MAP Migration – Data Cleanliness

This is Blog 5 in our series, “Digital Pi Dives Deep Into Marketo Migration”
After such great feedback on our Migration planning blog, we decided to dive deeper into each of the steps first mentioned there. Click on these links to check out all of the blogs in the series:

WEBINAR – Anatomy of a Marketo Migration

Clean data is always important, but it becomes paramount when approaching a migration as you want to only map/transpose and move the very best records.

If you don’t already have a strong data management program in place; you’ll want to start by determining/reviewing deletion rules with your stakeholders. Communicating the plan and getting buy-in is important for the success of this project.

Leads are usually the purview of marketing operations; while sales operations typically have the final say over contacts. Frequently sales operations will agree to delete old/unengaged contacts not associated with an opportunity, a support case, or a customer. If you don’t have a field where you are already tracking the last marketing activity date in Marketo, you can add a field now and populate it as best you can with activity data from Marketo.

Having this data will allow SOPs to check contact activity from both sales and marketing when determining good contact candidates for deletion. Work with SOPs to determine if they will delete the records in their CRM or if you will delete the agreed-upon records from within Marketo.

Within Marketo, we recommend looking at a couple of areas to find records/leads that are candidates for deletion:


If any records are deleted from SFDC, they will not be deleted from Marketo. You’ll need to pull a list of records with “deleted from CRM” set to TRUE and delete these from Marketo. We recommend having this as an ongoing data management program.

HELPFUL TIP: If you are doing a Marketo to Marketo migration; please note, that If you build programs in your old instance, you can import them directly into your new instance. This will allow you to build these programs once, use them to clean up your current instance and then utilize them, go forward, in your new instance.


Do you have records that haven’t been synchronized to your CRM? Depending upon the reason why they haven’t synchronized, you can:

  • Append the necessary data to qualify for CRM synchronization
  • Delete records
  • Migrate records manually to your new instance

In all likelihood, you may want to try a combination of all three. We recommend creating data normalization programs to clean up and/or populate missing data in your key fields.

HELPFUL TIPS: Using a 3rd party tool to append data to your records to further populate key fields is also a best practice. While you are thinking about “MAP only” records; you’ll want to remember your “MAP only” fields/data. IF you are not syncing all of your “marketing management” fields to your CRM, you’ll need to include in your project plan, a section on manually moving this data to your new MAP. For example, a lot of companies keep their GDPR & general compliance tracking data (date/time stamps, source of the consent) in their MAP only. This is important data that shouldn’t be left behind.


Personal data starts to decay quickly. You’ll want to determine a “recency” definition that works well for your sales cycle. With a lot of our clients, we’ve seen them land somewhere in the range of 12 – 18 months.

If you don’t have a “last marketing activity” date stamp field; you’ll want to create a filter that pulls all records that were created prior to this recency definition and who haven’t opened, clicked, visited, filled out the form, have no opportunities, etc. within this timeframe.

With the number of migrations these days, your “created at” date in Marketo may not indicate the “true” date this person entered your company’s systems. Hopefully, you’ve created “original created date” fields to keep track of this information. If not, utilizing the SFDC created date may help.

People who haven’t engaged within this recency definition are good candidates for deletion.

HELPFUL TIP: Utilize a 3rd party vendor to review email addresses for unengaged people. Marketing Suspend anyone that has had their email address identified as invalid. This is another good practice to follow in your new instance as well. It will allow you to identify potential spam traps and keep your deliverability numbers high.  Some 3rd party tools will also indicate when a person has left their company so that you can research where they have moved to and update your records.


Test data should be removed after you are done testing, but this doesn’t always happen. And if your instance has been around a while, these test records can pile up.

We recommend creating a test data management clean-up program to identify all potential test data. You’ll want to delete these records manually,  to ensure that you don’t delete records from potential clients that have “test” in their actual names (first, last, company) and/or domain. For example, “John Durtest” or “” When you find these anomalies, add them to the select list as exclusions, so that they won’t show up on the potential deletion list again.

If you still have test data in your system; be sure to remove it before the migration.


Not everyone who fills out your forms and/or provides their information does so with correct information. Shocking, I know. In addition, people who are not your ideal candidates may also enter your database.

We recommend that you create a junk record collection program. Again, you’ll want to delete these records manually to ensure you don’t delete records that aren’t undesired. Start by looking for data with junk in the name fields (first, last, company) and/or emails. For example, “asdf” or “aa” or records from certain domains. You might want to look at certain titles as well, like “student” or “consultant.” This will be an iterative process, where you continue to refine your filter criteria to collect the right candidates for deletion.


Duplicates cause all kinds of problems – which record is getting the score? Which one is getting the email? Which one should sales follow-up with? Furthermore, they skew report data and waste Marketo instance record numbers (to add insult to injury, you are paying to keep your duplicates!)

Prior to merging all of your duplicates, we recommend trying to get a handle on the source of these duplicates. Are there any sources that point to process improvements that would prevent this duplicate creation?

Try to merge as many duplicates as you can before you start migrating your data to your new instance.

The recommendations we’ve made will help you reduce the number of records you have to map/transpose and migrate; and if implemented in your new instance, will help you ensure that your data remains clean/current. Let us know if you have questions or want help, email us at

Your marketing technology experts.

At Digital Pi, we use technology to connect revenue to marketing efforts. We fuse marketing strategies, processes, data and applications to make marketing technology solutions work for clients' businesses.

Learn More
Share this resource

Cookies help us keep the site running smoothly and inform some of our advertising, but if you’d like to make adjustments, you can visit our Cookie Notice page for more information.