Temporal Patterns of New Product Introductions and IPO Value: The Importance of Recency,Dispersion,and Asymmetry

Abstract

A firm's innovation activity is often judged by the number and type of new products it launches, but when these products are introduced may be equally important, especially before going public. This research investigates how the temporal pattern of new product introductions (NPIs) influences a firm's initial public offering (IPO) value. The analysis uses data from 298 firms that went public between 2006 and 2023, and focuses on three patterns of timing: recency (how recently the latest product was launched before the IPO), dispersion (the degree to which NPIs before the IPO are spread out over time), and asymmetry (how disproportionately NPIs were launched over time before the IPO). Results demonstrate that firms with more dispersed NPIs achieve higher IPO value, whereas firms that introduce a product just before the IPO or introduce several new products disproportionately closer to the IPO tend to perform worse. Findings show that 1% increases in recency, dispersion, or asymmetry result in a 1.34% decrease, 3.65% increase, and 1.03% decrease, respectively, in IPO value, and these vary depending on the firm's industry growth and product innovativeness. Results provide new insights for theory and practice, highlighting that innovation timing is a key determinant of IPO performance.

Private companies aiming to raise capital often pursue an initial public offering (IPO) to improve their growth prospects. Although going public dilutes the existing owners’ equity, it provides the firm with additional capital to pursue growth (Steinbach 2018). Notably, the firm's market capitalization on its first day of trading indicates its overall valuation, reflecting investors’ perceptions, confidence, and expectations of the firm, setting the stage for future investment as well as the firm's positioning within its industry (Binder, Steiner, and Woetzel 2002; Nasdaq 2024). For example, DoorDash was valued at $71 billion on its first day of trading in what is considered a blockbuster debut (Hussain and Franklin 2020), whereas Boingo Wireless had a less successful debut with a market capitalization of $439 million (Savitz 2011). Although pursuing an IPO has risks, private companies continue to seek IPOs as “IPO momentum surges to the strongest levels since 2021” (PwC 2025).

Indeed, IPOs generate considerable fanfare in the stock market, impact economic growth, and consequently receive substantial attention from academic researchers across various disciplines. Research in marketing demonstrates that various pre-IPO activities can affect how IPO firms ¹ perform on their initial trading day, as investors assess the firm's activities before the IPO to gauge its future potential (for a representative literature review, see Web Appendix A). Investor perceptions determine IPO value; fewer concerns about the firm's performance potential generate better investor perceptions about the firm and higher IPO value. ² To gain initial insights on investors’ views, we surveyed 17 institutional investors from global institutional investor firms (M_{trading experience} = 9.12 years). When asked about the importance of a company's innovation activities before IPO, all but one investor indicated innovation is important or extremely important (for details, see Web Appendix B). Emerging research in marketing validates the importance of innovation, showing that new product preannouncements (Cao et al. 2023) and breakthrough new products (Wies, Moorman, and Chandy 2023) impact IPO performance. However, the investors we surveyed also indicated that the timing of new product introductions (NPIs) is important, highlighting NPI temporal patterns that have not been studied. ³

Theory also highlights the importance of NPI timing and reinforces our framework, which is grounded in institutional investors’ firsthand accounts of the significance of the temporal patterns of NPIs. We find that investors track the timing of a firm's NPIs, and this visible activity signals important information to investors. Indeed, signaling theory suggests that not only do NPIs convey credible information and reduce information asymmetry (e.g., Wies, Moorman, and Chandy 2023) but timing also influences the efficacy of a signal (Bafera and Kleinert 2023). Of managerial importance, NPI timing is readily accessible to investors (Moorman et al. 2012).

