Poor Salesforce Data? 5 Ways AI Can Help

At Cien we speak to sales leaders every day, and most tell us how frustrated they are with the quality of their data in Salesforce. After all, decisions are only as good as the data on which they’re made.

Sure, even when they think they do, no company I know has absolutely perfect CRM data. But poor CRM data prevents sales operations teams from identifying improvement opportunities or the finance department provide accurate forecasts.

Forward-looking sales organizations can make use of new advances in Machine Learning and Natural Language Processing to clean, standardize and enhance their Salesforce data in ways that were previously impossible.

Here are five:

1. Fix Missing Updates

It’s not uncommon for reps to miss entering in activities and opportunity stage changes that would have reflected the true state of a deal. Predictive analytics tools can detect and score a reps’ propensity to do this so sales ops can easily identify those reps and instruct them to do better. Many organizations can also invest in AI-powered automated call recording systems or outreach management tools to minimize the number of missing entries.

2. Suggest Incomplete Updates

Sometimes the data entry fields in Salesforce are not useful for complete entries of data records. Not every company can automatically collect that data by recording all email, calendar and phone activity straight into Salesforce. In some cases AI can predict how a certain field should be completed and fill in the data automatically to complete the record.

3. Update Untimely Updates

Most sales professionals understand the importance of proper record keeping, but are simply too busy to update their CRM during the day. Instead they keep loose notes, and enter it in later, e.g. on Friday afternoons. With the help of AI, you can detect the entries that fall behind and adjust weekly activity levels to accurately reflect the team’s efforts.

4. Standardize Inconsistent Entries

Many freeform entry fields, like lead source, or city (think LA versus L.A. versus Los Angeles) will have dozens of unique entries. This can make a significant analysis difficult. You can use Natural Language Processing to standardize and consolidate data entries. This is particularly helpful for making address or lead source data (eg. Dreamforce and Dreamforce ‘17) consistent. Your marketing and finance teams will thank you for ending inconsistent entries that hamper their reporting abilities.

5. Understand Changes in Definitions and Methodologies

Shifts in definitions or methodologies such as moving to account based sales and marketing are very common in organizations as they grow. This makes time-based comparisons difficult to perform, and causes confusion in reporting. AI-powered apps like Cien use Machine Learning to automatically understand and adjust to changes in your data methodology. So if you used Salesforce’s lead module in 2017 and the account module in 2018 to prospect, AI will detect and account for that. You can then have a consistent prospecting report for both periods.

While there are plenty of other use cases for applying AI to your Salesforce data, having a workable dataset is definitely a good place to start.

The original version of this article first appeared on Salesforce Ben.