AltruNext Associate, Diana Shea, was recently featured as a guest author with tips about donor data migration. The full article was originally published on Foundant.com and reprinted here with their permission.
Congratulations! You’ve positioned your organization for development success by investing in donor management software. Now what?! The software selection process is usually full of hope and promise of things to come. There’s excitement in the air as your board and staff feel eager to get started. With the team poised for greatness, there’s only one thing standing between you and implementation: data migration.
In contrast to the selection process, migrating data from your current system(s) to a new system can feel daunting. How should your organization best prepare your donor data to move into the new system? What are the tried-and-true strategies for nonprofits to handle this overwhelming process? Our responses to these frequently asked questions will address your concerns head-on and help you formulate a plan to move forward. By choosing to get your data organized and leverage the support of a donor management system, you’ve positioned your organization to expand its capacity and impact.
1. Do all historical donor records need to migrate to the new system?
This is a great question! If the pricing tiers do not depend on the number of records in the database, there’s no limit to the number of records you can migrate. Does this mean that you should take all of them? Maybe. Much of the data from your current system provides valuable historical information that will be lost if not brought into the new system. Analyzing historic donor behavior can help to shape future strategic initiatives. Provided that storage space is not an issue, we recommend bringing all historical records, including donor data from deceased persons and lapsed donors. If you are able to segment these inactive records out of relevant reports and dashboards, there’s no harm in keeping them, and they may even serve you later on. Is there any reason NOT to migrate a record? Sure. If you have the time and resources to analyze your donor data, you might find some records that you don’t want to bring over. Reasons for not migrating certain donor records may include incomplete or incorrect data in critical fields that render the record useless.
2. What important data cleansing activities should be prioritized prior to loading donor data into a new system?
There are two types of data cleansing activities: data formatting and data quality. Data formatting cleanup activities are relevant for any software migration project. Examples of this type of data cleansing include standardizing addresses and phone numbers (e.g., two-letter codes for state abbreviations versus shortened abbreviations of the state name with a period, hyphens in phone numbers versus parentheses and a hyphen).
For data formatting, consistency is king! There’s no right answer. These standards are more about style and readability in various reporting forms than about substance. Which fields need greater formatting consistency for your organization? Review the existing data by sorting and scanning to see trends and identify relevant cleanup needs.
By contrast, the data quality cleanup activities are highly contextual. From a development perspective, building relationships with constituents is critical to an organization’s sustainability and impact. First and foremost, this requires good contact and communications records. In particular, it’s a good idea to spend time reviewing and updating relationship codes and household records, checking for deceased records, and updating mailing lists or flags accordingly. Here’s a short list of specific data items to review and address as time allows:
Bounced emails: Which emails have bounced from your email marketing service? If the email marketing service is external to your current contact database, consider how to update the records by removing the old email addresses (or indicating that they are old) in the contact records.
Missing addresses: How many email addresses are missing? If the number is too large to tackle, segment the list by focusing first on major donors and other strategic individuals. (This assumes that your organization uses email communications as a primary contact method.)
Status change codes/flags: Review mailing and other status change flags or codes. What is your organization’s process for updating records when you learn that a person is deceased? Or changed addresses? Or changed households? Review the relevant codes/flags to ensure that all are aligned with other contextual information available in the record.
Remember, any upfront effort to clean the data will help build users’ trust and confidence in the new system and result in higher quality reporting.
3. Does all of the data for each donor record move into the new system? How does an organization decide what data to keep and what to leave behind for an individual donor?
The data cleansing exercises referenced above will help identify the records to load into the new system. With this list in hand, the next question is: should all data fields come along? Most of it will, and some might not be necessary. Data tracked historically may no longer be needed. A field-by-field analysis can determine the value of the data and where it should be stored in the new system.
To conduct field-level analysis, first, create a list of all donor fields. Next, go through each field and answer the following questions:
What is the purpose of the information stored in each field?
