10 Best Practices for Integrating Customer Data
Scribe Software, a Datix, Inc. partner, held an educational webinar about the 10 best practices for integrating customer data. Pierre Hulsebus, Scribe’s Senior Sales Engineer, presented a clear and practical approach to implementation. His tips come from years of integration and IT experience. We agree and wanted to share our takeaways from the session, too.
Establish Common Terms
Integration might not mean the same thing for each person. What one person calls integration for a certain department might in actuality be migration. From the start, establish shared understanding of key terms will help prevent misunderstandings later on. This extends to setting up what the goals of the integration process is for the business, too. Let’s say a company wants to eliminate manual data entry, then that should be clearly expressed from the beginning of the project. “Build a lexicon of what we are really talking about,” Hulsebus said.
Understand How the Integration Benefits the Business
Have a plan with a measurable outcome. The ROI will widely vary for the type and size of company. “Organizations with 100 people or fewer may not have the same compliance as a $500m manufacturer with ISO 9000 compliance,” adds Hulsebus. This could mean goals such as reduction of the quantity of re-entry or reducing returns by 4 percent or could even mean improving customer satisfaction by a certain amount. Be very specific about your goals upfront. This creates accountability with your integration partner. Did they meet your criteria in the end?
Evaluate Your Real Budget
Staff commitment is a very large part of an integration budget. You will need to dedicate SMEs to specific portions of your project. This could mean time away from their other daily tasks and will involve planning accordingly. Also, you will need to consider first time expenses. This could mean bringing in a knowledgeable consultant to finish a task in a week versus the company taking much longer to learn and architect something you may only use once. “Have the right assets for the task,” Hulsebus advises.
Know the System
Know what system you are wanting to integrate to ensure interoperability. It often happens that the API (Application Program Interface) is use case specific and you might be using the application in a different way. For example, something like SharePoint has been around for many years so the web service portion is used in many different ways. Knowing the system and the APIs will make a huge difference. Some APIs have vendor limitations, throttling or licensing fees. Understand the workflows and the cascading effects of integration in conjunction with the limitations of the API.
Map Twice, Integrate Once
Put everything you want to integrate on paper. Some projects can be iterative (expressly laid out) or agile (learn as we go), but integration projects should be iterative. Get the specific processes that will be mapped, get screenshots and documentation from users and stakeholders. Get all the mappings before starting the integration process. This will uncover issues. Let’s say you needed a quoting tool that integrated with CRM (Customer Relationship Management) and ERP (Enterprise Resource Planning) . It may be more valuable to not just integrate the two systems, but to show the sales person inventory levels or product information. This could save your sales team time calls and emails to other departments. “Planning is more important than learning as you go,” notes Hulsebus.
Know Why Integration Projects Fail
Data quality is a top issue for user adoption. Bad data will cause users to stop using the system. Addressing data quality prior to integration will make sure the best process is in place. Successful integration projects make users lives easier. If you wanted to integrate ERP with CRM you will need to integrate types of data sets. ERP may include invoice information such as a business address or account information. Your CRM may have more complicated or detailed information about the specific actions or buying habits of your customer. “Cleaning up the data along the way is imperative,” Hulsebus said. “Stop the project and make sure it is correct.”
Select the Best Approach
Stay agile with your technological choices. We just mentioned how important it is to be iterative in your project planning, but you will want your technology to adapt with you. Make sure your technology enables your business to move forward, instead of hampering progress. This may mean updating the software you are using to meet your future goals. This will mean taking a look at the costs associated with moving to the next version. If the platform you are currently using is not agile it will not allow you to move forward.
Recognize Design and Performance Capabilities
“We are in an API economy,” Hulsebus said. An example of this is integrating with Twitter to post a tweet each time you have a product update. Twitter will change the application overtime and they built an API, a programmable way to connect Twitter to your tweet. So you will never directly access Twitter’s main database only the application layer. This is how many modern applications, including Scribe work.
Make sure your team knows what is expected. Get your pilot team together from the start. They will become liaisons and will go through the training and relay that information to their teams. Clear communication is critical here when working with an integration partner. Statements of work set the stage, but continued communication is necessary. “Check in all the time and stick to the choices you made early on,” Hulsebus suggests. Integration is a process and takes time.
Happy Clients Means a Happy Project
Including you, the client, in the process of use cases allows gathering of data and business practices. This is something that will affect people’s jobs in the end. For a type of project, like this, we want to know what is working and what is not to continue on the path of success.