How to build processes and systems that create a data-driven culture.

All companies need data to drive decisions, but becoming a data-driven company takes more than just accumulating data. As stewards of cross-functional data, finance can play a role in developing an effective data management strategy.

Rachel Bradley-Hass is Co-Founder of Big Time Data, which provides data enrichment and automation services. She joined Airbase CFO, Aneal Vallurupalli, in a What I Wish I Knew webinar called How to build a data-driven company to share her experiences in building data infrastructure. Rachel said she’s seen a lot of chaotic situations connected to data in the companies she works with. “I’ve never walked into a place and thought, ‘This is better than I imagined it would be!’”

Instead, she typically finds that data is pulled from multiple disparate sources, without any thought to how they connect. Then, as companies grow and acquire even more tools, their data becomes increasingly unruly. “The migration of data from tool to tool means things are all over the place,” she explained. That can lead to messy practices like a lot of manual entry, or prorated numbers with different incorrect dates in different systems.

“What sets the MQL status in Salesforce? Is it Salesforce? Is it Marketo? Because I’ve seen companies where it’s both and then they can’t figure out why it keeps changing back and forth,” Rachel said.

Aneal has experienced the effects of disorganized data firsthand. In his experience, data that is all over the place is hard to reconcile and report. “I can’t build a model to forecast anything if I can’t get the data reliably and consistently.” He added that one big drawback is that he has less time for analyzing data if it takes a long time to gather and organize it in the first place.

Best practices for building a data-centric company.

1. Understand why you need each piece of data. 

When working with a client, Rachel likes to meet with each member of the leadership team to talk about what success looks like for their department. What numbers matter to them — and why? The point is ultimately to understand the story they want to glean from the numbers. If you don’t have a complete understanding, you are more likely to build something that won’t be scalable.

2. Approach data architecture from a neutral perspective.

Rachel acknowledged the cross-functional nature of creating data architecture. She recommends having one person or department take ownership, but it’s important that they’re able to maintain a neutral perspective.

Because they typically have insight across departments, finance often acts as the agent of change with regard to data. This is partly because poor data greatly impacts the finance function. Rachel described a typical pain point that acts as an impetus to get a handle on data:

3. Define your terms.

Common SaaS metrics like MQL, ARR, and net retention can all mean slightly different things to different departments, different industries, and even different people. Everyone brings their own biases and history. That’s why it’s important to create very precise definitions of relevant terms. In Airbase’s recent Off the Ledger LIVE! session, Finance’s role in telling stories with data, Aneal said he’s become infamous with his coworkers for his insistence on having a very specific definition of customer because many metrics can vary depending on exactly how a customer is defined.

He recommends maintaining a central repository of definitions that can be easily accessed. That also helps employees understand which numbers make the business successful. “People immediately have clarity on what matters,” Aneal said.

For each term, it should be clear:

• What are we measuring?
• What is the layman’s definition?
• What do you want the number to mean?
• Does every department align on the meaning?

Having precise definitions helps create the data to support the narratives of each operating unit.

4. Connect the data.

Rachel said she’s seen a proliferation of data visualization tools in many companies: “Everyone wants their own particular visualization.” She is an advocate for a data warehouse-first approach because it creates one consistent, reliable single source of truth. She explained that data warehousing is not as scary as it sounds because there are a lot of plug-and-play tools that can assist once the data is defined and mapped.

Gaining a thorough understanding of the purpose of each piece of data is essential to understanding how to map it in a way that will scale. “I think if you don’t understand the data needs, you end up building something that will need to be revamped.” Rachel said.

5. Data as the driver of change. 

With consistent data across departments, it becomes easier to use numbers to shape policy and communications. Rachel noted that customers increasingly expect that a software vendor will be aware of usage patterns with their customers. For example, if a customer success manager has data showing how a platform is being used, they can then shape communication about upgrades accordingly. Similarly, data showing the time lapse between signing a contract and going live can inform the onboarding process.

Discover how Airbase’s consolidated platform provides consistent, automated data for AP functions. Book a demo today.