Steps to Successfully Changing Your Business Software Part IV: Implementation

You've selected and purchased your new software, or signed up to a cloud-based application.  Now it's time to implement it.

There are two major challenges to implementing new software.  One is change management, and the other is data migration.  The new software should be as close as possible to what you documented when you were analyzing your needs during the business analysis phase.  However, the new software's design might require you to make some changes in your processes that you wouldn't have made otherwise.  If you involved all your employees in the design and selection phases,  change management will be easier because they will have had the opportunity to participate in decisions about the new system and be much more likely to be positive about making changes.  If you intend to import data from your old system into the new one you'll need to plan how to do it.  Data migration is often not given the consideration it should when implementing new software.  Lack of planning and appreciation for the difficulty of this task leads to frustration with getting a new system running and working right.

You should follow these guidelines when implementing a new software system:

- First, you should expect, and accept, that things will go wrong.  If you expect that unplanned problems will crop up you'll be much better prepared to deal with them.  Almost any new system implementation requires more work and time than can be foreseen.

- Be careful about setting deadlines for implementation.  Don't base deadlines on arbitrary dates and wishful thinking.  It's good to have a ballpark date for final implementation, but you should be ready to adjust this.

- Avoid the temptation to do a complete cutover from the old system to the new all at once.  This places your business at risk.  You can manage risk by following a staged plan for transition from the old system to the new.

- Include a testing phase in your plan.  You should schedule testing the new software with sample data to be sure it works as expected.  Testing can reveal problems you can get corrected before committing to the new system.  Break up the testing for vaious parts of the software.  Be sure you do thorough testing.  Don't assume everything will work without a hitch in the new software.

- Don't try to migrate all the data from the old system before testing the new system.  You should import only enough sample data to test certain parts of the new software at a time.  Importing small sets of data will reveal issues with data migration which you can address when doing a complete migration plan.

- Be sure to include data cleanup in your data migration plan.  Usually the data in an old system lacks consistency, which makes it difficult to import correctly into another system.  For example, a data field in the old system may have been used to record phone numbers at one time, but then used to record emails at other times.  But the new system has separate fields for phone numbers and emails.  Or perhaps the old system had a field for a phone number which could be entered in any format (with or without parentheses and dashes, with or without long-distance prefixes), but the new system uses phone number fields that require a certain format and number of digits.  You'll need to identify data that's entered inconsistently or incorrectly in the old system and get it corrected before trying to import it.  This might require hiring data entry temp help.

In some cases a business can adopt a new software system without migrating any data from the old system.  In such a case you'd be using two separate systems without the need for any sort of  data synchronization.  In most cases a business will require running two systems in parallel with some kind of data synchronization until the old system can be dropped.  In the latter case it would be wise to plan for a phased transition from the old system to the new one.  This means using only parts of the new system in stages.  This helps to mitigate risk, and affords a fall-back strategy in case issues are uncovered with the new system.  By keeping the old system running and up to date you can fall back to using the old system for everything until issues uncovered in the new system are resolved.  

Keeping data in sync between two systems can be a challenge, and must be carefully planned.  It might require doing double data entry, or  running data synchronization software, or a combination of the two.

Of course, there should also be time planned for training on the new system.  But training should not be used as a substitute for early employee involvement in the software selection process and testing it to be sure there aren't issues you didn't anticipate.