We recently launched a new partner and were surprised by the results of their first demand prediction report. While many of our customers in that region showed demand trending upwards or flattening out based on their operational performance (marketing, sales, operations), our new partner’s demand prediction was declining rapidly. Here’s how we figured out they were artificially disrupting their momentum, obscuring visibility to the market, and how AI was able to uncover the patterns of market demand
Identifying the Issue
First, we analyzed a set of features that best correlate with future demand. Our Machine Learning Model processed around 180 features, and identified their sales momentum as the primary factor impacting the prediction line. This was surprising because, in most cases, the prediction line is influenced by a variety of features.
Diving into the Data
We met with the partner at their offices to review the data, and show them their results. Due to confidentiality agreements (your data at OpenHouse is your data), we could only share that their report differed from others.
During a two-hour session, we discussed various market patterns and consumer behavioural patterns identified by AI that serve as powerful lead indicators to market demand while trying to understand why sales momentum was overwhelmingly dominant in the predictive model until we discovered a missing factor, the VP of Operations nicknamed the “Hand of God.” This referred to the manual capping of sales that occurred irregularly to prevent their slowest construction projects from being negatively impacted by sales deliverables.
Rectifying the Prediction
Our AI had interpreted this manual capping as a change in demand. To correct this, we trained the Machine Learning Model with four years of capping exercises from our partner, and instructed it to recognize these capping exercises as future occurrences. The updated prediction showed a more consistent demand line.
The Exciting Discovery
What’s more exciting is that our AI learned to recognize human interventions—those tactical plays by senior leadership. This means our AI can now acknowledge the strategic prowess of senior leaders and proactively respond to ensuring the right features are prioritized at the right time.
The VP of Operations told us that if we could update the model to reflect their input, it would be game-changing. Well, we did.
Unlock the Power of AI in Homebuilding
Ready to take control of your demand forecasting? Schedule a consultation with our experts today to see how OpenPredict can transform your sales strategy and operational performance.
This experience highlights how our AI can adapt to real-world, human-led adjustments, providing more accurate and actionable predictions. It’s like giving your VP of Operations a digital twin who understands their every move. By incorporating even the subtleties of senior leadership decisions into our models, we’re ensuring that our partners always have the most reliable insights. Ready to see your own ‘Hand of God’ in action? Let’s talk about how we can bring this game-changing technology to your team. 🚀
