Migrating to a new CRM or marketing automation tool, can be a great way to add new capabilities or save on cost. The risk of migrating into a new system is the data.
The quality of data, or lack thereof, is one of the main reasons for the failure of migration projects. Many companies not using proper migration processes could end up with far larger expenditures than the cost of migrating.
The importance of data quality
When business as usual processes are run by the business, most of the time there is a good awareness of where data quality issues lie. Duplicates, data formatting, outdated and inaccurate data should be addressed prior to migrating data.
What are the considerations with data quality?
There isn’t really an official definition of data quality as it depends on business circumstances. For example, in a business that relies on direct mail, incorrect postal address data might be considered poor data quality but within a fully digital business, whilst that data might be collected the impact of bad data is nowhere near as severe (albeit it should still be collected for a purpose).
In general terms, data is considered as being good quality if it is “fit for purpose” and during a process like migration, that purpose is likely to change hence planning is of the utmost priority. Many migration projects are put in place with the objective of moving from local to cloud storage for businesses to enable new, modern technology. The new technology will often only be a success if the data is of high quality.
Implications of poor data quality in migration projects
In most data migration projects, teams will test using a small sample of data to ensure everything still functions as expected. In the main, the user testing tends to go well and the problems don’t rear their ugly head until moving to go-live. It is quite typical for people to believe that the test data is representative of their entire database but with so many unknowns, it is almost impossible to predict every eventuality.
Some of the key issues from poor data quality are:
- Increased costs as post-migration data needs to be cleansed and reconciled back with the legacy system
- Incomplete or inaccurate data meaning the business cannot service customers
- Poor formatting where constraints in a new system are different to the previous one
- Integration issues between systems where all data connections haven’t been fully tested
- Duplicates are not found, therefore continuing to cause headaches for sales, marketing and operations
In regulated businesses, poor data quality can easily lead to prosecutions or fines resulting from breaches meaning it must be addressed during migration.
Preventing issues with data quality
Unfortunately, there isn’t a magic wand to wave over your data and turn it into good quality. The projects need to be incremental and fall into some sort of framework that is followed by everybody in the organization. For example, in Call Centres, it is common for agents to update customer details when they phone in but they all need to have the same guidelines otherwise it would be counter-productive.
Here are some simple steps to follow:
- Impact Assessment
Before migration, businesses should undertake a deep data discovery to work out the relationships between disparate data resources. This should also involve creating a catalogue of terms, also known as a data dictionary to highlight any gaps or misalignment of definitions. An impact assessment will also put focus onto the end users of data to make sure their needs are accounted for.
- Data Quality Team
For migration projects, businesses should put together a team from all areas of the business to be responsible for the data relevant to them. This will ensure technical and business subject matters are present at every step who may have their own system idiosyncrasies.
- Data Accuracy
Part of the project should involve reconciling data to ensure it will integrate with the new system. Once a framework is in place following the initial impact assessment, this task is far easier as the business will be aware of where the different definitions and terms lie.
- Data Cleansing and De-Duplication
It is important to remove duplicate records for storage, cost and experience benefits. Quality data needs to be fully identified, ready for migrating over to the new system. In some cases, a strategy is put in place to deal with records falling below the necessary quality standards
- Notes and relationships
During migration, it is important to comment and note records that have been migrated for future reference. I’m sure most of us have had a conversation with a company that couldn’t access our data because it is on an old system. Steps need to be taken to avoid this so as to maintain customer relationships and the potential for churn.
To sum up
Whilst data migration often highlights the importance of data quality, frameworks and processes should be incrementally applied in the business. There is a clear need for enforcing a data quality architecture with the right people so that come migration time, the potential impact of poor data quality is reduced or ideally, eradicated. If you are looking to migrate, StrategicDB can help with your data quality needs.