If you missed our recent webinar, “Calibrations with Confidence,” don’t worry—we’ve got you covered. We’re delighted to share the highlights from our latest research self-funded study.
At Beyond Blue, what drives us is our desire to seek out the truth and deep dive into complex prescribing and predicting behaviour; our Calibrations with Confidence study does exactly that.
Passion Project: Discovering the Influence of Behavioural Biases on Forecasting
This is the culmination of a passion project of around 2 years. We aimed to answer a critical question: Can physicians accurately predict their future prescribing behaviour? By digging into the data around behavioural biases and personality traits, we were able to even out any ‘bumpy’ forecasts.
Uncovering the say-do gap
We worked with over 250 US Physicians, who predicted their new prescriptions for SGLT-2s over a week. We then compared these predictions with their actual prescriptions to measure the gap between prediction and reality.
We found 3 types of estimators:
- Accurate Estimators: Approximately 75% of physicians gave accurate predictions
- Under-Estimators: About 10% underestimated their prescriptions
- Over-Estimators: About 13% overestimated their prescriptions
Personality Traits and Prediction Accuracy:
Weaving this in with the OCEAN or ‘Big Five’ personality profiling test, we were able to demonstrate correlation between personality traits and prediction accuracy.
- Accurate Estimators showed higher conscientiousness and lower extroversion
- Under-Estimators had lower neuroticism
- Over-Estimators were more agreeable, often giving “good” answers rather than accurate ones
What This Means for Forecasting and Targeting
Our findings have big implications for improving forecast accuracy and refining targeting strategies:
- Focus on Recalibration: High-share physicians need more recalibration attention due to their greater inaccuracy.
- Adjust Targeting: Without recalibration, over-estimators might be over-prioritised, and under-estimators, who could be frequent users, might be overlooked.
- Integrate Personality Profiling: Using this profiling helps us categorise and calibrate segments more accurately, enhancing forecast reliability.
Rolling Out Calibrations with Confidence
We’ve developed a bespoke typing tool that sorts physicians into accurate, under, or over-estimators with 90% accuracy. This tool ensures we only calibrate segments we’re confident in. We’ve also expanded our database to include various markets and specialities, continuously improving our approach.
Real-World Benefits
Our enhanced calibration model offers:
✔ More Effective Targeting: Identifying the right priority targets
✔ More Reliable Data: Correcting inaccuracies and filtering out synthetic responses
Let’s Continue the Conversation
We’re thrilled by the insights this study has provided and how it will help us, and our client partners achieve more accurate forecasting. If you’re interested in learning more about our methodology or want to see how we can support your forecasting needs, get in touch with us. We’re here to help you make better assumptions, better calibrations, and deliver better data.
Or, if you would like to learn about what type of a predictor you are, take our quick test here.