Written by Dillon Cox

Introduction

The idea of collecting data is not new within the nonprofit world. Today, community-facing nonprofits are realizing how critical data collection is in achieving and reporting on their missions. According to the most recent edition of Salesforce’s Nonprofit Trends Report, organizations that ranked higher for digital maturity, or their ability to leverage data to inform decision-making, had a greater likelihood of exceeding goals during (and after) the COVID-19 pandemic.  Furthermore, these data-driven organizations had a more optimistic outlook on the future, adapted to the pandemic better, and were more successful in mission-critical areas, such as program management, fundraising, and marketing.

Data-driven decision-making is a current buzz-phrase in the nonprofit community, but it is surrounded by a certain ambiguity. What does data-driven decision-making really mean, and how can nonprofits use their data to inform this decision-making in more sophisticated ways? According to “The State of Data in the Nonprofit Sector,” 90% of respondents reported that they sometimes or always tracked data, but only 40% used data to inform their decision processes. Of these respondents, 5% said they used data in each final decision. Furthermore, only 6% felt that they were effectively using the data at their disposal. These survey results indicate there is a disparity in the nonprofit world regarding data usage. So, where is this disparity coming from?

Using Business Intelligence in the Decision-Making Process

While the nonprofit sector is in the initial stages of integrating data-oriented resolutions into their paradigm, the parallel field of business intelligence is being utilized regularly in larger for-profit or larger non-profit corporate settings. Business Intelligence, or BI, is an organization’s ability to convert its capabilities into knowledge. In other words, BI transforms raw data into real knowledge that can be used to inform managerial choices. This process produces large amounts of information that can be used to identify new opportunities for businesses. Once these opportunities have been identified, organizations can implement strategies that provide a competitive edge in the market and ensure long-term stability.

The ability to present data in easily-understood formats is a vital element of business intelligence, especially for those who may not understand data in its raw form (Beibei). A challenge facing many organizations is that upper management, or anyone with decisive powers, do not know how to access the data. This issue is exasperated as IT data providers often do not fully understand an organization’s bureaucratic processes. However, BI serves as the link between these two siloed departments by serving as a tool to gather and encourage finding data-driven conclusions.

What is Business Intelligence?

Modern Business Intelligence Application Process

Business Intelligence is an umbrella term that encompasses the methods of collecting, storing, and optimizing data for operational-related activities. Common business intelligence applications include, but are not limited to:

• Reporting: Sharing analysis with stakeholders to influence decision-making.

• Querying: Asking a particular dataset specific questions.

• Benchmarking: Comparing current performance-related data to historic data for goal evaluation.

• Descriptive Analytics: Referencing data to determine the cause of certain outcomes.

Most nonprofits operate in this space. It is not uncommon for organizational leaders to track key performance metrics (KPIs) and/or evaluate financial reports to assess company performance. The sector primarily operates in the “What happened?” category, indicated in the above graphic. While this section contains necessary first steps, it only encompasses dated information. By utilizing the full spectrum of BI tools, nonprofits can transform raw data and retrospective insights into more thoughtfully-predicted future opportunities and strategies. However, the definition of modern business intelligence is shifting; as more BI tools enter the market, the domain knowledge threshold required to conduct intelligent analytics diminishes. Furthermore, modern BI tools are becoming more refined through integration with sophisticated artificial intelligence. BI is encroaching upon what was traditionally labeled “business analytics.”  The applications included in business analytics are:

• Data Mining: Using Machine Learning to uncover hidden patterns in data.

• Predictive Analytics: Determining outcomes based on a given set of parameters.

• Optimization: Maximizing likelihood or success of a predicted outcome given limited resources.  

Popular BI Tools: Salesforce & Tessitura

The most approachable entry point for a nonprofit into the realm of business intelligence is through their Customer Relationship Management (CRM) system. While CRMs primarily serve as a donor management software, many CRM companies are expanding their offerings to include more advanced analytics and business intelligence tools.

Salesforce is one of the leading CRM vendors and earliest adapters of incorporating AI into its platform. In 2016, Salesforce launched its new cutting-edge AI tool, Salesforce Einstein, to enhance its CRM systems. Einstein gathers data on every user action in order to improve its predictive capabilities. This guarantees users with more analytical power as the AI continues to learn. The video below demonstrates the capabilities of this technology and how it can empower organizations to examine their data in new ways.

Einstein Platform Overview

In 2019, Salesforce acquired Tableau, an immensely popular data analytics tool. This acquisition has positive implications within the nonprofit industry. Salesforce has announced plans to merge Tableau’s technology with Einstein’s analytics, releasing sometime in 2021.