For IPO firms, we expect different patterns in their NPI timing prior to the IPO to influence investor perceptions of the firms’ performance potential. Theoretically, the dispositional perspective of time in psychology emphasizes the importance of pacing style (e.g., Chen and Nadkarni 2017; Mohammed and Angell 2004). Pacing style refers to the pattern of effort distribution over time in working toward deadlines (Blount and Janicik 2001). Drawing on pacing style, where the deadline is the IPO date, we investigate three patterns in NPI timing that collectively reflect pacing style: (1) recency, which refers to how close the last NPI is to the date of the IPO, (2) dispersion, which reflects the extent to which the NPIs before an IPO are spread out over time, and (3) asymmetry, which represents the skewness (or imbalance) of NPIs over time before an IPO and ranges from NPIs disproportionately clustered near the IPO (i.e., left-skewed asymmetry) to those disproportionately clustered further from the IPO (i.e., right-skewed asymmetry). While firms introduce new products before the IPO to lower investor uncertainty, the recency of the latest introduction and the asymmetric pace of introductions may inadvertently signal negative information to investors. In contrast, greater NPI dispersion may signal that a firm is prepared for longer-term competitive advantages. Accordingly, our first research question is how NPI recency, dispersion, and asymmetry prior to IPO shape IPO value. ⁴

To develop a nuanced understanding of the signaling effect of NPI temporal patterns, we also consider moderating factors that could shape signal efficacy. Following signaling theory (e.g., Bafera and Kleinert 2023; Connelly et al. 2011), we focus on two relevant factors: (1) the signaling environment and (2) signal message. A firm's industry forms the basis for the signaling environment (e.g., Epure and Guasch 2020), and industry growth has been shown to influence a signal's efficacy (e.g., Gao, Gopal, and Agarwal 2010). Industry growth is also a critical determinant of new product success (Henard and Szymanski 2001). Firms have more and better opportunities to generate demand under high growth (e.g., McDougall et al. 1994) so that NPIs in a growing market are likely to generate different investor perceptions than those in a stagnating market. However, rapid advancements in technology and a consistent influx of market demand in high-growth industries (e.g., McCarthy et al. 2010) tend to generate environmental noise that compromises signal effectiveness (e.g., Bafera and Kleinert 2023; Connelly et al. 2011). In addition to the signaling environment, signal message is a key characteristic of the signal that refers to “the information that a signal communicates” (Bafera and Kleinert 2023, p. 2426). Regarding signal message, product innovativeness represents the degree of newness a product exhibits relative to existing market offerings (Szymanski, Kroff, and Troy 2007). Highly innovative products incorporate new technologies, serve new market segments, or create entirely new categories (Chandy and Tellis 1998). Such innovative products tend to carry greater market uncertainty but also offer differentiation and potentially higher returns (e.g., Chandy and Tellis 1998; Sorescu and Spanjol 2008). Thus, the innovativeness of a firm's new products provides additional relevant information to investors that influences information asymmetry around the temporal patterns of NPIs on IPO value. Our survey of institutional investors also reinforces the importance of the industry's growth and the significance of a firm's product innovativeness in evaluating the timing of an IPO firm's NPIs and assessing the value of an IPO. Thus, our second research question investigates how the signaling environment and signal message, observable from the firm's industry growth and product innovativeness, moderate the relationships between the patterns of NPI timing and IPO value.

To investigate these relationships, we collected data on 298 firms that went public (i.e., had an IPO) between January 1, 2006, and December 31, 2023, and had more than two NPIs. ⁵ In line with extant research in marketing, we use market capitalization on the first trading day to measure the IPO firm's performance (e.g., Saboo, Kumar, and Anand 2017; Xiong and Bharadwaj 2011), which captures firm value and offers a systematic forward-looking assessment of an IPO firm's future performance and growth opportunities. We also control for other factors that may affect IPO firm value, including venture capital backing, underwriter reputation, IPO market intensity, firm and founder characteristics, industry characteristics, and the year of the IPO. We use a two-stage Heckman model to address self-selection bias in the focal firm's decision to go public and employ a control function approach to correct for potential endogeneity in the NPI timing variables as well as the firm's product innovativeness.