Is the information required, or will it be helpful going forward? What is the value of the information?
Will the information continue to be tracked for new records created in the new system? If not, then the data will likely become less useful.
Is there a place for it in the new system?
If yes, what is the field name it will map to in the new system?
If no, is there value in it? Do you want to align with the way this data was stored in the past or start anew? Do you need to create a custom field for it? Will new donor records be expected to track this data?
4. Should the donor data be updated before migration?
It’s not mandatory to update all records before migration. Life happens. People change jobs, get married, change addresses, and even get new phone numbers. Some portion of your data will likewise always be outdated at any given point in time. However, incorporating data hygiene practices into your data management processes on an ongoing basis will reduce the number of records with stale information, thereby improving the quality of your data overall.
With that said, is this a good time to undertake a data cleanup project? Could be. Since you’ll be translating the data from one system to another, it makes sense to spend some extra effort to make sure that it is cleaned up. Plus, the staff will appreciate the software that much more and likely adopt it at higher rates when they experience better reporting with higher quality underlying data. What motivation do they have to learn the new system if the information is not better than before?
We’ve provided specific examples of cleanup activities to consider in the following two questions. The value of each activity must be weighed against the time and effort required and figured into the overall software transition plan. In general, it’s a great idea to clean up the data as much as feasible before loading it into a new system.
5. Does missing data need to be filled in before migrating a donor to a new system?
In some cases, yes. Contact information is critical to the donor relationship-building process. Without either a phone, mailing address, or email address, the development team will face significant hurdles when seeking to build a relationship with a person. However, identifying which pieces of information are important to track down and for whom is a customized exercise. Some questions to help guide you:
What are your organization’s methods of communication?
Is one form of contact information more critical than another?
What information do you need to build a relationship with someone?
Run reports to find out how many records are without crucial data as defined by the questions above (e.g., How many records don’t have email addresses? How many records don’t have mailing addresses? How many are missing both?).
Once you have a list of records to analyze, look for other key indicators. Do you want to fill in the gaps for any of these records? This is a judgment call about how much value is in the records in question. If there is very little other information available in the system, you may choose to delete records with limited/no contact information and no donation history. Create your own definition of what situation warrants further investigation and what type of missing data warrants removing the record. Here are some examples of questions that might be of interest when identifying important donor records despite lack of contact information:
Has the constituent made a donation in the last five years?
What is the constituent’s total lifetime giving?
How often does the constituent make a gift?
Are there any notes from previous conversations with this person in the system?
How are they connected to the organization?
6. What are the best data formatting standards in a new system?
Data formatting standards create order out of chaos. They provide consistency in a sea of unchartered waters. What is a data formatting standard? Simply put, it’s a formatting rule for a specific field or set of fields. For example, an email address should always have an @ symbol and end with a common extension (.com, .org, .edu, .net, …). Simple rules such as this streamline the data entry process and increase the likelihood that a user will understand system-generated reports.
In some cases, the software does the work for you by restricting fields to allow specific characters, pre-set values, or character patterns. Restricted fields automatically enforce data entry rules (email address fields that require an @ symbol or phone numbers that only accept numbers).
In other cases, fields allow free form text entry (category labels, codes, tags, or other user-defined custom fields). These situations require standardization to rein in the possible formatting variance. In these cases, think about how data is used and in what context. Not every field will need a formatting standard, but many do. Use common sense to evaluate which fields would benefit from a formatting rule and then consider the following questions to determine which formatting standards to use:
How is this field used?
What reports use this field?
What are some downstream effects of this field’s formatting?
Do mailing labels use this field?
Do email newsletter merge tags use this field?
Do thank you letters, appeals, or other automatically generated mailed letters use this field?
The most important rule of thumb: consistency! Making a rule, documenting it, and sticking to it will ensure easily digestible database reporting. Also, be sure to enlist the support and advice from those using the data when making these decisions. Hearing from multiple users will often provide perspectives that might have otherwise been overlooked.