A crisis in data leadership

Introduction

At the heart of every nonprofit organization is its “mission” — some definition of impact that serves as the reason for the organization’s existence. Increasingly, our mission-driven work is playing out digitally as we educate through websites, advocate through social media, and otherwise engage online. As a result, our measurement and evaluation of mission success similarly rely more than ever on data from these digital tools.

For most organizations, the heart of their outreach is their website. Data collected from the website can be used to measure audience engagement, assess the success of key outreach campaigns, and learn more about the needs of constituents. To do this, the vast majority of nonprofit organizations turn to the free, flexible Google Analytics. The flexibility and ease of setting up Google Analytics, however, can just easily result in problems.

Earlier this year, we conducted an original study of more than 1,500 nonprofit websites to check for visible trends and common mistakes in how they’ve implemented their analytics. Among the things we learned: As of today, nearly 1-in-10 nonprofit websites are double-counting their Google Analytics page views. The presence of double counting is a detriment to data reliability, and we’ve written about the downstream effects on measures of reach and engagement, and steps to diagnose and remediate.

The real problem revealed by this, though, is the dire state of data as a neglected resource throughout the sector. The fact that double-counting occurs so often in the first place suggests underinvestment in data strategy and management, while the fact that it persists and can even linger for years at a time speaks to underutilization of data as a strategic resource. Many organizations check the box (“Yes, we have data”), but aren’t making enough use of it to discover fundamental problems with data accuracy, much less harnessing data to continuously optimize content and outreach strategies.

Moreover, high-profile data lapses are heightening global concern about how carefully organizations collect and manage their data. The inattention highlighted by our new findings can result in potential security and legal risks — if so much double-counting can go undetected, to what extent are other faults, such as the unintentional collection of personal information, also being neglected?

Why is this happening?

To understand why this is a leadership problem, we first have to understand why it happens. Most often, double-counting page views begins during a website redesign or marketing campaign. Some change is being made, and a well-intentioned staff member adds (or “moves”) analytics tracking in the process without checking to see if it is already or also configured elsewhere on the site.

In these cases, the team is tasked with building a website or adding a new capability, and while they may be told “turn on analytics,” the details of what’s expected and the resulting data are neither seen as a part of that process nor as a part of the budget (particularly when external vendors are involved). As a result, data becomes a casualty of the change in an accident of inattention and a process that treats analytics as an add-on rather than a key requirement.

Failing prevention, organizations still have a chance to correct the error. Each organization that began double-counting likely passed through a period where their data showed a sudden, dramatic, and wildly unnatural change in their bounce rate (the percentage of visitors who navigate away from the site after viewing only one page). Signs like this quickly lead many motivated analysts to discover the truth and correct the issue.

In many cases, however, the data is not sufficiently questioned if it is reviewed at all. The person or team with responsibility for data assumes that it is merely the result of changes to the website and accepts the false data as a new baseline.

As you can see, there are a number of things that have to go wrong for an organization to start double-counting and then to continue doing so. It’s actually a bit of bad luck to end up in this state, and many organizations with the same gaps in data strategy and operations end up avoiding this situation altogether through unrelated choices or vagaries of the development process.

It’s not enough to have data

One of the downsides of how easy it is to collect data is that it makes it just as easy to declare analytics “accomplished” without really understanding what you’ve collected or putting it to use.

Very few nonprofits can afford to hire dedicated analytics staff, and as a result, it becomes a “collective” responsibility to use data to better the organization. Because of this, it ranks relatively low among each employee’s priorities and quickly becomes an abundant but neglected resource—organizations have reams of data, but aren’t doing anything of value with it.

Yet when it’s made a high priority, nonprofit staff can use data to:

  • Support expert intuition and accelerate decisions about outreach and publishing based on past performance
  • Encourage collaborative discussions across teams to celebrate and learn from successful strategies
  • Correct the record about audiences and content, using data as a myth buster to free the team from misapprehensions
  • Settle internal disputes about strategy by providing a single source of truth about what is effective
  • Begin with automated marketing by using data to segment audiences and design experiences for each of them

There are so many things you can do with data, and pursuing one or more of those uses bring direct benefits to your organization, while indirectly contributing to a culture of data that improves the quality of work overall and helps prevent major issues like double-counted data.

