Equitable Data Collection: Data Stewardship and Analysis

This is the final part of our series about equitable data collection. Previously we covered the foundations of equitable data collection, how to design for your audience, and the best ways to ask for demographic details. In this final part, we will dig into the best ways to include data in your Salesforce CRM so that you can learn and act on the data you’ve collected while protecting and respecting the privacy of your constituents.

Once you’ve collected data, the final piece of the puzzle is how to structure, store, and protect this data in your CRM so that it is both useful for informing your work and secure. Key considerations: where is the information stored, how long do you plan to retain it, and who has access. The more sensitive the information is, the more steps you’ll need to take to protect privacy within your system.

Privacy and Security

How you store information should be driven by its purpose. If you are using only this data in 1:1 interactions, it likely makes sense to store this information directly on or related to the Contact Object. If, on the other hand, you are using this data in aggregate and for analysis to answer questions about trends and patterns, you could keep the most sensitive data you gather from individuals separate from Contacts in your system, reducing the risk of that information negatively impacting them.

Information should only be accessible to those in your organization who need to view it. Just as you only want to collect data for a clear program purpose, you should only give access for a clear program purpose, and only keep the data as long as it will be useful. In addition to setting up field-level security to limit who has read/edit access, you can use built-in tools like data classification metadata and field encryption to further label and protect sensitive information. If visibility in aggregate is all that certain staff need, consider creating dashboards with a specific running user to summarize information for staff while restricting their access to individual records to keep individual points of data private.

Structuring for Analysis

If you plan to do analysis within Salesforce, it’s important to keep data points in fields that work in the ways you need them. Generally, the field type should match the question format you used when collecting data. Getting the data from an external tool into Salesforce may require some processing before it can be imported.

When collecting demographic information like race, gender, and sexuality, multi-select options are recommended as a best practice. However, using multi-select picklists, the corresponding equivalent within Salesforce, can make importing, reporting, and processing of that data difficult. This is often a necessary evil, and one great workaround is using checkbox formulas that translate the selection of individual options so that you can easily report on totals. This works best if you only have a handful of reporting categories that are relatively static, so you don’t have to create and maintain a huge number of checkbox formula fields.

Another alternative is building a junction or child object related to a contact or form submission record. You can then store each selected choice as a separate record for easier reporting, but extending your data model like this can be more effort than it’s worth.

Here are some considerations for using different types of fields in Salesforce:

Analyzing Data

A full review of how to do equitable data analysis is outside the scope of this series, but there are great resources on keeping equity at the forefront of your analysis and data visualizations such as the Do No Harm Guide, by the Urban Institute.

Case Study: Hiring at Percolator

In 2021, the Diversity, Equity, and Inclusion (DEI) Committee at Percolator took on the work of developing a transparent policy and process for staff recruitment and hiring to promote diversity. A driving goal for our updated recruitment process was to ensure that we were reaching and attracting a balanced pool of diverse applicants, including those who are traditionally underrepresented and/or historically marginalized.

We did this in two steps. First, we compiled a list of job boards that targeted underrepresented communities in the technology sector. Second, we developed a confidential survey that applicants can fill out after applying for a position to help us evaluate, within each hiring cycle, if we are reaching our goals for a diverse pool of applicants.

It was important to the Percolator DEI committee that we collect and evaluate the data in this survey in a way that protects the anonymity of those who have responded. Even though the survey is anonymous, we knew that if we collected granular demographic information, or reviewed individual survey submissions individually, it might be possible to accidentally identify an applicant.

Our approach for this is three-fold. First, we ask about demographics in umbrella categories so it’s easy for applicants to respond, and so we only get as much information about applicants’ identities as we need to assess the diversity of our hiring pools.

Second, we decided to host our survey in a Salesforce flow, and to process submissions into summary records (rather than individual submissions) so that accidental re-identification would be difficult.

Third, we review the different categories of information gathered in this survey separately, only review the data in aggregate, and only when we need the information to make decisions (i.e. we don’t monitor the data in real-time) so that we can make informed decisions about our hiring processes, and can ensure privacy and anonymity to our job applicants.

In this way, using tools like Salesforce flow to post-process submissions can give you flexibility in how you store and report on demographic information in ways that protect your respondents and still provide your organization with the data you need to make informed decisions. If your organization can get clarification on the exact questions you need answers to, you can work backward to best design a process that builds equity into every step.

Conclusion

Equitable data collection requires intentionality, experimentation, and openness to feedback as you find the best way to ask for, interpret, and protect information about your constituents. We hope this this guide helps you make your data collection process more purposeful, responsible, transparent, and grounded. When done well, your data collection reflects the values of your organization and treats all your audiences with respect.

Highlighted Resources

• DO NO HARM GUIDE | Urban Institute

• APRA ETHICS AND COMPLIANCE COMMITTEE DIVERSITY, EQUITY AND INCLUSION (DEI) DATA GUIDE