With the vast ecosystem of information technology systems available to management along with the wide-ranging analytical techniques that can be applied to data, knowing where to start the journey toward a data-driven decision-making approach can be daunting to management. We argue that this journey should begin with and be tailored around the Key Performance Indicators (KPIs) for a particular business. In other words, depending on the industry and business model, the activities around data collection and the resources allocated to analyzing that data should be prioritized based on improving firm performance and referencing KPIs.
This article intends to provide clarification on some of the terms and concepts that are often thrown around as implying importance for business practices but are conceptually opaque with respect to firm execution. This article further intends to provide clarity on how data can be used effectively. To further illustrate this, we focus on a subscription-based business model. What follows is a discussion of analytics strategy, an illustrative case study presenting the considerations of a KPI-centered approach used by a hypothetical Software-as-a-Service (SaaS) firm along with some concluding thoughts.
The Relationship Between Business Strategy, Data, and Management Practices
Business Strategy
Alignment or the fit between IT and business strategy is important.[4] A starting point in assessing the value of IT or technologies, which in this discussion comprises technology systems, data, and analytics, is to trace the impact back to business processes such as supplier relations, customer relations, product and service enhancement, production and operations, and sales and marketing support.[5] From a strategy perspective, given that firms may focus on some areas more than others based on industry and business model (e.g. customer service is often a significant focus area in the restaurant industry), there is a justification to prioritize data and analytics practices around processes that make the firm more competitive. Additionally, this is an important consideration even when firms are competing on a similar end product or service but are providing different value propositions.[6]
Data Sources and Types of Analytics (A Brief Review)
Sources of data that could be categorized as either being internal data or external data where internal is data what the firm collects and owns while external data comes from a source that is outside of the firm.[7] Internal data sources might include transactions from internal operations and interactions with suppliers and customers, firm website traffic, RFID/IoT sensor data, and data collected from machines and/or vehicles. External data sources on the other hand might include data purchased from a third-party, publicly available data that is free (i.e. U.S. Census Bureau), and community data from social media that is not explicitly owned by the company.
In addition, there are several types of analytics that have been identified: descriptive, diagnostic, predictive, and prescriptive. Descriptive analytics and diagnostic analytics are often referred to collectively as business intelligence while the latter two, predictive analytics and prescriptive analytics are often referred to as advanced analytics.[8] Additionally, the technology, tools, and skill set required are more advanced as one goes from descriptive to prescriptive (see Figure 1).
Figure 1: Descriptive, Diagnostic, Predictive, and Prescriptive Analytics (Source: Authors)
More importantly, there has been a shift in attention toward better management and prioritization of data referred to as data governance. Data governance includes structural practices (e.g., identifying responsible parties for data ownership, analysis of data value, and cost management), operational practices (e.g., actual organizational activities related to governance including data migration, data retention, e-discovery and archive activities, setting and monitoring of user access privileges, implementation of backup processes, and execution of storage procedures), and relational practices (e.g., educating data users and non-IT managers on storage and costs and flow of communication across executives, IT, and data users).[9]
Given the blossoming of data collected, it is important to consider what data can result in valuable, actionable insights as well as what practices are in place to prioritize the collection and access to such data to facilitate analysis. Furthermore, this may require not just attention to the source of data but also what fields are retained that may allow someone to make sense of it. This prioritization or attention to value can lead to an outcome where data costs can be minimized and value or return can be maximized.
More importantly, we find that key data points that are often not described within the analytics realm, as they are not in themselves traditional datasets, could come from disparate industry reports, market surveys, or other key external sources of knowledge that shed light on the competition. Thus, while it may be easy to fall into the trap of focusing on traditional internal and external data sources and insights derived from this data to improve decision-making, one should also think critically about data or information that allow for benchmarking.
Collectively, we argue that a KPI-centered approach should drive the development of management practices around data analytics strategy as this would help managers identify what matters and what the firm should prioritize.
A KPI-centered Approach
Key Performance Indicators—Measure What Matters
- If you could know exactly how your business is performing merely by looking at one screen—on your mobile phone even—what would that screen look like?
