If you’re like most associations and non-profits, you probably have tons of data tucked away in disparate, cobbled-together systems. It may be old and unreliable. Or, maybe you're not too sure how and where to access the information you need.
Regardless of the dimensions of data quality, you cannot let these issues fall to the wayside. It may be easy to dismiss poor-quality data as an "IT struggle.” But the truth is, data quality is an invaluable organizational asset in today's data-driven, digital economy.
In the association and nonprofit industry, we typically assess data quality across 10 dimensions: confidence, importance, clarity, accuracy, currency, completeness, hygiene, availability, entry quality, and uniqueness.
|Dimension||How It’s Measured|
|Confidence||How confident are you with the quality of data found in this application?|
|Importance||How important is it to business results?|
|Clarity||How satisfied are you that it’s clear, easy to understand, and interpret?|
|Accuracy||How satisfied are you that it’s correct and reflects reality?|
|Currency||How satisfied are you that it’s up-to-date and not outdated?|
|Completeness||How satisfied are you that all required records and values are captured?|
|Hygiene||How satisfied are you that time-dependent activities (e.g. tasks, events) contain clean, quality data?|
|Availability||How satisfied are you that information is available when you need it (and with the appropriate security and access levels)?|
|Entry Quality||How satisfied are you that data is easy to enter?|
|Uniqueness||How satisfied are you that each data element is captured in one spot (not duplicated across multiple fields)?|
These dimensions are really both concepts and metrics.
These concepts help you assess business confidence in data quality, ensuring stakeholder needs are met and decisions can be made with quality data.
These metrics are calculated across departments and functional areas as well as against benchmarking data. You can see how you measure up to get a sense of what’s working and quick-win opportunities by system/application, department, and business leader.
How confident are you in the quality of your information? If people don't have confidence in the quality of your data, then data doesn't get used in decision-making. Instead, they go with gut reactions and best guesses.
How important is a piece of data in delivering business results? How much do you need the data to execute tasks and complete a job? If people don't think that data is important, then it impacts practices related to that dataset in terms of proper data governance and data maintenance.
How satisfied are you that your data is clear? How easy is it to understand and interpret?
If you have two different fields with similar names (for instance, education and university) then that information is surely not clear. The data label should clarify the information you're looking for.
So, in this scenario, that may be, the highest level of education versus the actual university name. You want to be sure data users know what each data point represents and that everyone in your organization is in agreement. In other words, everyone understands the same thing.
Does your information truly reflect reality? Or, do you have concerns with stale data?
If people in your industry move jobs every two years, for example, then the employment information should be updated routinely within that timeframe. If it hasn’t been updated, then that field likely doesn't contain the right information. You want to be sure data users are confident that your data depicts reality.
This doesn't mean timeliness/recency is always a factor within this dimension of data quality. It simply means your data should contain the correct information.
How satisfied are you that your information is appropriately up-to-date and not outdated?
Take a data field like a birthday. It doesn’t change. But the data field indicating the member's areas of interest sure does.
A seasoned executive is probably interested in different topics than a new professional just entering the industry. Twenty-five years down the road, that professional is likely in a more senior role and their interest areas have probably transitioned. Their original responses are most likely no longer accurate.
How satisfied are you that all required information is captured and no crucial data is missing? Is your data as comprehensive as you need it to be?
You may be trying to come up with a target market for marketing for young professionals and defining it based on the graduation year collected. Yet if you only have that data point for 20% of records in data set, then you may be missing potential customers due to incomplete data.
How satisfied are you that time-dependent activities (e.g. tasks, cases, jobs) are appropriately managed to maintain data cleanliness and quality?
Or is your database full of garbage? Sometimes data ends up in the wrong field, like having a website address in an email field.
How satisfied are you that your information is easily accessible with the appropriate data security measures and access levels? Are the correct authentication and authorization protocols followed?
The right people should know how to run a report to pull the data they need when they need it.
Permission is another factor. If they don't have the right permissions for files and information they need, this impacts availability.
Donations are a good example. People on the foundation's staff may be able to see donation history, but membership people may not have the right access levels to include this critical data in touchpoints and marketing campaign strategies.
How satisfied are you that your data is easy to enter? How confident are you in how fields are validated when you're entering data? Does each key field only accept the correct type of information? For example, do date fields accept dates and free-form fields accept text fields?
If people type records in all caps when filling out forms, this could likely complicate searches in some systems. Data searches may also be affected by abbreviations in job titles and employer names. Data entry standards used by staff can help maintain consistency.
How satisfied are you that each dataset is unique? Is data captured in one spot? Or, are data elements duplicated across multiple fields?
This dimension of data quality is similar to clarity. However, clarity is about defining fields and uniqueness is about duplication of data. For instance, when you have the same thing in both education and university fields or when you have the same email address in the work and personal email fields.
You want to avoid redundant data storage within a database at all costs! Duplicate records could cause membership teams to waste resources browsing through records and processing transactions. This data quality problem could, in turn, impact the member experience when duplicate records cause login issues.
Duplicated files also cause confusion during updates. If the duplicate is updated instead of the original data asset, this causes significant reporting inconsistencies and search problems.
Read More: Creating a Data Quality Dashboard
Based on survey results, each dimension of data quality is scored. These scores are based on how satisfied survey respondents are with the quality of the information in a particular application or system. The breakdown is as follows:
Your Data Quality Scorecard displays your results at a glance.
Each scorecard looks at one application at a time, across departments and functional areas. Your scores on each dimension are also compared to benchmarks on the same scale. These benchmarks include industry baselines and changes in performance year after year.
You can find a snapshot of your scorecard at the beginning of a full data quality assessment report. It tells you right away the areas that are most important and error-prone. And it gives you a big picture of your training needs and key stakeholders who require better data quality.
Want to see how it works? You can download a sample report here.
Not all data is created equal. Some data is better than others when it comes to making predictions about your association and members. When it comes down to it, quality data is data you can confidently use for forecasting to meet the goals of your association.
Compared to low-quality data, high-quality data is data that’s consistent and unambiguous. On the other hand, poor data quality results in inconsistent and ambiguous data.
Read More: 6 Red Flags of Data Quality Problems
Achieving and maintaining high-quality member data in your AMS software is a critical factor in making faster, better decisions to reach your organizational goals. You can’t reach true data maturity without quality data. In other words, you cannot make strategic business decisions without good data.
It helps you meet performance metrics and increase ROI on your marketing campaigns and membership programs. In fact, organizations that make decisions based on high-quality data are 10% more productive and save 10-20% on vendor costs. Beyond improved efficiency, informed decisions based on high-quality data enable you to identify opportunities and create innovative products and services.
On the other hand, poor quality data has a systemic, organization-wide impact.
Assessing data quality on a regular basis allows you to prioritize departments and individual business leaders who are most reliant on data but least satisfied. Use the results to execute a plan around your immediate data needs, including displaying KPIs to track progress.
Are you ready to uncover the step-by-step process to get where you want to be? Take our Data Quality Assessment for Associations to pinpoint the root causes of your data quality issues and define the strategy that you can follow right away for effective decision-making.