Data analysts who sit in each business function (i.e., sales, marketing, finance) are critical to the success of a self-service analytics strategy. They are the ones who find and shape data sets, conduct ad hoc analyses, and create reports and dashboards on behalf of their departmental colleagues. Since they live in the business, they are uniquely suited to marry business data with business needs. With proper training, data analysts can develop most of the solutions the business needs, offloading these tasks from an overloaded corporate IT or shared services group.

The problem is that most data analysts don’t receive the training and support they need to be proficient with self-service data and analytics tools. The classic data analyst is an MBA graduate and spreadsheet jockey with some knowledge of statistics. Most can pick up and learn self-service tools, but few can become proficient without sufficient training and support. And there is a lot to learn: besides creating reports and dashboards, they need to know how to correlate variables and visualize the output; they need to know how to find, clean, and combine data sets; they need to know how to create and run statistical models; and ideally, they know how to create complex SQL and navigate both relational and NoSQL databases.

Isolated. Most data analysts are hired by business unit or functional heads. As such, many become isolated in their departments, and may not know that there are other data analysts in the organization, maybe dozens, who are doing the same type of work in other domains. Isolated analysts are wholly dependent on their own skills and knowledge to generate insights for the business; they don’t gain the benefits of collaborating with others. Often, these analysts recreate data sets and analyses that already exist in the company. I’ve met some who didn’t even know that a corporate data warehouse existed or how to access it, even though it had most of the information they needed to do their work.

Power User Networks

Regular Meetings. The easiest way to improve the skills and satisfaction of most data analysts is simple: bring them together. Companies with high-performing data analysts have a formal power user network with scheduled meetups and online forums where analysts can share ideas, present problems, and review success stories. Most power users are hungry for information and eager to share tips and tricks about how to use tools, navigate databases, and solve data problems. This type of peer-based learning has social benefits that go far beyond knowledge gained: data analysts gain confidence knowing others to whom they can turn if they have questions or problems. This increases job satisfaction and reduces churn.

Many business intelligence (BI) teams foster power user networks. They assign someone on the team to facilitate a power user network and help individual analysts learn the nuances of self-service tools and upgrade their analytical skills. At one global manufacturing company, a BI staffer ran a monthly webcast on various topics attended by hundreds of data analysts. Other BI teams hold “lunch and learns” where power users take turns sharing success stories or teaching other analysts the finer points of a BI tool, SQL, or R. Although BI technologists can deliver this type of training, data analysts learn best from each other.

Gamification and Gurus. Once there is an established data analyst community, the possibilities are endless. Many BI teams encourage gamification—visualization bakeoffs, hackathons, and other friendly contests where data analysts can exercise and demonstrate their skills individually or in teams. In some companies, data analysts who demonstrate a high level of mastery gain “guru” status in exchange for devoting time to building the power user network and working individually with analysts. Some BI teams encourage these and other technical specialists to hold “office hours” so others, especially, newbie analysts, can schedule 1:1 help. Other BI teams have created a lab environment where data analysts can test their data or analytics solution in a controlled setting with the support of the BI team.

Summits. Some global companies convene analysts for a few days each year to share best practices and hammer out policies, processes, and requirements that improve the data analysis workflow throughout the company. Others hold internal summits geared to the business where department heads and their top data analysts can share success stories and increase awareness of the power of data analytics to achieve business goals.

Certification. With a power user network in place, companies can start to implement a training and certification program. Rather than giving licenses to self-service tools to anyone who asks, some companies are starting to require data analysts to demonstrate their knowledge and skills prior to receiving a license. These certification tests require knowledge of corporate data, self-service tools, and dashboard formatting styles. Some companies have multiple levels of tests to determine the level of access an analyst receives. For instance, the most skilled data analysts (i.e. data scientists) might get access to raw data in a sandbox or staging area, while traditional data analysts can access a data catalog and data warehouse.

Tools. A power user network can also help standardize the use of tools throughout a company. Once data analysts begin sharing knowledge, the best tools begin to bubble to the surface. The power users themselves often push for a corporate standard as a way to accelerate their access and productivity.

To facilitate collaboration among data analysts, many organizations are implementing data catalogs. These tools enable data analysts to find relevant data sets and collaborate with peers across the organization. By reading comments and ratings of data sets and reports left by others, data analysts can help each other understand the nuances of data sets and their strengths and limitations for addressing various types of analytical problems. The catalogs also help them build a strong web of relationships among other data analysts in the organization. This is especially helpful for newbie analysts who can “follow” more experienced analysts and learn by seeing what data sets they use and the analytics they create.

Summary

Power user networks are great ways to bring data analysts out of the shadows of individual departments and upgrade their skills and knowledge at the same time. When data analysts feel part of a larger group, they gain confidence, perspective, and skills which turbo-charge their productivity and insights. Organizations that have strong power user networks can offload more development from IT, streamlining the delivery of information and insights to business users. 

EXTRA

Frustrated. Data analysts in organizations that do not have a strong culture of analytics or robust data infrastructure spend more time collecting and massaging data than analyzing it. These so-called “human data warehouses” often get frustrated and leave the organization because they don’t have the power to fix the processes that generate poor quality data. In other cases, these data analysts become glorified report writers; rather than analyzing data, they simply maintain and run reports.