Our results generate several theoretical and managerial contributions. First, we extend the extant IPO literature in marketing by demonstrating the importance of the temporal pattern of a firm's NPIs before its IPO. We find that recency and left-skewed asymmetry are negatively associated with IPO value, whereas dispersion is positively related to it. Our research shows that, on average, introducing the last new product prior to IPO approximately 3.5 days closer to the IPO (i.e., a 1% increase in recency) decreases IPO value by 1.34%. We also show that, on average, a 1% increase in NPI dispersion results in a 3.65% increase in IPO value, whereas a 1% increase in NPI asymmetry where multiple product introductions are clustered closer to the IPO date is associated with a 1.03% decrease in IPO value. These shifts in IPO value represent millions of dollars, where a 1% shift in IPO value represents approximately $23 million for an average firm in our sample. Second, we contribute to the literature on innovation timing (e.g., Sharma, Saboo, and Kumar 2018) by revealing how the recency, dispersion, and asymmetry of NPIs are differentially associated with IPO performance. With the IPO date a crucial deadline, our research reveals new insights into the temporal pattern of a firm's NPIs as it prepares for its IPO. In contrast to research that shows a delay in NPIs can result in higher returns (Moorman et al. 2012) and a delay compared with a competitor's product release can result in greater profits (Harutyunyan and Narasimhan 2024), we show that delaying an NPI so that it occurs close to the IPO will hurt IPO value. Third, our research not only highlights the importance of temporal aspects espoused in signaling theory (e.g., Bafera and Kleinert 2023; Connelly et al. 2011) but also reinforces the need to consider multiple temporal aspects of signals, including both immediate and cumulative aspects of time. Finally, our results provide important guidance to managers preparing for an IPO, revealing how patterns in NPI timing influence a firm's value on the first trading day. Strategic timing is an accessible way to signal an IPO firm's commitment to innovation and growth potential. We offer additional, detailed implications for scholars and managers in our “Discussion” section.

Theoretical Background

The first trading day presents a pivotal opportunity for IPO firms to secure a substantial influx of capital. Consequently, firms seek to reduce information asymmetry with investors to capture as much of this first-day momentum as possible (Beatty and Ritter 1986). Before an IPO, private firms are typically subject to higher information asymmetry than public firms because investors have relatively limited access to information about private firms (Ritter 2011), leading to uncertainty about their potential. Public firms communicate with investors during regularly scheduled earnings calls, whereas private firms lack such channels. Thus, it becomes essential for IPO firms to reduce information asymmetry and lessen investor uncertainty so that they can be evaluated positively about their future prospects.

Information asymmetry creates an adverse selection problem, whereby the party with less information is unable to verify the other party's true potential (Akerlof 1978). A firm can offer irreversible investments and actions to signal its performance potential and future outlook (Spence 1973). Credible signals, which offer diagnostic information about a firm's potential, are costly and prevent firms with low performance potential from trying to signal deceptively. One credible signal of a firm's innovation competence is its introduction of new products, as it is a costly endeavor (e.g., Simeth and Cincera 2016). In addition, the timing of the signal is also crucial in reducing uncertainty and information asymmetry (e.g., Bafera and Kleinert 2023; Connelly et al. 2011; Islam, Fremeth, and Marcus 2018). Indeed, investors rely on the timing of a firm's NPIs as a signal to infer unobservable firm capabilities (Moorman et al. 2012). Investors observe the timing patterns of a firm's NPIs, which are observed from publicly available press releases that enable investors to track and assess historical information (e.g., Godsall et al. 2025; Salmon and Stokes 2010) and are considered on the initial trading day to influence the closing price as well as the number of shares outstanding and thus the IPO value. Therefore, we expect the IPO firm's NPI temporal patterns to provide important information to assess the firm's value.

