What’s one of the most important questions to ask before beginning ERP implementation?
Is our data clean?
If your answer is no, you’re not alone. Most companies have data spread across multiple platforms and formatted in various ways. Our advice to these companies is to cleanse and import data before implementation to ensure their ERP software delivers accurate, real-time data. Proactive data migration also ensures minimal downtime at go-live and reduces the duration of operational disruption.
1. Assign Responsibility for Data Cleansing
There is a general misconception that the IT team can handle all the data cleaning. While there is data cleansing software, these packages can only help identify potential areas to be cleansed.
Someone still needs to know, for example, out of two different address records for a company, which one is correct. Also, someone knowledgeable needs to verify that a closing balance is in fact correct.
In most cases, the person who should cleanse these records is not in the IT department but is within the specific department that owns these records. These are called data owners, and they are critical for establishing standards and guidelines to maintain quality data.
IT staff should assist data owners by identifying the data flow, integration points where the data is being updated and what data sources exist.
Examples of Data Needing Cleansing
Duplicate customer or supplier accounts
Master data that isn’t classified with a common taxonomy
Blank description fields for products
Old product codes no longer in use
Data for a customer that has gone out of business
Critical data in Excel that is not in your system of record
2. Remove Duplicate Data
Duplicate items can be identified in two main ways: 1) Direct duplicates include two or more items possessing the same manufacturer name, part number and description. 2) Fit-form-function duplicates include two or more items that possess different manufacturer names and part numbers but have the same fit, form and function.
3. Don’t Migrate All Your Data
Companies continue to persist in the idea that they can migrate all the data from the legacy system over to new ERP software, and the software will somehow magically scrub and standardize it. That’s just not the case.
If you’re moving to a new house, would you put trash and clutter in the moving truck? Probably not, so when you’re moving to a new ERP system, you don’t need to bring over every single piece of data. Some of this data is unclean, unnecessary or pure junk.
The more useless data you bring over, the harder time you will have finding the data you really need, which can delay go-live.
Companies tend to hoard data, believing that someone, somewhere down the line just might need that one nugget of information. An ERP project is a great opportunity to clean house.
4. Determine What Data to Migrate
This can depend on what industry you’re in and your forecasting needs. Determining what data is important should be a collaborate effort that involves all departments.
5. Determine Resource Requirements
You will need developers to convert data and data owners to review and cleanse data. In addition, you may need executives to review certain types of data. Most likely, this will include data elements that relate to the core of the organization’s culture. These may be items your organization has avoided because they seem too hard to address.
Other resources you may need include resources for change management and business process management as you likely will need to train employees on how to input, manage and analyze data once the new system is implemented.
6. Consider Industry Regulations
Different industries have different regulatory requirements. For example, some regulations restrict the ability to change and/or export certain kinds of data records, such as HIPAA with electronic medical records.
You should carefully consider your method of data cleansing (e.g., in-system vs. Excel) if your company has industry-specific regulatory requirements.
7. Consider the Data Complexities of Global Projects
In global ERP projects, you’ll encounter different systems speaking a variety of ERP languages layered on top of cultural and language differences. It is vital to identify these complexities early and communicate to your project team where you see the need for executive decisions. This will eliminate slow periods in the data conversion, allowing developers to stay engaged.
If your project is global, you also will need a signed document from the executive steering committee stating who is responsible for data at the global, regional and country level for both master and transactional data.
8. Define Taxonomies and Attributes
Most ERP systems use some sort of taxonomy to classify items. Master records must be classified correctly, completely and to a level of detail that makes the record easy to identify for search and reporting functions.
While it is not necessary to choose one particular taxonomy, it is necessary to have a taxonomy that supports your company’s business initiatives. Therefore, you should ensure your ERP consultant has experience with taxonomy selection and deployment.
Item record attributes play a similar important role. Attributes define the item and are important for successful parametric searches. Incomplete or incorrect attributes prevent items from being found, resulting in proliferation of parts and bloated inventories.
To ensure a successful ERP data migration project, we recommend extracting, normalizing and completing item attributes beforehand. Because of the sheer volume of attributes to be extracted and enriched, an automated approach is the only practical way to execute this.
9. Develop New Processes
Once the initial data cleansing is complete, you will have more data to cleanse – unless you ensure employees adopt new processes that enable data accuracy. Even though these processes may be designed for legacy systems that will soon be retired, these processes are still worth developing. Consider the fact that most ERP projects last years. Can you endure two more years of dirty data?
We recommend that companies redesign their business processes and train employees both for the future state and this interim state. Both business process reengineering and organizational change management play a critical role in maintaining data cleanliness.
10. Test Before Migrating
Discover if you’re on the right track with your data by moving data to test environments. This will be time consuming and typically requires the development of unique code. Don’t make the mistake of leaving it until the last minute.
11. Use Your Data to Make Business Decisions
With your data migrated, your new processes established and your employees on board, you can almost breathe a sigh of relief – but not quite yet. Now, you must figure out what to do with your data insights.
Companies that have their eye on realizing business benefits from their ERP systems tend to address this issue early with change management and business process management. Knowing that an ERP software solution will enable increased data visibility, forward-thinking companies take advantage of this benefit by establishing data analysis processes. These companies also spend time training employees on these new processes.
Data migration is not usually a priority at the beginning of an ERP project. Instead, the focus is often on planning and design, while data migration is a muddled topic that people tend to avoid.
When companies avoid or delay data migration, their new system will provide unreliable data leading to technical challenges, customer dissatisfaction, low system usage and low benefits realization.
While data migration and cleansing may add cost to an already expensive ERP project, the effort will pay for itself almost immediately through the identification of excess inventory, reduced equipment downtime and improved data insights.