The first operator - the case for 'finance and analytics'

Prior to joining Intercom, my background was in finance. I’d worked in FP&A at a large company (IBM) and then FP&A again within a smaller 150 person startup. When I joined Intercom I managed analytics for the first time. I directly hired the initial data analyst team as well as our first data engineer.

My running joke is that they gave me analytics because, well, numbers! They’re all the same! The reality though is that I was quite deliberate and intentional in bringing those two teams together. In my role prior to Intercom, I consistently saw the work of finance and analytics collide, overlap, and duplicate efforts.

As such I’ve grown to believe that in the early days of many startups, finance and analytics are better together (another day we can debate the long term home for analytics), and it’s because there’s so much overlap in what those teams do.

Early finance is analytics. Early finance isn’t accounting, it’s not accounts receivables, it’s not payroll or procurement. It’s not even forecasting and planning. As I’ve written about before, the role of the first finance hire is to grow the business through organizing frameworks and analysis.

The early days of finance involve: organizing and defining metrics, diagnosing data issues, building scrappy first marketing attribution and customer lifetime value models, establishing a culture of metrics and quantitative analysis, and using quantitative analysis to create insights. And if that sounds just like early analytics, well, that’s the point.

Furthermore, the next most important finance responsibilities - forecasting and planning - are only as useful as the analysis on which they’re built. Said another way, you can’t forecast and plan effectively if you haven’t pulled apart the business to understand how it works. The argument extends in the other direction - understand your business and your forecasts and plans will build themselves (ok, for the most part… 🙂). At the highest level, great finance teams are built upon a foundation of great analysis.

At the same time, many of the more sophisticated responsibilities of an analytics team are not useful at an early stage. Unless core to the product, most startups don’t need sophisticated data scientists building machine learning models despite thinking they do (please don’t hire one, you’ll be unhappy as will they). Similarly, the more elaborate the marketing attribution or LTV model, the less likely it is to be useful. Most early analysis is directional, scrappy, and favors correlation over causation. Speed matters, and simple analysis can make a big difference.

All analysis ultimately points towards the fundamental unit of every business, return on investment (ROI). As I've written about before, it’s why every analyst is a finance analyst and so much of that early analysis needs to incorporate financial tradeoffs. As the peripheral responsibilities of finance and analytics are stripped away, the core goals of these teams collide. Brought together they can more powerfully work to bring insights and a shared understanding of the business.