If you’re writing checks and throwing the bills away, you’re throwing information away. Invoices hold much more information than simply goods purchased or services rendered and balance due. Taken in aggregate, invoices provide insight into spending patterns, and can be correlated with the business’s overall development and financial health.
Identifying those insights can’t be done by reviewing invoices as they arrive. Instead, the data needs to be collected and then studied through analytics methods. Here’s what you need to know to start putting an analytics process in place.
Identify Your Goals
You’ll be more successful in your analytics project if you identify what you hope to accomplish before you begin. Some common goals of payment analytics projects might be:
- Identify fraud and payments that don’t adhere to policies at lower cost than through audits.
- Monitor compliance with contract terms.
- Gain insight into true costs by reviewing intra/inter company expenses.
- Improve cash management and cash flow by optimizing payment agreements/schedules with vendors.
- Improve payment efficiency through identifying key sources of errors and bottlenecks.
- Compare spending patterns and efficiency between different organizational units.
- Identify opportunities for reducing spending through bulk purchases.
Get the Technology Right
The success of analytics projects depends heavily on the technology environment. It’s easy to get started simply by defining reports that can run directly on top of your database; many accounting packages will have built-in reports that support these goals. More specialized, customized analytics will require building your own environment.
This environment will typically sit in the cloud; cloud computing can support the large volume of data needed for analytics as well as make the generated insights available from any location or device.
As the volume of data grows, specialized analytics tools rather than simple database queries are needed to work with the data. Cloud computing vendors often provide specialized services for analytics, providing technology such as Hadoop that can work efficiently with big data using a distributed cluster.
Get the Right People
Implementing analytics requires a combination of technical and business people. The business intelligence skills needed to create simple reports are common, but more specialized skills are needed to create a full-fledged analytics environment. Your technical team will need to be skilled in setting up the analytics environment, which—if you don’t use a cloud provider’s environment—can require configuring both hardware and software. Much of the team’s time will be spent collecting and manipulating the data to be analyzed, extracting it from one system and transforming it to be loaded into another. Database skills, data warehouse and data lake knowledge, plus ETL (extract-transform-load) experience are key. Additional technical knowledge is needed to implement the analytics programs, such as skill with Hadoop or Spark, as well as general programming ability.
But asking the right questions and knowing what to do with the patterns and predictions identified by an analytics program is a business problem, not a technical one. The analytics project shouldn’t be driven solely by a search for patterns in the data, but by a search for patterns that can impact and improve business decisions. This knowledge of what questions are interesting to the business can only come by having a business user guide the analytics team.
Results on paper are meaningless; it’s when the results of an analytics project are put into action that the analyses have value. Getting buy-in from senior management is necessary to ensure that the results are used. Make sure the project’s goals align with key business concerns, but start with a small project that will demonstrate the value of analytics and support further investment into these projects. Once management sees the benefits of an analytics project on the business’s operations and bottom line, they’re likely to support bigger, more extensive analytics projects that take longer to complete but have more impact.