Having the same company called by three different reps in the same week is a familiar embarrassment for sales teams. From the customer's perspective, it leaves an unprofessional impression.
The source of this is almost always the same: records imported from different lists and never merged.
In this article, we explained list import, field mapping, and duplicate record cleanup step by step.
Table of Contents
List sources
Prospect lists come from many sources, and each source has its own unique quality issues.
Fair lists are usually incomplete; information read from business cards contains typos.
Web forms are cleaner but may contain fake records. A verification step reduces this risk.
Purchased lists are the riskiest group; both their freshness and consent status are uncertain.
Your existing active list is the most valuable source; priority should be given to it during import.Current account cardyou can check out the article.
Preparation before import
Cleaning up the file before the import is much easier than cleaning it up afterward.
Standardize phone numbers into a single format. The same number written differently ruins duplicate checks.
Simplify unnecessary suffixes in company names; when company type abbreviations are written differently, matches are missed.
Remove blank rows and columns without headers. Half-records turn into neglected clutter after migration.
Perform a test migration with a small sample; realizing an error after migrating hundreds of records is costly.
Field mapping
Field mapping is the correspondence of columns in the file to fields in the system. Incorrect mapping leads to data being written to the wrong place.
Mandatory fields must be mapped first; a record is meaningless without a company name and at least one contact information.
Source information must definitely be added. A record whose origin list is unknown cannot be analyzed later.
The migration date must also be recorded; this information is an indicator of data freshness.
We covered the general logic of file-based migration in the file integration article.
Duplicate record detection
Duplicate detection depends on which field is considered an identifier. An incorrectly chosen key produces either too many or too few matches.
Tax number is the most reliable key; however, it is often not present in candidate lists.
Email address is the second best option. Phone numbers can also be used; however, switchboard numbers can point to multiple companies.
Matching by company name works approximately; spelling differences and branch names can be misleading.
The soundest method is to evaluate multiple fields together and leave ambiguous matches to a human.
Merge decision
If two records belong to the same company, they should be merged; however, a decision must be made on which information will be retained.
The general rule is that the more recent one wins. Historical information in the older record should be preserved.
Activity history and notes are merged; no contact record should be lost.
If the responsible persons conflict, a manager's decision is required; two representatives working on the same company must be resolved separately.
The merge operation must be reversible and leave an audit trail; refer to the audit logs article.
Continuous cleaning routine
Data cleaning is not a one-time task. Every new record carries a new risk of duplication.
The most effective precaution is control at the moment of entry. Showing similar records as a warning when a new candidate is entered solves the problem at its source.
A monthly scan is also useful; matches that could not be caught at the time of entry thus become visible.
Records that have been inactive for a long time should be archived; dead records distort both searches and reports.
It is possible to automate these controls; automation rules we explained in the article.
Consent and data responsibility
Every transferred record may contain personal data, which entails legal liability.
Where the data comes from and with what consent it is processed must be logged. A list of unknown origin should not be transferred.
Commercial communication consent is a separate issue; being on the list does not mean that sending messages is permitted.
Deletion requests must be fulfillable; this requires the transfer order to be established from the start to support this.
We covered the details in the KVKK-compliant customer data article; records are kept in the CRM module.
Executing the transfer work
Transferring carries risks when planned as a single large operation. Proceeding in batches is safer.
First, a small sample is transferred and the result is checked. Errors in field mapping become visible here.
Then the list is divided into parts based on the source; each batch is marked and transferred separately.
Batch marking makes it possible to rollback a faulty transfer in bulk.
After each batch, the duplication control must be run and suspicious matches must be examined.
When the transfer is complete, randomly selected records must be individually verified.
Things to consider
Migrating all data from the legacy system is almost always the wrong choice.
Records that have been inactive for a long time will remain inactive in the new system; they only make searching more difficult.
Therefore, it is much healthier to set a date threshold and leave older records in the archive.
Migrating free-text fields also requires caution; unstructured notes cannot be analyzed later.
Enabling duplicate warnings at the moment of record creation prevents the issue from recurring.
This single setting eliminates the vast majority of cleanup work in the long run.
Frequently asked questions
Which file formats are supported?
Common spreadsheet formats can be used; we verify your file structure together during the setup phase.
Can a failed migration be rolled back?
If the migration batch is tagged with source information, a bulk rollback becomes possible; therefore, the source field is critical.
Should current accounts in the existing ERP also be migrated?
Synchronization is preferred over migration; ERP synchronization see the article.
How many records can be migrated at once?
Large lists are migrated in split batches; this makes control easier in case of an error.
List migration is the most underestimated and problematic step of CRM setup. A system that starts with dirty data continues to stay dirty.
Take the time to clean up your file before migration. Cleanup done afterwards is always more expensive.
Make sure to enable duplicate warnings at the moment of creation; this single precaution solves most of the problem at its source.
By consulting with the EQLEM team you can create your data migration plan.