From time to time, our customers face a decision as they are looking at integrating two or more of their systems with a platform like Bedrock Data. The question comes up, usually from an IT resource, “Should we code the integration ourselves?”
This usually comes from companies with a sizable budget to fund a third party systems integration project, or an in-house IT team that is considering performing the work.
We caution marketing, sales, and operations teams before jumping into a project like building your own integration. These are five reasons why coding your integration will likely be more trouble than it’s worth.
#1 - It will take you longer to code than using a pre-built connector
Using a pre-built connector means your integration project centers around mapping data and defining your data workflow triggers, which can be easily configured via a web interface.
If you’re building your own connector, you are in an IT/development project paradigm. This means a lot more time and risk needs to be built into your timeline.
Those resources will need to learn the API schemas, perform the development, and leave time for testing and troubleshooting.
And, don’t forget, development projects are often subject to Murphy’s Law. Meaning issues will come up and timelines will often be extended beyond the original estimates of development teams. That’s just the nature of the beast.
#2 - Multi-directional sync rules are complex (it’s more than just pushing data)
Marketing and sales systems integrations need to be multi-directional sync to be effective.
It’s more than just passing a lead over to sales. There is a need for ongoing back and forth between the systems to ensure all relevant data is up-to-date in all systems where prospects and customers are engaged.
To do this well, complex use cases need to be handled elegantly. Once that lead is in the CRM, how are updates handled? What happens when a salesperson updates information like a phone number? What happens when that person fills out another form with additional information? What happens when a lead that already exists in your marketing automation system gets created by sales in your CRM system? What happens when a lead that already exists in your CRM systems gets created in your marketing automation system?
All these scenarios need to be taken into account to ensure you have accurate, de-duplicated data.
#3 - Expertise in both APIs and connecting systems is hard to find
Coding an integration is just the start of what you need to do. Don’t overlook the second half of the equation: ensuring the integration is set up in the right way for your end users. The data is connected, but are all of your reports working correctly? Can data be used in the system the way you intended?
There’s nothing worse than realizing there’s more work to do with a project when you thought you were done. And what happens at that point? Does it cost more money for you? Will you lose your internal resources? This is one of the reasons why less than a third of all IT projects were successfully completed on time and on budget over the past year. (From Standish Group via Capterra).
#4 - Managing APIs requires ongoing resources
To manage the API, ensure resources are involved to troubleshoot errors and issues. It starts with error handling and troubleshooting issues. Then it extends to managing updates as the API changes or new schemas are introduced. It’s not a one-time project, it’s an ongoing project and will need to be resourced accordingly.
If working with an outside systems integrator, this likely means an ongoing retainer will be needed beyond an initial integration project - one that will likely cost more than a Bedrock Data license and support.
Or if this is an internal IT project, ensure you have access to IT resources to manage the APIs on an ongoing basis - which will mean one of two things: 1) Either get ready for a fight with IT staff for resources or 2) You won’t get the resources, and there will be trouble when issues arise.
#5 - Don’t get stuck where new requirements will require another development project
With an integration platform, when there are inevitably changes to your needs, you’ll want to be able to make these changes using a web interface, in minutes.
Let’s say sales wants more fields added for visibility in the CRM.
If you’ve built your own integration, you’ll be looking at another development project as you get new requirements. There’s scheduling, development and testing. Updates could take weeks, or longer.
One of our customers described her previous experience:
“When I needed the smallest change, I would need to submit a ticket to the systems integrator. Then, when I didn’t hear back from them after 48 hours, I’d need to submit a new ticket to ask them to follow up with me on my original ticket. It didn’t work for us.”
There’s a lot of risk and cost that you may not originally anticipate with coding your own integration. An integration platform such as Bedrock Data helps you take away the risk and deliver a better ongoing solution for the business.
Want to discuss your data integration challenges with one of our specialists? Contact us to schedule a conversation.