4 steps to successfully lead analytics teams
Leverage your data
Norman Pai | July 16, 2015
Growing competition is putting an increasing amount of pressure on leading companies to evolve their approach to analytics. Many are stepping beyond operational dashboards and modeling spreadsheets and finding new ways to leverage their data.
While these insights fuel valuable innovation, many organizations embark on these initiatives without tackling some of the upfront challenges—namely, collecting evidence and getting buy-in from leadership.
I have seen how great managers and executives can lead and be force multipliers to their analytics teams, from startups to FORTUNE 500 companies. From those experiences, I have developed a four-step path to solicit support from decision makers and lead an analytics team. The following framework is designed to tackle sequential challenges that exist in infrastructure, data, analysis, and results.
1. Assess the status quo
By first assessing existing people and their capabilities, missteps can be avoided down the road. An organization will always need a variety of technical skill sets to meet current and future demands, but the best people are dynamic and can adapt to the organization’s needs.
These smart creatives are both analytical and business savvy risk-seeking individuals who will make a significant positive impact in your enterprise. These individuals must also recognize the need to hire and recruit.
The second critical assessment is on technology infrastructure. When the organization is set to move, make sure there are not insurmountable technology gaps between what is conceivable and what the available technology will realistically support. Do not be timid. Let the team discover their potential to solve challenges by supporting them with the appropriate infrastructure.
“The only way for businesses to consistently succeed is to attract the best smart creatives and create an environment where they can thrive at scale."
2. Remove data roadblocks—and political ones, too
An important part of establishing a viable environment for analytics is being able to access the necessary data and understanding it. At larger enterprises, data is often splintered across department silos, requiring several levels of necessary approvals. Having active and engaged stakeholders can save days—if not weeks—of project time.
At small companies, the data may not be ready at the outset of the project. One client I worked with, a rapidly growing tech startup, needed a formal data warehouse to enable the business to easily consume and analyze data around revenue, users, and products. Support from the company’s directors and CEO to access visibility to the original data sources saved valuable time. Our work included connecting with customer service representatives, who knew the details of complex customer billing—invaluable context for designing a data warehouse that opened up product and financial metrics to the whole organization.
The lesson: Getting rid of data roadblocks often works top-down, but sometimes it should work bottom-up.
3. Select the right analytics tools and processes
The appropriate tools and project methodologies can be powerful productivity enhancers. As a leader promoting analytics, understanding the business cases and relative priorities will help define the problem at hand. Having gone through the initial technology stack assessment, infrastructure needed for the organization’s new analytics needs should be available.
For each technology, consider whether it holds a place in the organization’s short or long-term architecture. The latest generation of business intelligence and advanced analytics can provide your team with quick training and hands-on experience.
Following tool selection, you should formalize the team’s operating protocol. Agile or lean methodology often ensures flexibility and an effective use of the tool you’ve selected. Once you have empowered your analysts and managers with productive tools and a powerful operating environment, they will be ready to execute the analysis and produce results.
4. Communicate feedback and socialize results
After obtaining data, analytics teams should be able to run analysis and output results. What is sometimes missed is that these results need to be shared outside of the team.
If we take a look at a popular cross-industry process for data mining (CRISP-DM), this feedback exists in the loop back from Evaluation to Business Understanding. Facilitating this feedback loop is a critical part of leading the analytics team. This process is a great opportunity to sync everyone’s understanding and expectations.
While there will always be significant technical and staffing challenges in tackling and understanding data, your role as an effective manager and leader can be the difference between success and failure.
Norman Pai is an information management and analytics consultant in Slalom's San Francisco office. He is passionate about transforming raw data into actionable insights and delivering innovative solutions that solve critical business problems.