Issue link: https://resources.tangoanalytics.com/i/1465902
tangoanalytics.com 2 Predicting sales to optimize store positioning and revenue necessitates a comprehensive approach that takes into consideration all aspects that impact new and existing locations. Limiting your analysis drives down sales forecast accuracy and value. Consider all factors that affect location and market performance: · Utilize operational ratings and customer satisfaction to assess impact on sales performance. · Analyze historical sales ramps to better predict short- and long-term performance. · Forecast sister store cannibalization and sales transfers to model overall market effects of new store locations. CANNIBALIZATION MODELING Avoid cannibalization conicts. Identify differences in sister store performance by real estate location type and the relative distance between locations: · Evaluate prototype-specic cannibalization. · Analyze by EDITDA rather than percent impact on existing stores. · Utilize exclusive data partnerships to help determine the impact of proposed new locations on existing stores, quickly and cost-effectively. · Measure cannibalization across different asset types. DATA Collect data, analyze competitors, prole customers and forecast performance: · Geofencing and smartphone tracking to rene trade areas. · Mobile pulse data to aggregate foot trafc and day part analyses. · Remote sensing to improve model accuracy. · Utilize credit card transaction data to assess co-tenant attractiveness. ARTIFICIAL INTELLIGENCE / MACHINE LEARNING Combining cutting edge approaches with informed insight. Advanced techniques of machine learning blend deep domain knowledge and key learnings from traditional techniques to yield the most accurate models: · Harness new data sources—social media, credit card, trafc and cell phone—to speed model development and improve reliability through articial intelligence. · Factor in real life conditions to ensure the most accurate, reliable models. · Discover patterns that would be undiscernible by traditional modeling techniques. Real-time, regenerating insights. The challenge of traditional modeling techniques is that they fail to consider the changing interactions and new data that are essential to developing living, breathing models that provide the true pulse of the market: · Quickly and seamlessly refresh the model as parameters change. · Introduce new transaction data for different channels to automatically adjust the model. · Make the right decisions the rst time and save both time and cost by relying on more accurate ndings. TANGO PREDICTIVE ANALYTICS (CONTINUED)