Joy Zhang is a Customer Growth Manager at Vareto. Past experience includes Finance Manager, Prime International Expansion at Amazon, supporting the success of the Prime Program in recently launched countries worldwide, and Finance Manager of Decision Support at UW Medicine, where she supported strategic growth programs across the UW Medicine system.
When I went into finance, I didn’t expect to become an expert in data analytics. I anticipated a very traditional FP&A or accounting-adjacent role, but my work often focused on wide swaths of data, most of which wasn’t financial. These experiences gave me a new perspective on finance’s role in setting the trajectory for the business.
We know the finance function is evolving into more strategic and future-forward business partners. Finance teams can’t operate in silos anymore, and data analytics provides a more holistic picture of the business. With the additional data, you can achieve much more meaningful insights and use those insights to spend your dollars and time more effectively.
If your team is trying to move beyond the financials, there are a few key questions to answer as you get started:
- What’s your goal?
- What data do you need?
- Where does the data live?
- What’s your data governance framework?
- Is the data clean?
Where to start: What’s your goal, and what data do you need?
There’s a sweet spot for the amount of data you need to succeed. Most people start by asking if they have enough data. While that’s an OK question, I’d counter: Enough data for what? What is the business goal?
Yes, you need a significant amount of data to draw conclusive insights. But it needs to be the right data for the problem at hand.
Avoid the temptation to pull anything and everything available to you. Large companies often have an abundance of data. You can have too much of a good thing. I ran into this problem when I worked for a large retail organization. We housed terabytes of data. It could take hours to query what I needed. This meant that I had to be very specific and purposeful.
Narrow the scope to match your objectives. If the problem statement is too open-ended, you can spend hours, days, or even weeks wading through an abundance of data. Be efficient with your time by 1) making the business goal explicit, 2) evaluating hypotheses that your data will address, and 3) scoping what types of data will get to your answer.
For example, At Amazon Prime, our finance team would ask—How do we increase topline revenue for our international countries where we just launched? Then drill down even further—What levers can we control that will make a difference?
From there, formulate a hypothesis to explore. If you can control the shipping speed, examine whether increased shipping speeds will drive more conversions.
So, in this example, we have the business goal, and we have the hypothesis. That helps us dig into specific data sets that need to be analyzed. We could pull shipping details or historical purchasing behavior to determine if we can draw any relationships or correlations.
The next challenge: Where does the data live, and who has access?
Siloed data is a challenge for many organizations. This isn’t always due to ineffective tools or processes. Data governance frameworks include quality, stewardship, security, compliance, and management rules. Because of these concerns, there’s a level of access control and restrictions that need to happen to protect both the business and the consumer. Robust data governance frameworks are a good thing.
Data governance rules can pose challenges, though, for data analytics. If we only have access to certain data sets, it can be hard to validate if what we’re seeing is true on a broader scale. To continue our shipping example, if you only have access to one country’s shipping information, it’s hard to determine if the trend you see is a global consumer trend or a local trend. You can't validate your findings without access to data from other countries. To do so, you would need to request access to additional data or request help from someone who has access to conduct the same analysis on their side and deliver the results.
Once you have more processes in place, things inevitably slow down. That’s another reason it’s so important to thoroughly scope the project before you start. It helps us understand potential blockers ahead of time and makes the project more efficient.
Last but definitely not least: Is your data clean?
You may have heard this before: garbage in, garbage out. You can’t use messy or incorrect data to draw reliable conclusions. Data warehouses like Snowflake are great for pulling data using SQL. But if you’re seeing discrepancies in the data, you have a problem.
Manual data entry is often a primary culprit for messy data. Inaccurate entries, tagging mistakes, or inconsistencies in nomenclature can spoil an otherwise effective analytics project. Data cleanliness issues can grow significantly once multiple sources of data are introduced, with teams spending lots of effort to reconcile between their different systems. Companies that reduce manual data entry and consolidate to a single source of truth scale more efficiently.
While the finance team may not be positioned to lead a digital transformation for your entire organization, you can start with financial data. If your team relies on manual entries and stagnant reports, consider a tool with advanced automation capabilities and robust integrations.
For example, with Vareto, integrations automatically compile siloed data streams from different business systems into a single source of truth. New actuals data is automatically categorized and placed in the right home for analysis, reporting, and modeling purposes. This eliminates manual, ad-hoc maintenance and increases accuracy and efficiency.
Finance for the future
The gold rush to implement automation and AI has started. Every Finance team wants to reduce manual work related to data gathering and analysis and have more time for strategic activities. With these changes, the finance function needs to evolve. While modern software like Vareto means the Finance team won’t be writing SQL queries, it helps to have an understanding of datasets outside of Excel so you can effectively partner with other teams across the business to consolidate data and get a 360 view of the business. Taking a basic data course can help Finance teams gain an appreciation for how data is structured and what issues might arise with data entry, data pipelines, and more.