Welcome to the second installment of my blog series describing Value-Based Design (or VBD), which encompasses my best practices for driving meaningful business outcomes with analytics.
As described in my first blog post, I have had the pleasure of working with companies of all sizes in a wide array of industries over the course of my career, and I have learned a great deal about what leads to analytics success or, alternatively, failure. In my experience, the Value-Based Design methodology has proven to drive the greatest results, as it leads to shorter implementation times, higher user adoption and greater business impact.
In my first blog, I discussed VBD’s central tenet of identifying the most important metric for each group of people in an organization. This metric, called a Key Value Indicator (or KVI), captures the primary metric for each group, typically identifiable as the metric that determines whether the leader of a group of people gets a large bonus or, alternatively, fired. I also discussed organizing the KVIs from various groups into a Value Plan in which the specific order the KVIs will be addressed via an analytics solution is identified and tracked.
Once the first KVI to be addressed is identified, the work begins to turn that KVI into impactful, role-based dashboards that everyone in the organization can use to improve the overall group’s performance.
In this blog, I will discuss the process by which the critical items that most influence the KVI, or that influence people’s decisions regarding the KVI, can be identified.
The two types of items that need to be uncovered can be summarized as being either (1) a driver metric or, (2) an action-point business attribute
KVI drivers are the metrics that most influence the KVI, either positively or negatively, such that improvement in the KVI drivers can be expected to significantly improve the KVI.
For instance, in an e-commerce business, the KVI might be Revenue and the drivers might be: (1) number of website visitors, (2) percentage of visitors creating carts, (3) percentage of carts with completed purchases, and (4) average dollars per completed cart.
Improvement in any of these drivers would substantially improve the overall performance of the Revenue KVI. KVI Drivers are best uncovered through conversations with business leaders because they typically know what metrics most influence the KVI for which they are responsible.
If, however, the business leaders are not yet analytically-minded and do NOT know the drivers for their business, a leading question can typically be asked to uncover the business drivers: “What numbers do you look at one month into the quarter to know if you are going to meet your goals at the end of the quarter?” The answer to this question is often the top KVI Driver.
Once the drivers are identified through conversation, a simple check can be performed to see if all the relevant drivers have been identified: Can the KVI be expressed as a formula of the drivers? Often, the KVI can be expressed as a formula of the drivers, as is the case in the example above.
This relationship often exists because the KVI is often an output of a business process – and the output of a process can typically be expressed as a function of (1) the number of items entering the process, (2) the percentage of items that successfully complete the process, and (3) the average value of the items completing the process.
In our earlier example, the Revenue KVI equaled the number of visitors to the website, times the percentage of visitors creating shopping carts, times the percentage of carts for which the purchasing process is completed, and multiplied by the average dollars per completed cart. This relationship exists for many KVIs throughout most organizations, but it is not necessarily the case everywhere. For instance, at Birst, we know that a higher Net Promoter Score (NPS) – a management tool that can be used to gauge the loyalty of a firm’s customer relationships – results in higher retention and upsell rates in our customer base. However, there is no formula that relates how much a given improvement in Net Promoter Score will yield in terms of increased retention or upsell. So, while NPS is a significant driver for renewals and upsell, it cannot be used in a formula that equates to renewals and upsell.
Another example might be stock-outs in the retail industry. Each time an item runs out on the store shelf, revenue is lost, but it is hard to determine by how much. So, the goal is always to express the KVI as a function of some or all of the KVI Drivers, but it is not always an achievable one.
After the KVI Drivers are identified, the next items to be identified are the Action Points. Action Points are the business attributes that most influence a business user’s decisions when trying to improve a given KVI. The Action Points are far more internally focused than most people expect, but the reason for that is simple: people can change things internally quickly – it’s hard to change things externally! While the specific action points for any business can only be identified through conversation with the business, there are distinct patterns in where those action points are most commonly uncovered.
The primary areas where Action Points are typically found are (1) people, (2) product, and (3) process. The first action point most business users think of is people, or, more specifically, the people in their organization who are responsible for managing a part of the overall business.
For instance, a sales leader will generally first want to know who on their team is doing well, and who is hurting the overall performance of the group, when trying to figure out how to improve overall performance. How the different individuals are performing greatly informs the business leader’s decisions, so it is imperative that people be included as an Action Point in most analytics solutions.
The second most common action point is product. Most companies offer multiple types of products, and they behave very differently when one product is in play versus another. For instance, a software company makes very different decisions when it’s dealing with high-margin software licenses rather than low-margin professional services. In addition, an insurance company behaves very differently when selling Life Insurance or Homeowners policies. Because product can dramatically influence a person’s decisions, it is often a critical Action Point to include in an analytics solution.
The third most common action point is process. Because many KVIs measure the output of a business process, the steps in that process often appear as an Action Point.
For instance, in a sales organization responsible for delivering revenue, the steps in the sales process leading up to the revenue event are typically an Action Point (e.g., Sales Stage). Another example would be a finance team responsible for driving Payments will want to analyze the outstanding invoices by Invoice Status (e.g., overdue, more than 90 days overdue, etc.). Because different decisions will be made, based on where certain items are in the process being managed, these process steps are an important Action Point in most analytics applications.
For each KVI, these KVI Driver and Action Point conversations need to be conducted with the business leaders. These conversations need to focus on the decisions the business leaders make and the factors that most influence those decisions. As described in my prior blog, it is imperative that these conversations avoid delving into the reports that people use today and instead focus on the decisions that people need to make.
Each business will be different, but the rules of thumb identified in this blog will help you engage with your business leaders in a new way. These steps will help the organization drive better, more valuable decisions once the development process is complete. I will dig into the next step in the process, which is identifying and exploring the decisions made by each functional group in the organization, in my next blog.
Paul Staelin is SVP of Go-to-Market for BI & Analytics at Birst, an Infor company. He leads Birst’s go-to-market activities, including product marketing/strategy, business development and demand generation. Paul is a co-founder of Birst and previously served as its Chief Customer Officer, a role in which he was responsible for Birst’s customer experience – including its customer success, professional services, customer support, education and community organizations.