The integration of applications into Salesforce is heavily dependent on many factors. While Salesforce has a very robust API, the capabilities of other applications often becomes the single contributing factor to the time and complexity of integrating with Salesforce. Obtaining an understanding regarding the structure and limits of these applications prior to initiating an integration will yield the greatest results. As the technology of these applications continues to evolve, the capabilities of the applications is becoming less of a concern; however in dealing with many legacy and dated applications there is often a need to account for additional time to ensure a seamless flow of data both into and back out of Salesforce.
When reviewing the time, effort and duration of any integration project the flow of the data must also be considered. For one-way integrations into Salesforce, there are fewer constraints; however as a practical matter it is a best practice to configure the integration to accept data back from Salesforce, specifically the Salesforce ID. While this may not be a necessary element of the integration today, establishing this up front will ease any transitions into a two-integration and make referencing data between two systems that much easier.
In dealing with multi-directional integrations, the time, effort and duration must be adjusted. There needs to be a clearly defined path around the flow of data to ensure that the most relevant and appropriate information is always accurate. Establishing clear data governance regarding the data entry points, master records and data monitoring all become factors to the complexity of the integration project.
Additionally, to best deal with errors and inconsistencies in the integration process, establishing detailed logging records will improve the speed of which any errors are resolved. While this adds to the time and duration of the project upfront, providing more granularity around the transfer of data is critical to the long-term success of any data integration project.
There are also many different methods that can be employed to facilitate an integration with Salesforce. By in large, the extensibility of the Heroku platform continues to deliver very positive integration results. The platform aligns very well with Salesforce through the API, it can be scaled up or down depending on client needs with zero maintenance, and it is delivered at a lower cost. While there is an argument to be had for prebuilt ETL tools and integration applications, most if not all of those require ongoing subscription costs; whereas a custom developed integration application on the Heroku platform does not and it allows for greater flexibility in the integration.
As stated, there are many factors that influence the time, effort and duration of an integration project; therefore making it hard to pinpoint averages. At the low end of the scale, there are applications that have a pre-built connector that require only configuration to establish and integration. An integration where that is the case could be up and running in a day’s time or less. However, for a multi-directional integration that requires data to be parsed prior to being added to Salesforce could take several months.