Salesforce is not the only CRM in the advanced business intelligence space. In 2019, Tessitura released Tessiture Analytics powered by Sisense. Sisense has vigorously pursed artificial intelligence and machine learning solutions, indicating why Tessitura was eager to employ their expertise. Tessitura analytics enables users to perform forecasting by using simple trendline analysis. However, Tessitura V16, a new version not yet released, will introduce true predictive forecasting capabilities. This new update will allow the user to assign weights and balances of analytical models. As a result, users adjust their models, or simply specify anomalous time periods to ignore, such as Covid-19. This version will also introduce new machine learning models that will anticipate user inquiries regarding changes in specified KPIs or trends, providing insight on the trend change in question.

Ramifications for Nonprofits

As new technology is introduced to CRM platforms, opportunities for robust BI emerges. Yet, most nonprofits are historically under-resourced and overworked. Thus, training and opportunities for automation will be necessary. Penultimately, those with adept use of BI via powerful CRMs will help nonprofits maximize their capabilities and use of resources, while more effectively fulfilling their mission.

Since BI tools and CRM systems are often not used efficiently in nonprofits, analytics are typically performed manually. Nonprofit data analytics often takes form in examining donor records and basing analysis on past and personal experiences – this information is then leveraged to make informed decisions. In looking at this process in an analytic lens, employees are manually sorting data, determining its importance, and testing these findings. However, AI can conduct these analyses faster, more accurately, and more affordably. As expected with technology, these advanced BI AI can sort through prohibitive amounts of data in seconds, and create more powerful and predictive models than any human could.

Since nonprofits are mission-centric, they often spend less time establishing a strategy for competitive advantage, instead focusing on mission impact. Despite this, competition for charitable dollars is at an all-time high. In 2019, 71% of all donations made to nonprofits came from individual donors. While the overall number of donors is stagnant, the average gift is trending lower. With less money circulating in the giving sphere overall, competition to earn these dollars increases. Fortunately, advanced predictive business intelligence can help solve this critical issue by increasing the number of donors and increasing the average donation amount.

In a recent study conducted by Ghent University in Belgium, researchers explored the possibility of harnessing the power of social media in CRMs. They built a predictive model to determine if it was possible to predict donation behavior based solely on Facebook data. The results strongly supported that in using a linear regression model, researchers could accurately identify a group of potential donors (who are likely to like their page on Facebook) and predict which were most likely to become actual donors. Traditionally, donor acquisition is expensive. Nonprofits often purchase new databases or mailing lists from third parties, investing a great deal of time and money in querying this new data to identify potential donors. Once found, more resources are focused into soliciting gifts from these individuals. This study exemplified that it is possible to build an accurate acquisition model using only Facebook data, saving considerable amounts of time and money.

Software is Only as Good as its User

Although advanced business intelligence tools and applications are promising, these systems still contain flaws.  While a common adage is “data is only as good as the software,” there is a truth that the software is only as good as its user. In “The State of Data in the Nonprofit Sector,” 5% of survey respondents said they took full advantage of data collected by software tools, and only 6% said they felt they were effectively using available data at their disposal. This is an indication that the problem is not with the systems that nonprofits possess, but that there is probably a lack of meaningful employee training in using these tools. Inadequate training in CRMs results in the inability of their correct and efficient use – nonprofits are not using these systems to their full potential.

The model built by Belgian researchers required an incredible amount of personalization and domain knowledge that is far beyond the current capabilities of publicly available CRM tools. However, it provides exciting insight on the CRMs trajectory and the possible capabilities of advanced business intelligence in the nonprofit sector. As advanced analytics become more accessible in CRMs, nonprofit organizations will be able to optimize better data-conscious decisions. 

The caveat, however, is that none of these solutions are free. These CRMs and BI tools can be prohibitively expensive; even the most lucrative nonprofits are unable to hire in-house data scientists to build personalized models. As this software becomes accessible to a wider user base, it could cause greater inequities within the nonprofit sector. Organizations that can afford to use these BI tools will severely outperform their peers and dominate the industry. Ironically, the nonprofits with the most limited resources, and thus the greater need for innovative and efficient solutions such as BI, will most likely be the last ones to access these tools.

Conclusion

Data collection, while not a new concept, is often difficult to apply in nonprofit organizations. Through the use of business intelligence (BI), nonprofits can leverage the tools and data they already possess to maximize its analysis and implementation. With the rapid advancement of modern BI tools, including more sophisticated AI and machine learning models, nonprofits can better use them to predict industry trends, strategize more effectively, and acquire donors. As more user-friendly technology are applied in CRM systems, the obstacles of domain knowledge are diminishing. This results in a positive implication of what these systems can do for nonprofits and how it can transform how these organizations are run. However, the lack of budget and training resources makes the adoption of these systems unlikely for many, particularly smaller, nonprofit organizations.