Money and a mandate

As a sector, we need to confront the fact that thousands of organizations can have this significant of a failure of their data without anyone noticing. It’s up to leadership and the funding community to help organizations improve and make mistakes like this increasingly rare.

Budget for data

The most direct thing you can do is direct funds toward increasing your capabilities in data — most commonly through technology that makes it easier to surface insights and staff time to extract and act upon those insights. Dashboards, training, upgrades to data that make it more relevant to your strategy — all come with costs, and without budget allocated towards them, they will forever be activities for “next year.”

Whether you are a philanthropist writing a grant or an executive writing a budget, you can make it easy for your team by adding a line item for analytics. Just as good, or even better, is to add more detailed outcomes expected of analytics to existing projects, such as a website redesign or campaign plan.

Encourage a culture of data

Most nonprofits have a central set of mission-driven “goals,” and many take the extra step of tying data to those goals. Where many fall short is exercising those metrics and creating a habit of internal measurement and evaluation, instead waiting until grant review time to pull together the numbers, often when it’s too late to act on them. Creating this expectation of nearly real-time analysis will help you catch problems with your data, not to mention optimizing your outreach.

The tone of this mandate is critical, however. Building a culture of data means creating a professional environment where people are encouraged to ask questions and ask for evidence, while simultaneously making it okay to say “I don’t know, and I need help finding out.” A lot of paralysis in analytics stems from fear—the fear of asking silly questions or being resented for them by people who fear the responsibility of answering them. Even when analyses end with no finding, the mindset and process of conducting them often reveal unexpected insights.

Recognize the challenges of working with data by celebrating the unanswerable and the discovery of problems. Make it safe to report failures in data, or they’ll never get fixed.

How funders can support data strategy & management

If you work for a grant-making organization, it’s possible you already connect grantees with valuable services to inform strategic planning or fundraising. Yet these essential areas rely on a foundation of reliable data. In addition, funders themselves depend on accurate data from nonprofits in order to make wise philanthropic investments.

Given the sector’s struggle in embracing ideal data practices, there’s a strategic opportunity for grant-makers to build capacity that empowers organizations to make necessary changes to their data practices and culture. When funders connect grantees with relevant technology, training, and consulting services, nonprofits can learn how to gather and use timely, actionable data to improve their effectiveness. As an example, one potentially game-changing application for many nonprofits is online fundraising from individual donors. Robust data systems and practices can optimize outreach and maximize giving behaviors by strengthening long-term relationship building with top donors. In addition, enhanced digital data helps nonprofits do other important things like tailoring their websites to meet the needs of constituents, strengthen their digital advocacy, and feed into data dashboards that empower executive decision-making.

Conclusion

When data is rendered relevant and actionable within organizations, it tends to get used. And when it’s used, there are strong incentives to ensure it stays accurate. The trick to preventing future issues like double counting, then, is for leaders to integrate data into ongoing decision processes. When this happens, organizations will have sufficiently robust internal systems, accessible expertise, and – perhaps most important – many people asking smart questions of their data on an ongoing basis. The latter is the surest way for unexpected data issues to be identified and addressed quickly in the future.

Stefan Byrd-Krueger
Chief Analytics Officer
ParsonsTKO
Stefan Byrd-Krueger is the Chief Analytics Officer of ParsonsTKO and leads the company’s data strategy practice. He lives in Silicon Valley, blending the area’s penchant for innovation in technology with more than 10 years of experience supporting digital projects in the non-profit and think-tank community. His experience includes managing a multi-disciplinary digital and grassroots outreach portfolio at a small nonprofit and serving as the first dedicated analytics specialist at The Pew Charitable Trusts.