Abstract:
The use of internet advertising as a primary customer acquisition strategy is becoming increasingly common among businesses. Internet companies like Google, Facebook, and Amazon have become platforms for these online advertisements. Standard metrics like impressions, clicks, conversions, click-through rate (CTR), and cost per acquisition (CPA) are used by marketing managers to evaluate the efficiency of advertisements. Managers mainly utilize these indicators to allocate funds to their advertising campaigns, which are then used for bidding on other advertising opportunities. Online advertising is dynamic, and advertising campaigns are susceptible to multiple shocks in demand. Using the data collected from an advertising company that places search ads on e-commerce websites on behalf of consumerpackaged goods companies, we developed a multivariate time series model to investigate the effect of impulse shocks on a specific keyword and its performance. According to the model, we observe the impact of these sudden, impulsive shocks on impressions, clicks, and conversions that define the efficiency of an advertising campaign, a short-run equilibrium among these metrics, and the evolving nature of the keyword in paid search advertisements, and forecast the metrics using the vector error correction estimates. This model can aid managers in their campaign budget allocation decision-making to ensure they can withstand these fluctuations in demand while avoiding either overspending or underspending and longevity of the performance of a keyword