How are you using predictive analytics to identify opportunities to improve revenue performance and efficiency?

We are still in the early days of this at Intercom but there are a few things we have introduced recently that I think all companies would benefit from.

1. Forecasting / opportunity management: Nearly every sales tool we use (Salesforce, Gong, Clari, Outreach) includes some predictive analytics elements. I'd encourage everyone to start here and explore what you have already paid for and probably are not using to its full extent. These tools can help identify opportunities at risk, forecast trends, next best actions and so on which is probably 80-90% of the value before trying much more complex things.

2. Account scores: At LinkedIn, our Data Science teams built propensity models for new business account which were hugely helpful in identifying which accounts to assign, how AEs and SDRs should prioritize them, and even which ones should get routed to and SDR vs. nurture campaigns when they MQL. At Intercom, we have a robust health score that is very predictive of churn or contraction risks. That allows our CSMs to proactively engage these customers before renewal to try to improve the outcome. Both of these require more sophisticated in-house analytics, but are hugely valuable in helping reps prioritize their activities to drive better revenue performance/efficiency.

3. Upsell/expansion opportunities: Most PLG/PLS motions include usage-based notifications for when to upsell or expand a customer. If you're trying to sell this way but don't have the ability to identify when a customer may be ready for more from you, it's a large gap likely leaving revenue on the table. It's also important here to have the right "play" ready for the sales team if they receive one of these notifications. That way they have a talk track, quick visibility into the data that is driving the notification, and we can track the performance of the notification and play to understand the effectiveness and adjust as needed.

Leveraging predictive analytics is crucial in identifying opportunities to enhance revenue performance and operational efficiency, and now this whole thing has become even more relevant with the eruption of generative AI as a way to complement and enhance any of the previous AI based predictive analytics models. Despite how mature is your AI strategy or foundations, I think there are some basics that would apply to every organization regardless on where they are in that journey, and here are the ones I consider more relevant:

1. Data Integration: We ensure that we have access to and integrate data (or signals) from various sources, both internal and external, providing a comprehensive view of customer behavior, sales performance, and operational efficiency.

2. Identifying Key Metrics: We pinpoint the key metrics that are strong indicators of future performance. This could include pipeline conversion rates, customer engagement/activation levels, or sales cycle lengths.

3. Historical Trend Analysis: We analyze historical data to identify trends and patterns. This analysis helps in understanding what has worked well in the past and where there may have been challenges. Here's where AI can drastically simplify and eventually make the process more accurate.

4. Predictive Modeling: Using statistical models and machine learning algorithms, we develop predictive models to forecast future outcomes based on current and historical data. These models can highlight potential areas of opportunity or risk.

5. Segmentation: We use predictive analytics to segment and/or prioritise customers based on their likelihood to convert, purchase size, or other relevant criteria. This allows for more targeted and efficient marketing and sales efforts.

6. Resource Allocation: By predicting which areas are likely to yield the highest return, we can more effectively allocate resources, ensuring that our teams are focused on the highest-value activities.

7. Continuous Monitoring and Adjustment: We continuously monitor the accuracy of our predictive models and adjust them based on new data and outcomes. This iterative process ensures that our predictions remain accurate and relevant.

8. Training and Support: We provide training and support to our teams, ensuring that they understand how to interpret and act on the predictive insights provided.

The world and usage of predictive analytics is definitely a journey that it's worth starting in any organisation, regardless of its size. By adopting the above framework, my team and I are able to proactively identify opportunities to improve revenue performance and operational efficiency, driving more informed and strategic decision-making across the organization.