With new technologies transforming the business environment at an ever-faster rate, there’s become a greater need for insightful customer analytics. When it comes to finding trends and making timely business decisions, it’s no longer enough to analyze traditional financial data. Finance departments also need to incorporate operational data – especially customer analytics – into their forecasts and reports.
This need is particularly apparent in the tech industry, where product updates and upgrades are frequent and KPIs like customer profitability are tracked with fervor. Users are distributed across different devices and are using different versions, signing up for free and paid services alike, and income statements paint only a small part of the total picture. To make profitable decisions regarding product roadmaps, bug fixes, sales and marketing, these companies need detailed insights into specific groups of customers.
The Pitfalls of Pure Financial Data
Income statements, balance sheets and cash flow forecasting – these are the typical financial reports that help companies choose how, when and where to push or improve upon specific products. A manufacturer might increase its production of a hot-selling item, for instance, while investing in R&D for a product with declining sales.
For enterprise software, mobile apps and other tech, however, you need more granular data – particularly when customers are making purchases in the application. You’ll often have a blend of new and old customers contributing to each line item, and pure financial statements only show you an aggregated view of that diverse pool. For example, you might see a particular product is growing in revenue, but you won’t see specific revenue contribution by new, month-old or year-old customers.
It is important to understand that traditional financial data treats all product versions the same. Customer acquisition, retention, conversion, average revenue per user (ARPU) and other critical KPIs can differ dramatically from one version to the next, but you won’t gain those insights from your ledgers. If one product version contains a bug or unpopular feature, those differences will have a major impact on your firm’s finances; you just won’t know exactly why or how.
Cohort Analysis: A Must for Forecasts and Trends
Fortunately, there is a solution to those pitfalls – cohort analysis of customer analytics. Tech-based products rarely experience bugs or crashes uniformly across its entire user base, you need to be able to understand which applications, versions and user groups are experiencing problems that could impact your bottom line. Similarly, cohort analysis offers insight into which product versions are more or less popular among users. If you split up your customers according to versions 1.0, 1.1, 1.2 and so on, for example, you might find out that different versions have wildly different rates of user engagement and retention. Instead of addressing every potential problem in your next update, you can focus on the fixes and upgrades that will lead to greater adoption, greater retention, increase virility and, ultimately, more profits.
This level of efficiency is especially important for updating mobile applications, which aren’t well suited to A/B testing in a production environment. With web applications, companies have the ability to simultaneously test multiple product versions in the production environment. With mobile, however, developers have to contend with the rules and approval process of iTunes and Google Play, which only allow the release of one well-vetted version at a time.
Lastly, cohort analysis is critical for effective marketing. What do your customer acquisition metrics tell you about cost by channel? How long do mobile users keep your apps installed after they’ve downloaded them, and are retention rates different among iOS and Android users? Which marketing channel is producing the longest lifetime, highest ARPU customers? These are the critical questions your marketing team needs to answer, but to get the necessary customer analytics, your finance team needs to do some extra digging. In fact, tracking even basic metrics – such as revenue per user – often requires your finance personnel to look beyond their ledgers.
Roadblocks to Timely Insights
Why do these responsibilities fall to the finance team? For most companies, the answer is related to a couple of common roadblocks that prevent other departments from getting the information they need. The first of these roadblocks is a lack of resources – a common issue with tech start-ups and other high growth companies. Many firms lack a dedicated BI or data warehouse team, and the finance department has to pick up the slack, gathering customer analytics from application servers, inventory databases and other disparate sources.
The second roadblock is that BI (Business Intelligence) personnel often lack the finance skills and business know-how to effectively relate those insights to marketers, product managers and executives – the people who are making decisions based on the insights. Even if you have a robust BI team, your finance department is uniquely positioned to work with those decision-makers and help them understand the importance of the insights they find. The tech industry operates at a breakneck pace, and this kind of close, efficient working relationship is crucial for quick changes to products and marketing campaigns.
Tools for the 21st Century Finance Department
With so much data to dig through, Excel won’t cut it for a finance department focused on customer insights. Spreadsheets are fine for analyzing smaller datasets once you’ve isolated them, but to get the right data in the first place, most finance departments will need to expand to more powerful analysis tools.
First, a database querying tool such as SQL is necessary for finding and isolating cohorts within enterprise-wide databases – for instance, the group of users who have downloaded the most recent version of an app within the last two months. Second, you’ll need a BI tool such as Tableau for combining data and visualizing cohorts. Relevant data related to a single cohort may be stored on multiple servers, and these tools are indispensable for quickly bringing it all together in an easy-to-understand way. Finally, A/B testing, regressions, predictive customer analytics and other statistical operations will require a stats package.
Want more insight?
We invite you to check out our new financial forecasting process guide, which details best practices for improving value and predictability.
The expertise needed to implement these tools and the insights they facilitate is just as important as the tools themselves. Many finance departments lack the tech know-how, BI experience and even coding skills to generate and translate meaningful customer insights to help their financial planning and analysis.