Salesforce Testing & Implementation Recommendations
Olenick has been involved in testing, managing, and implementing Salesforce. Salesforce is a CRM (Customer Relationship Management) cloud platform. CRM platforms allow users to store and manage customer contact information such as accounts, leads, and sales opportunities in one central location. In working with Salesforce, Olenick has gathered some observations and recommendations that warrant further consideration. Some of the main observations and recommendations captured involve Salesforce customization, data alignment, methodology, and validation points. Below are the main takeaways from each category.
Salesforce is very customizable; however heavy customizations require a lot of testing. Salesforce pushes out 3 releases per year so significant testing may be required, as some customized features may be affected by the releases. The recommendation for any Salesforce implementation project is to develop a regression testing suite (automated & manual) for testing new releases.
It was observed that Salesforce was often not aligned to the customer data. In some cases, data residing in Salesforce had not been cleansed and was out of sync. The addition of any kind of inaccurate data will cause data integrity issues. Olenick recommends a thorough data cleanse before any data is imported. The quality of data is extremely important especially when integrating with external legacy systems. It is much less time consuming and far less expensive to tackle data cleansing at the outset versus after the fact. One client did not cleanse data at the outset and spent 12 months cleaning up data related bugs, or as they called them “data gnats”.
Agile is not the best approach for a large Salesforce implementation project. It is recommended to approach an entire Salesforce feature in a given iteration, rather than building out fragments of a feature across multiple sprints. When breaking logical work components into smaller subsets to fit the Agile approach, it often leaves critical acceptance criteria out of stories and results in building fragments of a feature instead of approaching the entire feature. As a result, hybrid methodology is highly recommended in order to both prevent pitfalls of the short sprints and avoid a prolonged project.
Clearly defining requirements for all aspects of the implementation, including: user interface, business workflows, data conversion, system integration interfaces, and security (both for Salesforce application and data) is highly recommended. In addition, defining, aligning, and documenting processes within Salesforce functionality is recommended . Also, set lean, measurable objectives and guidelines before accepting customizations. Lastly, make sure to identify how different systems will be integrated with Salesforce and ensure progress on those integrations is tracked project-wide as a whole. When working with Salesforce it is important to keep in mind the associated interfaces, to ensure full test coverage.
As Olenick continues working on Salesforce engagements, these key learnings can be utilized to help identify and bridge gaps. In turn, some of the recommendations outlined here, can also potentially be used as best practices for future projects entailing Salesforce testing and implementation.