Research on the temporal disposition of firm actions indicates that pacing style is a crucial component of firm performance (e.g., Chen and Nadkarni 2017; Mohammed and Nadkarni 2011). Pacing style is a continuous allocation of efforts leading to a deadline (Gevers, Mohammed, and Baytalskaya 2015), and we argue that an IPO firm's NPI pacing style represents the firm's NPI efforts prior to the IPO deadline. As we consider the temporal pacing of a firm's NPIs, the institutional investors we surveyed also raised several timing-related aspects they found important in assessing an IPO firm, including “how long ago (or recent) they launched new products before their IPO,” “how regularly they launch new products,” and “how they are persistent in caring about innovation” (for more examples, see Web Appendix B). Accordingly, across theory and our survey, three key temporal characteristics emerge: recency, dispersion, and asymmetry. Importantly, we expect these three temporal patterns of an IPO firm's NPIs to reveal differential relationships with IPO performance.

Temporal Patterns of NPIs

We highlight the nature of NPI recency, dispersion, and asymmetry using an illustrative example in Figure 1, where four firms each introduce four NPIs prior to their IPOs. First, recency is distinct from dispersion and asymmetry in its focus on the time for a single product (i.e., the last product prior to IPO), whereas the other two temporal patterns involve multiple NPIs. As shown in Figure 1, Firms A, C, and D have higher recency than Firm B, with their last NPI (NPI_L) closer to the IPO date. Firm A has high dispersion, whereas Firm B has low dispersion. Finally, considering the continuum of asymmetry, Firm C has high asymmetry (left-skewed) and Firm D also has high asymmetry (right-skewed), whereas Firms A and B have low asymmetry.

Figure 1.

Illustration of Recency, Dispersion, and Asymmetry of New Product Introductions.

The recency of an IPO firm's last NPI before its IPO signals important information to investors. Theoretically, recency is indicative of a firm's deadline-action pacing style (Chen and Nadkarni 2017; Gevers and Peeters 2009). Firms with a deadline-action pacing style tend to concentrate their efforts close to deadlines, becoming fully energized and motivated primarily when deadlines approach (Mohammed and Nadkarni 2011). In the context of NPIs prior to an IPO, when the most recent NPI occurs close to the IPO date, investors may perceive this as deadline-oriented behavior (Gevers and Peeters 2009). While deadline-action pacing can create an impression of productivity before an IPO deadline, it may also suggest the firm's prioritization of short-term-oriented innovativeness over sustainable innovation capabilities.

Investors may also perceive the NPI very close to the IPO date as an opportunistic attempt to boost short-term performance metrics. Research shows that investors tend to discount firms they suspect of artificial performance enhancement (Mizik and Jacobson 2007; Teoh, Welch, and Wong 1998). A product introduction just prior to an IPO may be interpreted as an attempt to artificially inflate the firm's growth potential, which may lead investors to be more skeptical of the firm's valuation. Indeed, an institutional investor we surveyed indicated, “If it is right before the IPO, I might be suspicious that firms are introducing attention-grabbing products just to attract attention before the IPO.” Thus, we hypothesize as follows:

H₁: Recency of the last NPI before IPO is negatively associated with IPO value.

The dispersion of NPIs before a firm's IPO also serves as a critical signal as investors assess the firm's performance potential. Findings from our survey of institutional investors aligns with the importance of dispersion, emphasizing that “how regularly they launch new products” and “consistency” of NPIs are important to investors when evaluating IPO firms. From the temporal disposition theoretical perspective, dispersion reflects the spread of activities in the pacing style before a deadline (Gevers and Peeters 2009; Mohammed and Harrison 2013). In our context, dispersion captures the extent to which NPIs are distributed over time before the IPO date. We expect NPI dispersion to play a critical role because it can signal a well-managed, strategically planned innovation trajectory to investors. Indeed, firms with a balanced time perspective tend to be more effective in organizational activities (Ancona, Okhuysen, and Perlow 2001). Balanced pacing of strategic activities allows firms to allocate attention and resources more efficiently across multiple time horizons (Nadkarni and Chen 2014). A pattern of dispersed NPIs can signal a temporally balanced resource allocation, indicating that the firm values both short-term goals and long-term planning. Thus, NPI dispersion before the IPO will enable firms to signal a healthy outlook to investors.

