A study of chief marketing officer (CMO) tenure with competitive sorting model
Eun Hee Ko,Douglas Bowman,Sierra Chugg,Dae Wook Kim
DOI: https://doi.org/10.1145/3190645.3190717
2018-03-29
Abstract:With the average CMO tenure increasing significantly over the past decade, the business press has speculated about reasons for this climb while the academic literature has been relatively silent, remaining indecisive about the contributions of the CMO to firm performance [3, 7, 10]. These mixed results have resulted in calls for more systematic inquiry into the performance consequences of the CMO. This proposal investigates factors associated with CMO tenure. It develops theory based on competitive sorting model whose the underlying intuition is that when the competitiveness of an individual's talents aligns with a firm's strategic directions job tenure increases. We argue that the firm's strategic change has a positive impact on firm performance when the CMO has aligning skills with the firm's strategic shift. Rationales underlying our arguments rely on a competitive sorting model of the CEO labor market [6, 11]. The essential intuition of the model is that CEOs have discernable characteristics that are indicative of their expected productive skills and are matched to firms competitively [4]. We used the sales data for the firms from 2000 to 2014, which were retrieved from Fundamentals Annuals section of COMPUSTAT database [9] and tested the effects of the interaction between the firm (i.e., Long-term business strategy and Data-driven approach) and the CMO variables (i.e., Analytical Ability Index (AAI) and General Ability Index (GAI)) on CMO tenure and the performance implications of the interaction, focusing on the emergence of business culture which transforms diverse aspects of business foundations, data-driven culture. In specific, we suggest a positive relationship between the CMO's characteristics which match to the firm's strategic shifts and CMO tenure and firm performance, because CMOs' distinguished characteristics which are effectively matched to firms are the indicative of their competitive performance consequences [4], which, in turn, are associated with longer tenure. We adopted five proxies of General Ability Index: number of firms, number of industries, CMO experience, number of executive positions, and executive tenure. Following Custodio et al. [2], we reduced the five proxies into one-dimensional index using principal component analysis [14] which extracts common component. We used one component instead of five by employing the method of dimensionality reduction to avoid multicollinearity [5] and minimize measurement error. Because all the proxies for the GAI Index are static variables over time, the General Ability Index is calculated for each CMO, but the index is not varying over time for a CMO. The index is standardized and thus has zero mean and a standard deviation of 1. AAI is computed with the same method using three proxies: number of degrees, degree kind, and functional career experience. To extract the proxies from firm side (firm's valuation on the change in long-term business strategy and in data-driven approach), we use text analytics with `Business Section (i.e., Item 1)' in Form 10-K, which is the most comprehensive compilation of information on a firm's business that is in the public domain. We applied lexicon-based sentiment approach which involves calculating orientation for a document from the semantic orientation of words or phrases in the document [13] and automatically extracting the semantic values in a numeric format to our analysis. Lexicon-based approach can be created manually [12], or automatically, using seed words to expand the list of words [8, 12], and we adopted the first approach. We construct a proportional hazards model to estimate the effect of predictors on CMO tenure. The dependent variable in Equation (1) is the hazard rate, which is right censored for some individuals. That is, there are the individuals whose tenure end points are unobservable, because they still serve as a chief marketing officer at the end of the study period. Thus, traditional regression estimates will be biased. In addition, time-varying covariates (i.e., the firm's valuation for long-term business strategy, the firm's valuation for data-driven approach) are included in the estimate. Consequently, we formulate the model with proportional hazards model [1]. The tenure for an individual CMO is considered to be a random variable with p.d.f. f(t) and c.d.f. F(t), and hazard rate is h(t) = f(t) / (1-F(t)). Let h(t|x) denote the hazard rate for a CMO i with certain conditions captured by the vector x. The hazard rate takes the following form: [math here] where h0(t) is baseline hazard rate which does not depend on x but only captures time effects and ' captures the effect of predictors (xit) on the hazard rate. Figure 1 (a) and (b) illustrate a survival curve and a cumulative hazard curve which are estimated with the hazard model when no predictors are involved. In Figure 1, the relatively gentle decline in the early months indicate that there are only a few CMOs who leave from the position in the first few months. This is also indicated by changes in the cumulative number of events and number at risk. In specific, about a third CMOs left the position within 54 months, and about 50% of the total events occurred within 80 months. We make the following new contributions. First, to the best of our knowledge, this is the first study which attempts to identify the factors related to the increasing CMO tenure. Second, we employ existing lexicon-based sentiment analysis method to take the values of firms' strategic transformations but create new lexicons including the list of indicators and corresponding sentimentic orientation values and new algorithm to capture the accurate values that we would like to examine. Finally, this study offers a managerial implication by showing the effectiveness of certain CMO skills which align with firm's strategy direction. In the future work, a competing risk model can account for different paths to exiting the CMO position. Among the CMOs who leave the CMO position, we observe 77% leave for a different firm or retire and 23% stay with the same firm. We plan to further investigate the data with the competing risk model to observe the CMO's post-career or the reasons of departure from the position.