The goal is simple: transfer data out of one system and into another. But the process, not so much. Developers must simultaneously juggle APIs, recreate workflows and reports, and manage client expectations. All that considered, a “simple” migration project can quickly become overwhelming, especially for inexperienced developers.
But don’t fret! Our dev team has plenty of experience in handling these projects. Here are five tips to help your migration projects run smoothly so you can look like heroes to your clients.
1. Have a Game plan
Before taking on a data migration project, team members involved in implementation need to develop a sound game plan. This seems obvious, but we’ve found that whiteboarding processes upfront can help visualize all the different pieces involved so we can dedicate team members to specific tasks.
Use this time to come up with realistic timelines for deliverables and note any potential roadblocks that could prevent due dates from being hit. For example, you may need client input on finalizing custom fields before your team can move onto the next task. These sort of things need to be relayed to the client before the project begins so expectations are realistic on both sides.
2. Familiarize Yourself with API Documentation
Because there’s no magic button to make this all work, developers will most likely need to write custom scripts built around APIs to make multiple systems interact with each other and automate as many processes as possible. Thus, developers should familiarize themselves with the appropriate API documentation as early into the project as possible.
ActiveCampaign’s API, for example, will require different parameters than HubSpot’s API, and understanding these distinctions ahead of time will boost your efficiency once the implementation begins. Luckily, nearly all prominent platforms will have publicly available API documentation for developers to reference. These docs generally provide all the tools you’ll need to get started: endpoint URLs, required parameters, examples in multiple interface languages, and authentication requirements.
3. Utilize Sample Data
Once you’ve exported all the necessary data from the old system, it may be tempting to immediately upload it to the new system. But before this is done, we strongly recommend handling sample data to test how your import will respond in the new system.
For example, say you’re migrating data from Salesforce to HubSpot. Before you import the 100,000 contacts and 50,000 opportunities with associated activity history to HubSpot, import a random sample of 200 contacts and 100 opportunities and their associated activity history. This will allow you to analyze the success of your migration on a small scale so you can more meticulously examine how the data is formatted in the new system.
This method will help prevent potentially reaching daily API call limits with certain systems as well. If you import all 100,000 contacts and then discover you need to change a field and send the API request again, you risk exhausting your API calls and having to wait 24 hours to try again. By uploading small sample data, you can iron all the kinks out first and then confidently migrate full data sets.
Data Timeliness
It is essential to consider the time interval that data needs to pass back and forth between integrated systems. For example, say information has been input into system A. Does it immediately need to move to system B as well? How often does it need to move? Hourly, daily, weekly, or monthly?
The main question here is how quickly data must move across your systems. If you require the most up-to-date data, you’ll need to extract as it enters your system. Webhooks are a great way to link data across systems. A webhook is when an action happens in system A, and it automatically sends to system B. This is ideal for interconnected systems, like marketing automation platforms and CRMs, so both marketers and the sales team have the most accurate information at any given time.
Interested in joining the obo team?
Data Models
It’s essential to understand if the data models of the systems being integrated are natively compatible. It usually comes down to whether systems use Custom Object or Standard Objects. An object is a set of properties that make up an item. Standard objects could be Account, Contact, Lead, and Opportunity, while custom objects are objects you create to store information specific to your company or industry. A piece of data leaving the system must have a “home” to go into the system it’s being integrated. At OBO, we align these items in the systems we integrate to make data models interoperable.
4. Develop a Support System for Client Feedback
Depending on the extent of the project, the work may not necessarily be complete once the data has been migrated. Oftentimes, and especially for agencies, the implementation team will need to handle client questions and feedback surrounding the new system. Keep in mind that the client likely has zero experience with the new system and will need support in getting started.
Depending on the size of the client, your team may become overwhelmed with questions and feedback relating to navigating the new system. We’ve found the best way to handle this is to develop a “ticket system.” This could mean creating a form for the client to fill when they have a question or sharing a Google Doc between teams–essentially, any tool that will allow feedback and questions to accumulate for a few days to a week so your team can respond in an organized and timely manner will suffice.
When responding to questions and feedback, provide as much documentation (official or custom) as possible. This is essentially “teaching the client how to fish” instead of simply “giving them a fish” — by providing them with documentation, you provide them with tools they can reference and learn from, allowing them to feel more comfortable and confident in their new system moving forward.
5. Document Everything
The key to learning from each project is to be able to analyze failures and mistakes and understand how to eliminate them in the future. This is where documentation becomes critical. If your team is able to identify mistakes of a past project as a potential roadblock of a future project, then they have successfully taken a negative and turned it positive.
By documenting the successes and failures of specific processes, each project your team takes on will become more efficient and profitable as you will have gradually laid down the framework for successful implementation.
Contact us today to learn more about data migration.
Take me home!
Read more blogs.