In addition, a dispersed pattern of NPIs signals the firm's innovation capabilities, indicating that it is consistently investing in organic growth. NPIs reflect a firm's ability to sense and respond to market changes (e.g., Narver, Slater, and MacLachlan 2004; Sorescu, Chandy, and Prabhu 2003). The dispersion of NPIs before the IPO signals the firm's systematic approach of consistent investment in innovation and its future, thereby reducing investor uncertainty surrounding the IPO firm's performance potential. Indeed, one institutional investor we surveyed indicated that “companies that do not regularly introduce new products in the years preceding the IPO … might just strategically introduce products to grab attention.” Research also reveals that investors indeed value firms that demonstrate consistent innovation capabilities (Deeds, DeCarolis, and Coombs 2000). Therefore, we hypothesize as follows:

H₂: Dispersion of NPIs before IPO is positively associated with IPO value.

Asymmetry is another important aspect of pacing style and reflects effort exerted over a continuum toward a deadline (Gevers, Mohammed, and Baytalskaya 2015). Unlike dispersion, asymmetry reflects the extent to which events are clustered at one end of the continuum. Reiterating the importance of asymmetry, one institutional investor reinforced the need to assess the IPO firm's timing of NPIs to ascertain “how they are persistent in caring about innovation.” With high asymmetry, most events will be clustered either very close to a deadline or far in advance of it (Gevers and Peeters 2009; Mohammed and Harrison 2013). In our context, this continuum represents NPIs before an IPO, ⁶ with asymmetry indicating the degree to which NPIs before an IPO are clustered near or far from the IPO date. We note that elevated asymmetry could occur at either end of the continuum (see Firms C and D in Figure 1).

When NPIs are disproportionately clustered closer to the IPO date (i.e., left-skewed on the timeline), investors may perceive that the firm is introducing several products just prior to IPO to impress investors and boost short-term performance. Research also demonstrates that innovation behaviors in the “last mile” may signal inferior quality (Hermosilla 2021). Although this asymmetric pattern of NPIs closer to IPO is distinct from recency, where recency focuses on a single product rather than multiple products, each pattern will likely be interpreted as an attempt to accrue short-term gains. Furthermore, this emphasis on multiple NPIs close to the IPO date may signal a lack of long-term strategic planning, which increases investor concerns about the IPO firm's performance potential.

For an asymmetric pattern at the other end of the continuum, when NPIs are disproportionately clustered further from the IPO (i.e., right-skewed on the timeline), it may signal that the firm has exhausted its innovation pipeline. As one investor indicated, “If the new products have been introduced way before the IPO, then I would think the firm is not that innovative.” Alternatively, it may signal that the firm has an established innovation capability, elevating investor confidence in its performance outlook. Because firms’ prior knowledge and experience contribute to innovation capabilities in subsequent periods (Cohen and Levinthal 1990), such an asymmetric pattern is likely to demonstrate the firm's confidence about its innovation potential. This may reduce investor concerns about the IPO firm. Thus, we hypothesize as follows:

H₃: Asymmetry of NPIs disproportionately clustered close to IPO is negatively associated with IPO value.

With the NPI temporal patterns serving as a credible signal to investors, both the signaling environment and the signal message are likely to influence the effectiveness of the NPI temporal patterns. Therefore, we investigate the moderating roles of industry growth and product innovativeness as additional, important information in shaping the relationships between NPI temporal patterns and IPO value.

