What Data Democratization Looks Like in Colorado Tech

by Alton Zenon III
June 11, 2020

The most important data tool an organization has is its people.

Data democratization only occurs when team members across the business are empowered with accessible, shared data that makes it easy for them to tailor analysis to their team’s respective needs and workflows. When stakeholders aren’t in sync, dashboards are clunky or employees aren’t trained on how to best leverage data, the wealth of information at a company’s disposal is rendered useless.

Once a company decides how to store its data, leaders must train their teams to analyze the information efficiently. For some Colorado companies, that means creating data teams and giving them the responsibility of owning maintenance and organization before opening data sources to the rest of the company. Data analysts and scientists — some of which are department-specific — build the highly accessible dashboards necessary for successful data democratization. 

The following professionals leverage a diverse toolset, including analytics tools like Presto and data integration platforms such as CloverDX, as well as business intelligence platform Looker. These tools and integrations allow data teams to lay the groundwork for data-based collaboration across the entire business. 

Ibotta team members chatting

Laura Spencer
VP of Analytics • Ibotta

“We haven’t unleashed the full power of data until analytics-based decision-making is embedded into our culture,” said Laura Spencer, VP of analytics at retail payments platform Ibotta.

In order to create that culture, Spencer said employees across the business are trained on how to take advantage of the usability-focused internal tools the data team creates.

 

What were the initial steps you took to break down data silos across your organization? 

We have long believed in the power of democratizing data, so it was less about breaking down silos and more about selecting and building a technical solution to organize our data into a single source of truth. 

We chose to leverage the concept of a data lake because of the flexibility and scalability it provides. Our data lake provides us with the ability to store large quantities of structured and unstructured data sourced from a diverse set of places. This structure enables our decision science and machine learning teams to leverage their choice of analytics tools — like Presto, Spark and others — to quickly explore the data, pull business KPIs and build advanced models that are incorporated into processes throughout the business. 

Our data engineering team led this effort in collaboration with numerous stakeholders throughout the company, including teams that are upstream of these processes. It also included downstream consumers, both within and external to the analytics department.

We chose to leverage the concept of a data lake because of the flexibility and scalability it provides.”

 

What are some of the tools used to integrate your data and make it more user-friendly?

We have worked with Looker for over five years and they have become our primary business intelligence platform. Looker allows us to package key metrics and data queries and easily distribute them internally and externally to our partners. On the engineering side, we leverage Datadog for monitoring and alerting on all of our services. 

We recently introduced Mode into our suite of analytics tools. Mode combines SQL, R, Python and visualization capabilities into a single platform. This combination has been powerful to our analytics team for quick, ad-hoc exploration and interactive analysis.

 

What’s a specific win one of your teams saw from having better access to data?

Our client analytics team built a tool that we call the TrendFinder that highlights current marketplace trends by brand. This tool empowers our sales team to find powerful data stories to share with clients around the importance of maximizing market share during these unprecedented times. In a period where many brands are pausing on spend, these insights enabled our team to retain clients. These types of projects are engaging for the team because they drive clear value to the business.

 

ghx
GHX
Mike Doerner
Director of Data Operations and Engineering • GHX

When it becomes more difficult for a company to scale its data infrastructure, building a team responsible for that goal can help. Director of Data Operations and Engineering Mike Doerner said that’s precisely what happened at healthtech firm Lumere, a GHX company. 

 

What were the initial steps you took to break down data silos across your organization? 

We created a data operations team to own the intake, cleaning and classification of key data sources at scale. We also implemented data analytics roles across other teams in the organization such as product, services and finance. Empowering different teams with access to shared data sources enabled us to tailor any analysis to each team’s respective needs and workflows. 

We created a data operations team to own the intake, cleaning and classification of key data sources.”

 

What are some of the tools used to integrate your data and make it more user-friendly?

Sisense for Cloud Data Teams is an analytics tool that allows our individual teams to query data for reporting and supporting workflows. It also supports the cross-team sharing of analytics dashboards. CloverDX is our extract, transform and load tool for many external data sources. We employ it to automate the intake and validation of data sets from our customers and other sources. Stitch is an integration tool that allows us to combine data from other sources like Salesforce with our internal data.

 

What’s a specific win one of your teams saw from having better access to data?

Our data operations team works closely with the medical device and pharmacy subject matter experts on our research team. We rely on their knowledge to accurately codify, classify and curate our data. Analysts on the data operations team created interactive, analytical dashboards that allow the research team to monitor trends and identify and investigate potential outliers. 

This feedback loop improves our overall data quality on behalf of our customers. Most recently, the close collaboration between the research and data ops teams helped us more collaboratively partner with our parent company. We compiled a list of supplies that were at risk of shortage during COVID-19. We then provided complimentary evidence of alternate supplies to help health systems across the country better navigate the pandemic.

 

Jobs from companies in this blog

Colorado startup guides

LOCAL GUIDE
Best Companies to Work for in Denver & Boulder
LOCAL GUIDE
Coolest Tech Offices in Denver & Colorado Tech
LOCAL GUIDE
Best Perks at Colorado Tech Companies
LOCAL GUIDE
Women in Colorado Tech