Self-service analytics tools are all the rage right now, in large part because they're seen as user-friendly and...
relatively simple to install. But for many companies, the real work starts after implementing the software. Self-service applications force business intelligence and IT teams to reexamine how they operate and work with end users to ensure that the tools are used to their fullest.
"It's so much more than just buying the technology," said Michael Bridges, managing director for information management and business intelligence at consulting firm Accenture.
Bridges said getting adoption by workers throughout the organization is always the hardest part of the process for the companies he works with. There are many considerations to take into account when implementing self-service tools, like making sure that data systems are architected properly on the back end and that the software is put to work in appropriate use cases and configured properly for them. But Bridges said that when businesses fall down on self-service deployments, it's often because they haven't properly considered how the technology will fit into internal workflows.
For Josh Lipton, vice president of technology at startup SpareFoot Inc. in Austin, Texas, the key to gaining end-user adoption is to start small and customize. SpareFoot operates an online marketplace designed to help connect customers to self-storage facilities. Lipton said the traditional model for BI software and self-service analytics implementations, particularly in a small organization, is to provide universal access to a centralized tool. But he has taken a more personalized approach.
SpareFoot uses a cloud-based BI platform with self-service features from GoodData. It wasn't pushed out company-wide at once, though. Lipton said he prefers a more Agile methodology that lets him and his team do development in rapid iterations and deliver functional applications to specific users in a short timeframe.
Answering the call for analytics data
Lipton started with the company's call center. He set up the BI system to enable managers there to forecast call volumes, track worker productivity and monitor the outcomes of customer interactions. With all of that data at their disposal, the call center managers can better determine things such as when workers will be needed most, giving them more accurate information and increased flexibility for scheduling breaks and time off, according to Lipton.
The deployment for the call center has also led to a 15% increase in work rate, a measure of how much time customer services reps are on the phone, and a decrease in the amount of time it takes answer to incoming calls. Lipton, though, said he wouldn't necessarily look to implement the BI tool the same way in other departments. "It's very important to achieve a beachhead," he said. "But just because one group wants something doesn't mean everybody wants it."
Josh Lipton, vice president of technology, SpareFoot
Understanding users is also at the heart of making self-service tools relevant to front-line workers. Not everyone is going to take to new ways of doing things as quickly as others. It's up to BI and analytics teams to bring along such workers, regardless of how comfortable they are with using the software initially.
"It's in our interest to make them completely self-reliant," said Manoj Yadav, director of business analytics at online art and poster marketplace Redbubble. "If you don't make them self-reliant, then the analytics team ends up doing chores. We want to be doing more interesting analytic things."
As users like it on self-service support
Providing appropriate support to different types of business users is the key. The San Francisco company has implemented the GoodData tool in multiple departments, including customer service, marketing and supply chain operations. Yadav said in all the departments, interest levels among users have varied. Some workers are really excited about being able to dissect data, while others are more interested in just receiving prebuilt reports.
As part of his effort to support users wherever they may be in the adoption curve, Yadav and his team published a sort of dictionary that describes all the different ways the BI tool can be used. It also explains data attributes, data types and the metrics that can be manipulated in the software.
Adoption was initially slow when Redbubble first implemented the self-service analytics tool in 2012, but it has since started to pick up. That has led to a new problem: The more people see what the tool can do, the more they want to expand its use, which demands ever greater support from Yadav and his team.
"Power users are the ones who help initially," he said. "The problem is they keep coming back with new requests."
Get your data architecture right before moving to self-service analytics
Self-service tools can lessen impact of data scientist shortage
Self-service success is about more than just turning data over to users
Embedded tools for BI ease access to analytics data