Moderating Effects of Industry Growth and Product Innovativeness

Moderating effects on NPI recency

Although an NPI shortly before an IPO is expected to reduce IPO value, we anticipate that this negative relationship will attenuate when IPO firms operate in industries with higher growth. Importantly, introducing a product shortly before an IPO is likely to break through the information clutter within a high-growth industry and signal important information to investors. When firms introduce a new product in a high-growth industry, it generally reduces investor uncertainty. Investors recognize that firms operating in high-growth markets can tap into rising market demand (Aldrich 2008). As one investor from our survey indicated, “In a fast-growing industry, I would be more generous in terms of evaluating the new product introductions of the IPO company.” Furthermore, investors also recognize that market opportunities are often subject to time-sensitive windows in high-growth industries (Datar et al. 1997). Thus, investors may interpret a more recent NPI as indicative of a firm's market-relevant timing in high-growth industries, rather than viewing greater recency with suspicion. Instead, investors may view recency in a high-growth industry as a signal of the firm's ability to strategically align with market opportunities. In contrast, firms in low-growth industries will face heightened challenges in expanding and increasing profits. Fewer market opportunities exist for firms operating in industries with lower growth, which will intensify investors’ concerns. Therefore, we predict:

H₄: Industry growth weakens the negative association between NPI recency and IPO value.

We also expect product innovativeness to mitigate the negative relationship between NPI recency and IPO value. First, introducing a new product shortly before IPO will reduce investor suspicion about the firm's innovation and market capability if the firm has greater product innovativeness. A firm's product innovativeness signals its technological capabilities and market foresight (Chaney, Devinney, and Winer 1991), which can counterbalance concerns about timing when a firm introduces a new product close to its IPO date. Highly innovative products require sophisticated knowledge, specialized capabilities, and significant investments in research and development (R&D) that cannot be hastily or easily assembled (Chandy and Tellis 2000) and will likely attenuate investors’ suspicions of short-term priorities when the firm introduces a new product shortly before its IPO. Second, highly innovative new products generate disproportionately higher returns than incremental innovations (Sorescu, Chandy, and Prabhu 2007). It is therefore possible that the positive signaling role of a firm's high product innovativeness may overshadow the negative signaling role of its NPI recency prior to its IPO. Thus, we predict:

H₅: Product innovativeness weakens the negative association between NPI recency and IPO value.

Moderating effects on NPI dispersion

We expect industry growth to influence investor assessments, shifting the relationship between a firm's NPI dispersion and IPO performance. Although high-growth industries offer rich opportunities to firms, the technological advances, regulatory shifts, and heterogeneous customer demand that occur in high-growth industries tend to produce more uncertainty (McCarthy et al. 2010) and thus greater environmental noise. This environmental noise impairs signal efficacy (e.g., Park and Mezias 2005). With more noise in the environment, it becomes increasingly challenging for investors to observe and discern signals (Connelly et al. 2011; Steigenberger and Wilhelm 2018), especially when multiple signals are present (Bafera and Kleinert 2023) and when these signals occur over time.

Given that high NPI dispersion reflects a temporal pattern of new product introductions spread over time, this signal becomes less distinctive when the firm operates in a high-growth industry. In contrast, with less noise in low-growth industries, the dispersion of NPIs has the potential to convey a stronger signal of the firm's capability, as there is less noise to impede the signal. In a lower-growth industry, a firm's competitive advantages are more observable to investors (Park and Mezias 2005). Thus, introducing new products in a dispersed pattern will be more observable in a lower-growth industry, which reinforces the signal of an intentional pursuit of innovation. We therefore hypothesize:

H₆: Industry growth weakens the positive association between NPI dispersion and IPO value.

With dispersed NPIs, a firm's higher product innovativeness weakens the positive relationship with IPO value. A dispersed pattern of introducing new products signals the firm's innovation capabilities, highlights its strategic foresight and innovative planning, and shows it is consistently investing in innovation. This dispersed pattern of NPIs offers a valuable signal to investors, thereby reducing information asymmetry when assessing IPO firms. However, when firms introduce highly innovative products, it provides another powerful signal of their capabilities (Sorescu and Spanjol 2008), potentially lessening investors’ sole reliance on temporal dispersion to reduce uncertainty in their evaluation.

