Data Science - The Future of Corporate Finance
Corporate finance must change. Across industries, an organization’s Finance team should shed light on what’s happening today with revenue and other financial indicators, while also predicting what the future may hold. And they must do the same for the entire organization.
Data Science - The Future of Corporate Finance
Author
Mike Scarpelli
Mike Scarpelli
CFO, Snowflake

Data Science - The Future of Corporate Finance

source: https://www.snowflake.com/blog/data-science-the-future-of-corporate-finance/

Corporate finance must change. Across industries, an organization’s Finance team should shed light on what’s happening today with revenue and other financial indicators, while also predicting what the future may hold. And they must do the same for the entire organization.

Until recently, it would have been impossible to meet these expectations. Excel-driven forecasting requires herculean efforts to wrangle data and report numbers by the end of each quarter. But delivering daily insights continued to be a pipe dream for most finance teams. In the end, all that’s revealed is history, not the future.

These days, SaaS planning tools enable rolling forecasts and scenario modeling. This helps finance organizations understand what’s happening in the present and provides the ability to re-forecast business at a monthly cadence. But is that enough?

Data and reporting are usually locked within the SaaS tool, which means two things: 1) It’s difficult to measure in detail what’s changed between forecasts, thus stifling analyses and forward-looking insights; and 2) data can’t be shared with the broader organization, which prohibits a collective understanding across business functions and with key stakeholders.

Finance teams need to evolve and become strategic, data-driven superheroes across an organization’s entire business. Real-time feeds of operational data open up opportunities to be proactive rather than reactive. However, data alone will not fuel this evolution. You must also have the ability to create dynamic models that can re-forecast a multitude of indicators on a daily basis for real-time business insights. And you need to empower Finance to develop partnerships with other functions of the business.

If the future of Finance is data-driven strategic planning and forecasting, then what sets apart a finance organization is its investment in data science.

Embed data scientists in your finance organization

The best way to bring a data-driven strategy to life is to hire data scientists. I’m not talking about borrowing them from IT or resource-sharing with another group. You must bring data scientists onboard as finance team members who live and breathe finance, thus developing an understanding of the day-to-day work and pain points. Embedding data scientists means they can act as functional experts with data and with all the various aspects of finance.

Financial Planning & Analysis (FP&A). These teams, regardless of industry, are responsible for forecasting and budgeting. Data scientists who learn the intricacies of your company’s pricing structure can build a series of models that reflect FP&A’s primary requirement for accurate forecasting. When data science powers forecasting, you receive immediate feedback on how revenue is tracking and can see how it’s evolving over time, which enables real-time adjustments.

At Snowflake, we forecast revenue daily for our usage-based pricing model because a lot of our key financial revenue metrics depend on consumption. This daily reforecasting would be impossible without data scientists who truly understand our business.

Cost of goods sold. Data scientists can also build models to improve financials around the cost of goods sold (COGS). Organizations that either rely on supply chains, or consume external resources in order to deliver a product or service, benefit from analyzing cost structures and margins. Since customer usage evolves over time, opportunities may exist to increase profitability by switching providers or renegotiating vendor contracts. By understanding product demand, you can generate both a revenue and a cost forecast, illuminating opportunities to lower costs, increase margins, or adjust pricing.

Research and development (R&D). In a similar manner, companies may want to conduct an R&D assessment to determine whether it makes sense to develop something in-house or continue to purchase it from a third-party provider. With centralized data, data scientists can model out whether a large upfront investment will pay off and how long that payoff period will be before it yields positive financial results. Alternatively, data models can help determine whether an acquisition is a stronger maneuver in order to bring a specific capability in-house.

Tax and Treasury. Companies looking to launch entities in additional countries need to be aware of tax implications. Treasury teams will want to make sure entities are properly funded, while balancing cost and revenue to ensure the right levels of taxation. Rather than make high-level assumptions, data scientists can model when and where to launch entities based on factors such as customer location, sales, and renewals, and then determine what the impact is on forecasting revenue, costs, and cash flow.

Procurement. Data science can make a difference for procurement through sharing information and ensuring collaboration happens between the procurement function and teams such as IT, marketing, and sales. For example, it’s not unusual for the sales and procurement teams to be completely unaware that each is working with a common customer/vendor, which may present opportunities to negotiate better rates and terms that lower costs.

