Take it from Jennifer Pahlka, the Founder and Executive Director of Code for America: “Our ability to do great things with data will make a real difference in every aspect of our lives.”
Big Data is revolutionizing the way we live, communicate and do business. However, harnessing the full power of Big Data has only begun to become democratized. Though data has become cheaper to produce and thus more readily available, it can be costly to hire the right minds to analyze the wealth of data that exists. Of course, the alternative — ignoring the data — can prove even more costly.
By adopting a data culture, nonprofits can compete in this landscape and make more out of limited resources. At Whole Whale, we boil data culture-building down to three main principles:
- Abolish data fiefdoms
- People are 50%, process is 30%, product is 20%
- To create buy-in: Set the vision, cultivate participation, and build a prototype
Learn more about each component below with our 7 principles of a nonprofit data culture — and why building one matters for your organization.
1. Avoid data fiefdoms
You probably remember the word “fiefdom” from 5th grade history class. For the non-history buffs, a fief was a large piece of land owned by the wealthy medieval landowners when feudalism flourished throughout Europe (thanks, Henry II!). In a nutshell, a bunch of higher-ups had access to large chunks of land; the peasants had to work to receive a small part of it. Each fief essentially functioned as an island of its own, making central management nearly impossible.
In the 21st Century, this feudal behavior can frequently be seen within nonprofits that silo access and understanding of their data. The motivators of this behavior are often related to fears around misuse of data (especially in the age of GDPR), data sins made by staff, and fear that data gatekeeper’s job might be in danger if they’re no longer needed.
5 Signs your company may have a Data Fiefdom
- You have a CIO/Senior Technologist that restricts access to data and statistics with the team
- Your most important metrics are only contained in the computer of the data analyst
- Your organization consists of a bunch of disparate departments that rarely (if ever) communicate data insights
- You’re reluctant to train employees on slightly advanced operations related to data
- Metric-backed goals are siloed in departments without relation to the whole organization
The problem is that data fiefdoms can cause bottlenecks in the process of accessing or storing data for the entire organization. In an ideal process, every member of the team can access the relevant data they need. The flow of data does not stop at the data analyst, but reaches members at all levels of the organization.
2. Data culture requires total buy-in (aka, People are 50%)
Data culture depends on a collective buy-in from staff at all levels to measure outcomes, act based on available data, and build on existing knowledge over time. In fact, people are 50% of the equation when it comes to to building and sustaining a data culture. Employees at all levels must recognize the importance of embracing data and using an analytic approach to decision-making. CEOs and Executive Directors (aka the HiPPO) must lead by example through showing that they use data — and not simply rely on experience or instinct — in shaping strategy. (See below for a quick note on the HiPPO.)
The highest paid person in your organization is typically the one who makes the final decision. In doing so, a HiPPO may be reluctant to let data take precedence over their experience, or allow numbers go against their instincts. The data may even challenge what they aim to do. It is critical that the HiPPO knows how to operate within a data culture and still be able to provide leadership and vision. When JFK delivered his Moon Speech to Congress in 1961, he provided the vision and motivation of landing a man on the moon — not the engineering details necessary to do it.
However, change also needs to include buy-in from managers, coordinators, and directors who can articulate why analytical tracking should be part of employees’ personal goals, development plans, and strategic thinking. At every level, colleagues in various positions will need to participate in best practices, training, and tracking to make the shift successful and make data part of your overall culture versus another chore that you have to do.
This doesn’t mean that your graphic designer has to become a statistician, but it does mean that team members won’t balk at numbers or metrics. They don’t tune out when the data analyst in the room makes her presentation. Instead, they embrace the data and use it to inform decision making and challenge conventional norms.
Questions to ask at the People stage
- What is the staff structure as it relates to data reporting?
- Do staff members have the training they need to understand relevant data?
- Do staff members understand how to glean insights and actionable steps from data?
- Do staff members have good working relationships with data analysts?
3. Set the vision for your data goals
In tandem with building the people component of your data culture is the process of setting the vision.What is your company mission and what is the long-term impact of using digital analytics as it pertains to that mission? How can learning about your performance and modifying it help your company to realize its long-term vision?
Creating an overlap of between your goals and the organizational change you are trying to implement is key in creating successful buy-in. At a basic level, this stems from the idea of making a compelling case for change that fits with the already existing organizational objectives. Don’t create all new goals, and be sure to explain how the change in methods will contribute towards the objectives and values your company has always had.