Investors may also view a tension between a firm's dispersed NPIs and their highly innovative products. Highly innovative products require the firms’ concentrated focus and a greater level of resources (O’Reilly and Tushman 2004), demanding dedicated attention to technological development, market creation, and organizational learning that may be incompatible with maintaining a dispersed pattern of multiple product introductions over time. Research also notes that breakthrough innovations often require firms to temporarily divert resources from other initiatives (Christensen and Bower 1996). For firms with greater dispersion, investors may also place less emphasis on the firm's product innovativeness because they recognize that more innovative products require creativity, which does not result from a synchronous schedule of time and planned efforts (Amabile 1996). Thus, we hypothesize as follows:

H₇: Product innovativeness weakens the positive association between NPI dispersion and IPO value.

Moderating effects on NPI asymmetry

While firms in high-growth industries face the challenge of reduced efficacy across multiple signals due to environmental noise (Bafera and Kleinert 2023), this is heightened for multiple signals over time. For firms that disproportionately introduce new products closer to the IPO date in a high-growth industry, investors may be better able to discern these signals, given their proximity to the IPO. Moreover, this left-skewed asymmetric pattern may signal market-relevant introductions that mitigate investor concerns about multiple introductions close to the IPO. In contrast, in a high-growth industry where new products are disproportionately clustered further from the IPO (i.e., right-skewed asymmetry), investors will likely face more noise that reduces signal efficacy. If investors do observe the disproportionate introduction of new products further from the IPO date, then this inaction closer to the IPO in a high-growth environment rich with market opportunities is likely to generate concerns that the firm's knowledge about consumers or competitors has become outdated.

In contrast, low-growth industries afford advantages when NPIs are disproportionately clustered further from the IPO. When an industry has lower growth, there is less noise in the market environment and advancements in new technologies or demand are also lagging (McCarthy et al. 2010). Thus, firms may maintain market foresight on consumers and competitors even with NPIs clustered further from the IPO. However, a left-skewed asymmetric pattern of NPIs, where NPIs are clustered closer to the IPO, will reduce IPO value to a greater degree in a lower-growth industry, because it reinforces investor concerns of multiple NPIs close to IPO and elevates concerns that the products are introduced in a market with potentially limited room for success. Thus, we hypothesize:

H₈: Industry growth weakens the negative association between asymmetry of NPIs disproportionately clustered close to the IPO and IPO value.

We also expect the negative association between a left-skewed NPI asymmetry pattern and IPO value to weaken for firms with higher product innovativeness. When a firm with high product innovativeness introduces new products disproportionately closer to its IPO, investors may interpret this timing differently than they would for firms with less product innovativeness. A high degree of innovativeness signals that the firm strategically timed its NPIs to align with technological readiness and market windows (Bayus, Jain, and Rao 1997), so investors would view an asymmetric pattern of NPIs clustered closer to the IPO more favorably.

With multiple products introduced in a clustered pattern closer to the IPO, the quality-signaling effect of product innovativeness can counterbalance concerns about asymmetry. High-quality signals reduce consumers’ skepticism about firm motivations (Kirmani and Rao 2000), which may reduce investor skepticism about an IPO firm's asymmetric pattern of new products introduced close to the IPO date. Similarly, investors may be less concerned about opportunistic timing when the products demonstrate significant innovation capabilities. Innovative products require substantial development time and typically undergo more extensive testing (e.g., Griffin 1997), making it less likely that they were hastily introduced regardless of their asymmetrical temporal proximity to the IPO.

H₉: Product innovativeness weakens the negative association between asymmetry of NPIs disproportionately clustered close to the IPO and IPO value.

We depict the hypothesized relationships in Figure 2.

Figure 2.

Conceptual Framework.

