Trend analysis involves collecting and analysing data in an attempt to discover a "pattern" or "trend" in the data. Forecasting techniques leverage historical data and trend analysis to determine possible trends from that data and use the information to extrapolate what could happen in the future - if those trends continued. In business, forecasting is used to estimate some variable of interest at some specified future date. For example a sales or demand forecast involves trend analysis of past sales or demand data as a series over time, discover the trends, and then estimate the possible future sales or demand of the product, assuming that the same trends will continue to prevail.
Trend analysis and Forecasting techniques predominantly use time-series data analysis (because the data and the variable of interest need to be analysed over time to discover the trends). Compared to trend analysis and forecasting, predictive analytics is a more generic classification of techniques that can involve analysing data over time (also called time series or longitudinal data), or cross-sectional data (snapshot of data at a given point of time), or panel data (data that can include both longitudinal or cross-sectional data), and provide a predictive score about the likelihood of an outcome. For example, forecasting might estimate the total number of ice cream cones to be purchased in a certain region in future, while predictive analytics predicts which individual customers are more likely to buy an ice cream cone.
At KMS, we employ Trend Analysis and forecasting techniques to improve the planning processes for businesses:
Market share forecasting to estimate future market share by leveraging competitive analysis and market research data.
Inventory management and demand forecasting that can involve SKU-level forecasting to estimate demand at an individual product level.
Website traffic or network forecasts to estimate future traffic loads that might need to be supported leading to capacity planning.
Human Resource planning to estimate volume of talent pools or growth of talent in specific technology areas.
Real time anomaly detection by estimating and comparing - estimated versus actual data points and generating real time thresholds, to detect possible anomalies.