For example, knowing how a new donor heard about your website can help you to target outreach towards those sources to increase future fundraising goals. Looking for new newsletter signups? You can increase that number exponentially if you know where previous conversions came from.
4. Cultivate data culture participation
As noted above, culture change within an organization stems from successful participation at every staff level. The top management of your organization help to provide a role model by setting a meaningful example of how the rest of the staff should follow and embrace change. Hand-in-hand with setting a vision is cultivating participation.
You should appoint at least one data analyst on your team (if you don’t have a dedicated role for that already). Ask her to host weekly office hours with the team, where anyone in the organization can get consultations for anything data-related. At the very least, the data analyst should have good working relationships with members of an organization, and not simply be a number crunching hermit.
One way to foster this begins with open discussions for staff to ask questions and arranging surveys to learn about attitudes and preexisting skill levels. Turn these information-gathering sessions into staff-wide training and dedicated time to continue learning and building capacity.
5. Operationalize participation with process (aka, Process is 30%)
Process accounts for 30% of a strong data culture. The scientific method has been used for over 1,000 years as a way to build knowledge through a consistent iterative process. In a labor market where the average employee spends less than 5 years at a job, it’s important to have strong structures in place to institutionalize data knowledge. Data insights and findings from past years should be maintained for posterity to inform later employees. Maintaining records of prior decision-making also saves time when similar issues arise in the future.
That said, note that process follows people in this equation. In other words, people at all levels should know how to interact with and assess incoming data to ensure that all relevant metrics are being accounted for. Remember: The #1 killer of good data structures are Data Fiefdoms. Organizations cannot afford to be bottlenecked by having a limited supply of data analysts.
Questions to ask at the Process stage
- Are staff accessing and communicating data across teams well?
- Do staff act on data or regularly share learnings from experiments?
- Are goals set in a way that can be tracked through metrics?
- Does the organization use a Gather<Analyze<Insight method?
- How often do staff receive data feedback?
6. Build a prototype — even if it doesn’t work (especially if it doesn’t work)
It’s one thing for you to articulate why building a data culture is important, and it’s another to show your colleagues why. Running a data insight report or building a dashboard in Google Analytics after the first few weeks or months of monitoring your change in behavior can demonstrate to your colleagues concretely why it is so important to monitor and modify web analytics as part of your organizational functioning.
This is called building a prototype. And don’t be discouraged if your first round doesn’t work. In fact, embrace the failure. In a data culture, it’s important for the team to adopt a Build-Measure-Learn philosophy as described by Dan Ries in The Lean Start-Up. Team members should be asking themselves, “How can I measure the success of this product/campaign/program?” They should know not only how to seek out relevant metrics, but also appreciate why doing so is an effective strategy. And, most importantly, they should understand how to learn from the data. This entails utilizing the feedback provided by the data to make necessary changes and inform decisions.
When your colleagues can see the results of their hard work and time, it’s easier to convince them to keep putting in the effort. Reports and monitoring, especially at the earlier stages, is important to getting over the initial buy-in hurdle.
7. Use the right tools for the right effect (aka, Product is 20%)
You can’t fly astronauts to the moon without a good spaceship. In the same light, there are numerous tools and services that can help support a data culture and embody the build-measure-learn approach. Fortunately, many are simple to master and do not require an advanced degrees in statistics. These tools allow people, in all departments, to incorporate an analytic approach to their work.
Mailchimp, for instance, is an email service provider that your marketing team can use to test out different subject lines with your audience, and see which one leads to the highest open rate. Very intuitive. Very fun. A similar tool that can be used by web designers and non-designers alike is Google Optimize. This service allows you to experiment with different content and layouts on your web pages and compare how they perform. You pick the metrics that are important to you, and the site will help show you how to optimize your page. Effectively, it replaces guesswork with data.
Questions to ask at the Product stage
- Are tools in place to analyze large data sets (beyond Excel)?
- Are consistent naming and storage conventions in place across databases?
- Are dashboards and metrics updated as automatically as possible?
- Is data stored in a way that reporting can be done across the organization?
- Are semi-annual security audits and passwords changed?
Like Rome, data culture isn’t built in a day. It is a gradual process that calls for changes in habits, attitudes, staff, and even resources. But once it’s established, it becomes a lot easier to sustain and grow your nonprofit’s true impact. While each of the above components is individually necessary, like Captain Planet, their powers combined are far more powerful. Plus it might even let you pull off green hair and red tights like a boss.