"Why is it important to invest time and effort into the study of factor analysis and statistical Science when designing a dynamic pricing model?"
It is a race against time and tough competition to reach the finish line, first. Planning and preparation is essential. Breaking into the revenue optimization problem and complexities associated with perishable inventory management, those discussed in our first viewpoint – Make A Move Or Choose To Lose, we concluded with a set of pertinent questions that the business must answer when they decisively undertake a journey towards devising a dynamic pricing engine.
The next 100-meter relay readiness is about shaping a study of Factors Analysis and Statistics. Factors that represent commercials, competitors, customers, seasonality and the differentiated business model. Followed by, a prudent choice of statistical models, ones with evidential accuracy, that will drive price prescriptions for revenue optimization.
Judicious selection of data sources forms the fundamental basis for Factors Analysis and Statistics. Taking events data into perspective, by way of example, 26th January marks the Republic Day celebration in India. Parades showcasing India’s defence capability and its cultural and social heritage take place at Rajpath in the Nation’s Capital – Delhi. Amidst high security alerts, the Nation’s President, the Governing Machinery and a Chief Guest travel into Delhi to celebrate and pay their tributes. Necessary security measures enforcing preventive checks impact the state border, city and inter-state trains, metro rails, air traffic and road transport. A quick look at the week of 20th Jan to 26th Jan offers the following information:
Changing trends, otherwise unavailable in seasonality plans or pricing plans of airliners and hoteliers, can be converted into the business opportunities, using statistical algorithms.
Are You Flying High Enough? A use case that offers insight on how dynamic pricing helps optimize revenue.
Information on events is publicly available which could be aggregated to build an events repository. Then, events that have a strong correlation with booking curves must be shortlisted for a comprehensive study of demand trends. An equally relevant source of real time traffic information is also useful to understand the impact of no-shows on changing booking curves. Speaking of which, Google, Bing and Yahoo Map APIs (and soon Uber Maps) provide intelligent traffic data insights.
To conclude succinctly, statistical algorithms are only as good as the quality of data sources, business assumptions and the quantification techniques. While linear regression remains the favourable choice for arriving at the dynamic price range, for accomplishing revenue optimization, developing and implementing a statistical model and validating its decay rates requisites feature engineering capabilities. Real-time model operability for pricing decisions is most successful when technology infrastructure is architected to engender statistical modelling at its highest precision levels.
The buck doesn’t stop here or with the development and implementation of a statistical model once. Frequent examination of the changing characteristics of factors and validation of the statistical models for change in decay rates is essential to find the new optimized levels.
Next up, Investment In New-age Technology Architecture Prevents A Fall & A Fracture
Profisor Services Private Limited
Management Consulting Office
Welldone Technology Park, 1106 A & B,
Sohna Road, Sector - 48
Gurgaon, Haryana 122 018 India
mobile India: +91 70110 69795