested by a machine.
But RPA for Tableau does not end with detection. It merely starts there.
In the last few years, Gartner has said that somewhere between 60% and 80% of big data projects fail, and we think that might actually be conservative. We believe the problem for most of the projects occurs in the “last mile” of the analytics cycle, with insights failing to reach those who can act on them. We are not alone in this thinking. In an article in Harvard Business Review, they describe the “last mile” in data analytics as “the devilish gap between uncovering a data-driven insight and actually extracting value from it.” Here again, RPA for Tableau dashboards can solve what seems like an insurmountable problem plaguing all enterprise data analytics projects with a simple but powerful piece of automation.
Using RelayiQ, you can create prescribed actions tailored to the detection of specific analytic events and automate their delivery. There’s an important distinction here. We believe human expertise is critical to prescribing which actions to take in a given analytic event. Not only do we not want to replace human expertise, but we see RalyiQ’s RPA solution as a way of actually scaling human coaching to reach teams large and small with custom-tailored, data-driven advice and action plans. However, by preloading prescriptive actions for common scenarios tied to KPIs, you can automate the delivery of action plans tied to key events before they even happen! From what we have seen, this is the fastest way to get companies past the “last mile” problem in analytics.
Let’s make this less abstract by giving a simple example. You have a team of 30 sales professionals globally and each has an annual quota. Rather then stare at dashboards every day to figure out who is crushing it and who is falling behind, you can automate detection of top performers as well as those at risk of failing and automate the delivery of prescriptive notifications to them at the exact right time. Perhaps those ahead of quota receive an automated email with praise, a reward, and a calendar invite with their manager or team to share best practices. Likewise, those at risk of missing their quota will receive an email with a prescription to meet with their manager for additional enablement and coaching before it is too late to intervene.
That is, of course, just one simple example. We are blown away by the variety of ways our customers use prescription to turn data-driven insights into action. We have customers using RelayiQ’s RPA with IOT dashboards to detect and alert on machine failure. We have other customers that use RPA to make sure the data in their invoices is correct in order to expedite reimbursement. Several customers use RelayiQ RPA to monitor their supply chains to intervene when delays occur or orders spike. Pretty much anything that is being measured in a Tableau dashboard that requires detection around key events followed by prescribed actions can be automated with RelayiQ’s RPA solution.
And we have not even talked about the most powerful aspect of RelayiQ’s RPA for Tableau: Automating the delivery of custom-tailored insights and action plans to all the right people at the exact right time, in the channels they use the most.
As wonderful as Tableau is for analysts, we believe it suffers greatly from delivering insights in the wrong form factor for business decision makers. To be clear, every analyst we have met (thousands at this point) looooves making dashboards in Tableau - but for every analyst, there are dozens of business decision makers who don’t want to go on a scavenger hunt through dozens of dashboards to find actionable insights. So what do these business decision makers want? Simple notifications that let them know when a metric tied to their KPIs has made an important change along with prescribed actions on what to do about it. And one of the biggest things we have learned from deploying RelayiQ is that pretty much everyone is a business decision maker on some level with their own set of KPIs. That means that automating the delivery process can get complex fast! The good news is that RelayiQ’s RPA for Tableau is designed from the ground up to handle this level of complexity.
The first thing we did is make sure that we could deliver insights in the communication channels that business decision makers are already using and give them the power to subscribe to the ones that work best for them. You want notifications in email? No problem. But maybe you prefer getting push notifications or having them sent via Slack. We do that, too. Maybe you use some other communication tool in your enterprise? Easy peasy. We can hook you up with a custom connection. We also know that it is not always a human sending and receiving the notification. We make it incredibly easy to notify a machine to take action for IOT use cases and use cases with Freshdesk, Jira and even another RPA solution!
We know people (and machines) not only need control over the way they receive their alerts, but also control over the frequency and the type of alerts that they receive. That is why we built an insights-delivery catalog where you can pick and choose the notifications that are relevant to you. Once subscribed, you can always adjust the thresholds to throttle up or down the frequency of alerts based on your use case or make adjustments to the channels you want to receive insights in.
Ready to unlock your Tableau superpowers? We offer a free trial and free onboarding to make sure you are successful and up and running in no time!