Data Maturity Part I
I was introduced to the term data maturity by Karen Graham, Maddie Grant, and Jenn Taylor at their 2019 GoodTech Fest presentation. (If you dont know about GoodTech Fest, check out my blog post on it here and consider going in 2020! Its a great, reality-based conference about measuring what matters in the nonprofit sector.)
Ive been thinking a lot about this idea of data maturity, and how we can help organizations build capacity to provide data for themselves and to funders. I think about this all the time, of course, but data maturity provides another, very useful concept to hang some ideas on. It got me thinking about why some of our clients seem able to collect and use data better than others, and also about what it takes for an organization to really be capable in this way, both from a data perspective and a technology perspective.
Worth considering here is the continuing trend of funders (among others) pushing programs to collect more and more comprehensive data, especially related to outcomes and hypothesis testing. The problem is, there have not been parallel efforts for ensuring that organizations can collect and analyze what they need to for the purposes of even basic performance measurement and accurate collection of admin data, much less high quality outcome data.
Organizations must have certain structures, practices, and tools in place in order to be able to collect and use data, internally and externally. In the business world this is a given. Of course you need computers, software, expertise, and skilled staff to track how many customers you have, what they bought, what your costs and profit margin are, where your products are selling, and what your customers think of your products. Yet in the nonprofit sector, theres an irrational reluctance to fund the kinds of things that are the backbone of any high functioning organization, and donors prefer that their funding is going directly to services.
Given that a blog post cant adequately tackle why many funders dont want to fund backbone functions to deal with data and analysis (and if you find a blog post that does, please send it to me!), lets try to address what you need – in terms of know-how, tools, and policies and practices– to be a data savvy, data mature, and generally kick-ass, **DATA MATURE **organization!
What Youll Need to HAVE to Use Data for Internal and External Decision Making (and, ultimately, Hypothesis Testing)
- **A clear data model. **This means knowing what you are collecting, in what form, how it intersects with other data that you need, and (this is so important) what you are using it for. The data model describes how data elements relate to each other, where they come from, and what purpose they serve.
- Good data hygiene. So, once you know what you are collecting and why and where it comes from as part of an overall data strategy you will also need to have clean, complete data. We put a public health spin on this. It sounds basic, but after organizations that dont have a clear data model, the most common thing we see is a lack of good hygiene. To clean up your data act, you need clear organizational roles about who collects data and clear protocols as to how its done. And you need data quality checks, meaning someone needs to look at the data on a regular basis to see what is getting missed and to address it. (We will do another post later this year just on data hygiene. Keep an eye out for it.)
What Youll Need to KNOW to Use Data for Internal and External Decision Making:
There are two types of expertise needed to have a successful data strategy.
- Research and evaluation expertise: The capacity to understand and implement the right performance measurement data collection strategies. This involves knowing how the research literature connects to your program activities and outcomes, understanding the development of surveys and other data collection instruments, and the analysis and reporting of findings. Good research and evaluation staff (whether internal or external) tend to be a mix of data nerd and business consultant.
- Data systems (technology) expertise: What exactly is needed depends on the type of data systems and technology you are using but, generally, staff or vendors need to have expertise around how to collect data from a variety of sources, how to clean and aggregate it, how to store and transmit it (safely), and how to warehouse and integrate it. Knowledge of various reporting libraries and software is a plus. Of course they also need to be experts at whatever software you are specifically using, whether it is Excel or Python.
**What happens when you have one but not the other: **
- When an organization has the evaluation expertise in house but inappropriate or insufficient technology, they collect the right information at the right time from the right sources, but they lack the ability to effectively and efficiently clean, aggregate, analyze and report out that information. They also wont have the capacity to integrate data from partners or effectively warehouse and integrate data from multiple sources internally.
- When the organization has invested in technology and data systems, but doesnt have the capacity around expertise and knowledge management (evaluation staff or consulting partners) the organization may invest a lot of time and money into managing the system, but find that they dont have the right data collected at the right time to answer key questions.
Thats all for now! Stay tuned for Part II of our exploration of data maturity, which will further your understanding of data maturity stages and how your organization can become more data mature in order to achieve your goals.