- What measures would it capture?
We propose that some more explicit steps or guidelines are fitting to properly illustrate how resources can be better allocated to address meaningful measures. To distill this further, we argue that if the objective is to tackle business performance and build out the right KPIs using data that there are four main steps:
Figure 2: Key Performance Indicators—Measure What Matters (Source: Authors)
- Start with the Industry: Designing and refining key performance indicators for your organization should begin at the industry/sector level. This is where similarities in transaction types, products and service offerings, and operating models provide comparability for performance measures. Furthermore, there are generally available datasets compiled by industry research groups that provide business leaders with meaningful benchmarks to assess performance and make strategic operating decisions.
- Consider the Inputs: KPI measures may be financial (i.e., revenue, profitability, cash flows) or operational (i.e., supply chain, human resources, manufacturing) and the related business processes (i.e., supplier relations, customer relations, production and operations, and sales and marketing) that impact your attention to certain technology systems and data can be multi-faceted. A sound starting point for considering the inputs begins with identifying the related business process flow to which the KPI relates. Specifically, consider a process narrative or graphical representation such as flowchart or process video as a starting point.
- Identify and Capture the Right Data: With the proliferation of organizational data assets, business leaders must partner closely with technology leaders during the design and implementation of strategic initiatives. Quite often, companies do well to capture data about their customers, products, supply chains, operations, etc. However, the equally important final step of making the data available in useful ways poses difficulty. Delivering data in useful ways can mean system integrations (i.e., APIs), scalable reporting (i.e., automated and real-time), and making this information available to key users on any device.
4) Establish ownership for the KPI: A common saying we consider in working with clients is “The business doesn’t run inside a spreadsheet.” Certainly, we agree that turning these analytical insights into better organizational performance requires effective communication, aligning management incentives and generating stakeholder buy-in. To illustrate, apply this four-step process to the SaaS industry:
SaaS as an Industry
Worldwide End-User spending for public cloud SaaS is estimated to reach $123 billion in 2021.[10] The SaaS business model is unique, and so too are the measures of company performance. More generally, the “as-a-service” or subscription-based business models generate revenue from contracted subscriptions which provides a revenue base that is more predictable than non-subscription-based models where revenues are generated from individual sales and separate buying decisions from customers. This subscription-based revenue model provides the predictability that can help assist with near-term business decisions. Furthermore, as a company gains an understanding of its subscription customers, the rates at which they renew and the needs and wants for additional products or services, the predictability of revenues can extend beyond active subscription terms and instead over the lifetime of a customer. Additionally, unlike non-subscription-based businesses that must build, buy, or otherwise create each new product it sells, SaaS businesses need only to issue incremental licenses to already developed software which requires less incremental cost for each new dollar of revenue. As a result, these nuances in business model and value proposition are at the core of the KPIs used for assessing SaaS performance.
Example Case Study: Under the Bridge Software
To make this point, let’s look at an example company: Under The Bridge Software (“The Company”). In this example, we assume The Company has achieved a YoY Revenue Growth Rate from 2018 to 2019 of 21.42 percent and seeks to improve this YoY Growth Rate during 2020. The Company has an established data-driven culture, leadership, and management embrace data analytics, The Company has the infrastructure and processes to capture and review the data points relevant to the YoY revenue growth rate KPI, and has the required analytics resources available to visualize, model, and understand the components of this change, at a customer level.
1) Start with Industry:
In the SaaS industry for example, one highly regarded “dataset” used in benchmarking and assessing Company performance is the annual KeyBanc Capital Markets Private SaaS Company Survey,[11] which compiles data from more than 500 senior executives from SaaS companies around the world submitted anonymously and confidentially. Not only can this data be helpful in benchmarking performance against comparable companies, but it can also serve as a roadmap for the type of metrics and data that your organization might want to be capturing.
