Making Your Salesforce Data and Einstein Work For You

Data is important. It helps you measure your past business decisions, analyze current solutions, and informs the future choices you make for your company. Salesforce is a goldmine of data on your customers and their interaction with the various departments of your company, especially with the implementation of their new AI data scientist, Einstein. We’ve been hearing a lot about all of the amazing things Einstein can do, it’s pretty awesome – however, Salesforce reporting and Einstein can only do their best job if two things are happening:

1. You have data from the other areas of your business flowing into Salesforce
2. Your data is free of bad data ie. duplicates, conflicting data, incomplete data, invalid data, or un-sychronized data

For many, this seems easier said than done. According to The State of Salesforce Report, 2015-2016 conducted by Bluewolf, “Poor access to different types of data, including data outside of Salesforce, topped this year’s list as the number one barrier to deriving insights from Salesforce —and the best companies are focusing on integration to mitigate employees’ data challenges and improve their experience.” Overcoming this barrier is important and bad data is no joke. Bad data costs the U.S. economy 3 trillion dollars a year. That means that bad data is responsible for a monetary loss that amounts to over twice the amount of the 2011 Federal Deficit. From a business standpoint, studies have found that bad data could potentially be costing you as much as 10-25% of your entire revenue. Moving bad data back and forth will only cause a business to loose more money, more time, and more customers. In order to make your Salesforce Data work for you here are a few tips when getting integrated:

Get Rid of Pre-existing Bad Data

David Ciccarelli, Founder and CEO of Voices.com advises to, “take the time to clean up your data before importing it into Salesforce. Avoid duplicate accounts and contacts by merging accounts whenever possible. As the saying goes, ‘garbage in, garbage out.’” You don’t want bad data when putting your information into Salesforce and you certainly don’t want to send bad data into other applications, so take the time to clean up your data before you integrate with another app. Also consider if you have been using different labels or systems in Salesforce and other apps like your accounting apps. For example, if you store account names in Salesforce using the Company Name  but for “QuickBooks Customer Name” in QBO you use the name of an actual person in that company, you need to make sure it is consistent for all your Account/Customer records. If they are not consistent, is there another way to check? If these records are not yet synced with unique ID numbers, be sure to go through them and figure out a precise method of matching Account/Customers before performing a full sync to prevent duplications.

Use a Non-Changing Field to Match Systems

All one-to-one integrations allow duplicates to slip through and any service that doesn’t automatically take care of API generated duplicates will leave your data duplicate ridden. Though cleaning up your old data before integration is the best practice, it’s impossible to keep your data error free if your integration service cannot handle de-duplication logic. The next step to avoiding duplicates is to use a field that will not change like ‘Record ID.’ If you use a field like ‘Account Name’ and match it to ‘Customer Display Name,’ it is not a foolproof way of checking for duplicates because the name might change in one system and the integration will still search for the original name. When the system sees that there is no record of the “new” name, a duplicate will be created. Using an ID number, which will not change, is the best way to ensure your integration prevents duplicates.

Look for Custom Objects Support

Custom objects and fields are truly supported when a provider is capable of picking up your data dynamically, as opposed to statically pre-defining the fields and objects that will be available for you. An integration service that truly supports custom objects will allow you to interact with your objects using whatever you’ve already named them in Salesforce. When only predefined fields and objects are usable it’s called a “static schema” and will require you to go back into Salesforce and change the name of all the objects to the predefined field names. Limiting your organization to an integration solution that uses a static schema will make new and more complex workflows difficult and require extra work.

Be Aware of Different Data Structures Between Apps

In many accounting apps, customers and their contact information are stored in a single record. However, in Salesforce, accounts may have several contacts under a customer profile. This difference in data structure can be confusing when integrating the two. If you want a two way sync, which means you are not only sending information from Salesforce to accounting, but from accounting back to Salesforce as well, then sending contact information becomes more difficult. Because there is more than one contact in Salesforce, paying attention to the one you are choosing is essential. In The State of Salesforce Report, companies that had integrated or planned to integrate their cloud apps were more likely to think Salesforce’s overall value had increased in the past 12 months versus those that hadn’t integrated. Integration opens up new possibilities and better insights from Salesforce, letting you get the most from the system. Whether you plan to use Einstein or not, having good quality data integrated from system to system is absolutely key for any business that wants to continue to grow. The next step? Making sure your integration platform is flexible and can support agile growth for digital transformation.