Partner with functions across your organization’s business

When you embed data scientists into your finance organization, you evolve Finance into a more strategic and proactive function that can also help stakeholders throughout the business be data-driven and make better decisions. The possibilities for collaboration are endless, ranging from the identification of key metrics for individual teams to track, to determining how each team should forecast components of the business to grow and act more cost-effectively.

Beyond data science, one key requirement is that the entire organization shares a single source of data and metrics. Anything less (isolated business systems, Excel spreadsheets) equates to data silos, which prohibit information and insights from being shared in real time.

To promote alignment across an organization, Finance should maintain a set of data models related to bookings, revenue, costs, and other financial data sets that other teams can consume. This centralized source of shared data empowers business functions to speak the same language as the finance organization, work off the same assumptions, and drive their own analysis to improve their functional areas. The result is more effective and collaborative conversations between departments and a stronger alignment around company objectives and vision.

For example, the faster you get feedback on these metrics you’re defining, the easier it is to reinvest more in an area that’s seeing success. If Marketing is experiencing success with a campaign that’s driving a ton of demand, that team can ask Finance for more budget, and that decision can be made in real time.

Sales. Alignment is especially important between functions such as Sales and Finance, where a shared responsibility exists around customer bookings. While sales may be concerned with bookings that happen this quarter or within the fiscal year, the finance organization also cares about bookings one to 10 years out since they impact cash flow. While the two functions may take different approaches and look at different horizons, it’s important to bring together and synthesize that information to build a stronger forecast and a more predictable business.

Engineering. Their focus is on building new features, but the finance organization needs to know how changes pushed into production may impact the cost structure of the product. Alternatively, engineering may release new features that make the product more efficient. Finance needs to know ahead of time because these improvements may reduce consumption, which is good for customers, but may also reduce revenue, which should be adjusted for in the forecast. It’s an iterative process and reciprocal relationship between engineering and finance that requires an open line of communication and shared data models and analyses.

Product Management. Finance’s data scientists should also partner with product management to model the costs, pricing, and monetization associated with launching new features or moving into new markets. Product teams should be empowered to strategize and make decisions on their own, based on financial modeling and their own analysis. The outcome is more efficient discussions and easier decision-making during a product or executive review process because cost, revenue, margin, and pricing implications have been baked into the analysis.

Invest in the right data platform

As mentioned above, a centralized, single source of truth is extremely important for sharing information across the company. From a Finance perspective, a modern cloud data platform should be the backbone that makes all of this possible. It’s the only way to centralize a huge amount of data, process it quickly, and build innovative data models to make real-time decisions.

Scalability and performance. First and foremost, you must start with an extremely robust and scalable system that can easily ingest massive amounts of data, enable analytics to be run on that data at scale, and apply machine learning models in order to predict future business. Near real-time access. Rather than view a cloud data platform as a cost, it’s important to recognize the benefits received from real-time data and the speed at which data can be accessed. As a business, this power enables stronger insights and faster decision-making. Data types. The platform must enable many different types of data to be brought together into one platform. Structured, semi-structured, and even unstructured data need to be processed with speed and used as a single data set for analysis. Data enrichment. Third-party data, such as ratings, fact sets, and industry data should be available to complement your internal data and empower richer analysis. Again, this data must join seamlessly with your internal data. Governance and security. When all data is in one place, you must be able to govern who has access to it and make sure it’s extremely secure. Any time data is extracted, you should be able to see who extracted it, when, and have the ability to ask why. Using a single cloud data platform goes hand-in-hand with my belief that you should limit the number of SaaS point solutions used within the company. You want everything to work together seamlessly and for all data to exist within a central repository so you have one system of truth. Every additional system adds unnecessary headaches. The fewer systems you have, the easier it is to monitor them from a security standpoint so you always know what’s going on in your internal environment.

Wrangle data and invest in data science

Companies have so much data, and much more they can access from business partners, and beyond, in the form of second- and third-party data. Many Finance organizations suffer from siloed data, and manual work is often required to get data into shape to analyze it. Perhaps one of the biggest challenges is combining ERP data with CRM data. It’s not simple or straightforward. That’s why easily centralizing data with a cloud data platform is a critical first step.

Forward-looking CFOs know that you’ll never be able to harness the power of that data unless you can predict customer usage patterns, and many other crucial business insights. That’s why focus should be placed on analyzing data in real time and investing in data scientists who can enrich forecasting and predictability.

No matter what the future holds, one fact remains true for Finance organizations: you’ll always need strong accountants and financial analysts. But you’ll also need those who can build predictive models and understand systems, data, and processes to deliver you great success.

Comments