In the SaaS industry, revenue and revenue growth-related KPIs are most commonly used to compare companies and scrutinize performance. The components of revenue growth therefore become key considerations in modeling future expectations and improvements in these metrics alone can have significant implications on enterprise valuations. These components, namely revenue churn, upsell/downsell, and revenue from new bookings/logos serve as key considerations for improving performance. A visualization such as the FY18 – 19 Revenue Bridge below brings these trends into focus.
A Revenue Bridge is a form of a waterfall chart that provides a visual representation of increases and decreases in revenue during a given period at a variety of categorical levels effectively summarized for an executive audience. At a high level, The Company experienced low churn (-0.6%), considerable upsell growth (+13.6%)—primarily from expanding the number of users and new logo revenues of +8.5 percent of FY18 revenue.
2) Consider the Inputs:
Revenue—in a SaaS business model—is a function of the number (#) of users over a given period of time times (*) rate per user. Revenue-related KPIs are financial measures and the related business processes are within sales and finance functions of the business. More specifically, revenues from expanding relationships with existing customers are generated by a different business process (i.e., contract renewal order administration) than revenues from acquiring new clients (i.e., sales execution) and separate considerations are needed. The inputs used in the Revenue Bridge are the following: Revenue detail by Customer, Customer acquisition dates, Customer termination dates, User volumes, Rates per user, and Product Data (SKUs).
Table 1: Under the Bridge Revenue
3) Identify and Capture the Right Data:
The data points required for the Revenue Bridge in our example include billing and revenue details per customer, which are typically maintained within a Company’s ERP, and customer acquisition date, price, and volume which are quite often maintained at the CRM level. In many cases, system synergy between ERP and CRM enables multi-dimensional customer level analytics which can provide key insights and context into performance. In order to prepare the summary bridge, some data organization is required potentially by a junior analyst who has the ability to compile and analyze data from different sources. The revenue detail by customer for each fiscal year is compared and then additional customer level data points are used to stratify the components of the change. In the context of the problem or question, The Company utilized both internal and external data points and only needed to reach the Business Intelligence level of analytics. The table below summarizes the inputs of the Revenue Bridge visualization:
Table 2: Revenue Detail by Customer
4) Establish Ownership:
By analyzing the components of this YoY revenue growth, management can see that The Company is doing well in retaining customers and growing their use of The Company’s product. For instance, $975,440, or approximately 93 percent, of revenue generated in FY19 was from customers that existed in FY18. From this, management can see that the growth in existing accounts is a significant component of YoY increases in revenue. It validates that The Company has a compelling product market fit for its offerings and has performed well in on-boarding, supporting, and retaining clients.
Furthermore, it’s clear to see an increase in revenue due to upselling existing customers has been driven largely by an increase in user count (as opposed to an increase in prices per user) which is a strong indication that customers are expanding their use of The Company’s product without significant per user price pressure—a favorable trend for profitability. These trends in strong retention and upsell are fundamental strengths and validators of The Company’s value proposition. On the other hand, a significant portion of The Company’s revenue growth is generated by existing customers which means that the sales productivity of new logo pursuits has been a performance laggard and could be justification for management to increase its investment in its sales and marketing efforts toward new logos to increase revenues in FY20. Finally, management has to be willing to embrace the collection and analysis of data and also rely on the results to impact decision-making and strategy going forward.
Closing Remarks:
In closing, whether your business is in the SaaS industry or in any other industry, what remains constant for any identified KPI is that it represents a measure of what matters. For this reason, KPIs present a reliable recipe/starting point for developing your organization’s data analytics strategy so long as they are defined within the context of the business model and industry. To monitor these KPIs, firms may benefit from a dashboard pulled from a series of key reports. While the cost of collecting data can seem daunting, the lower cost of cloud computing has made it easy to store and analyze data without significant expense.
A starting point may be to look internally with several one-off projects to address key firm strategic initiatives executed by tech savvy, eager staff who may be able to utilize the now widely available, low-cost software tools and technical training. Finally, highly qualified, independent consultants detached from large consulting firms who do not charge exorbitant fees may be able to provide assistance in identifying the right technology systems, types of data, analyses that need to be performed, and practices that should be in place to make that happen.