Methodology

Data

To test our hypotheses, we assembled data from multiple sources, including the Refinitiv New Issues database, Factiva, the U.S. Patent and Trademark Office (USPTO) patent dataset, FactSet, and the U.S. Securities and Exchange Commission’s Electronic Data Gathering, Analysis, and Retrieval system (EDGAR). First, we obtained data for IPOs between January 1, 2006, and December 31, 2023, from the Refinitiv New Issues database. Consistent with prior IPO research (e.g., Ritter and Welch 2002; Saboo and Grewal 2013), we excluded closed-end funds, spin-offs, limited partnerships, leveraged buyouts, rights issues, real estate investment trusts, unit offerings, American depositary receipts, American depositary shares, firms with an IPO offer price below $5, and IPOs with total proceeds less than $5 million. ⁷ We retained only the first IPO of a firm if the firm has gone through multiple IPOs. To account for any incorrect or missing values, as highlighted in prior IPO research (e.g., Loughran and Ritter 2004), we cross-checked each firm's financial information and supplemented it using data from the Center for Research in Security Prices (CRSP) and corrections from Ritter (2019). We also cross-checked the information with the IPO prospectus, which was collected from the 424B statement (prospectus form under the Securities Exchange Act of 1933) available on the EDGAR.

We collected information on a firm's NPIs from the firm's press releases, following Warren and Sorescu (2017). We retrieved press releases on NPIs for each firm from Factiva and analyzed the content of these press releases. Following previous research on press releases (Pan et al. 2020; Rindova, Ferrier, and Wiltbank 2010), we limited our sample to articles with sources such as Business Wire, the Associated Press, and PR Newswire to ensure the use of press releases and not merely media coverage. ⁸ In Factiva, “new products/services” tags are associated with articles on the introduction, preview, or announcement of a new product or service. These include product/service enhancements, improvements, and new versions but exclude products in early developmental stages or news on facility openings. To identify NPIs as opposed to new product announcements, we identified the distinct new products/services discussed in the press release that were classified as “new products/services.” We content-analyzed each press release with such a tag to ensure the product/service is an introduction and not a preannouncement of a forthcoming product. We retained only the first press release about each new product/service to ensure the appropriate NPI date. We measured the NPI timing variables using a five-year window preceding a firm's IPO to ensure a sufficient time frame for measuring NPI dispersion. Further, this was supported by data from our survey of institutional investors who indicated they rarely track information about the IPO firm beyond five years before IPO.

We also collected data from Compustat, USPTO, EDGAR, and Refinitiv New Issues to construct the measures for industry growth, product innovativeness, and other control variables. Moreover, to ensure consistency in measurement, we dropped firms that are younger than five years, given our time frame. Following research on temporal aspects of innovation (Sharma, Saboo, and Kumar 2018), we excluded firms with zero or one NPI; we also excluded firms with two or fewer NPIs because they do not allow us to calculate NPI dispersion and asymmetry measures. ⁹ After compiling these data and accounting for missing values, the final sample resulted in 298 IPOs. We collected information about private firms from FactSet.

Measures

IPO value

Following extant research in marketing (Saboo, Kumar, and Anand 2017; Xiong and Bharadwaj 2011), we measured the IPO value of firm i (IPOV_i) as the firm's market capitalization at the end of the first trading day as follows:

$${\mathrm{IPOV}}_{\mathrm{i}}=\mathrm{closing}{\mathrm{price}}_{\mathrm{i}}\times \mathrm{number}\mathrm{of}\mathrm{shares}{\mathrm{outstanding}}_{\mathrm{i}}.$$

(1)

We used the log-transformed measure of IPO value that was measured in hundreds of millions.

Recency of NPI

We measured recency by the number of days between the date of IPO and the introduction date of the most immediate NPI just before the IPO. For ease of interpretation, we multiplied this number by −1 so that firms with higher recency are those whose last NPI before IPO occurs closer to the IPO day.

Dispersion of NPI

We measured dispersion using the standard deviation of the number of days between each NPI date and the IPO date, within five years before the IPO (see Equation 2).

$${\mathrm{dispersion}}_{\mathrm{i}}=\sqrt{{\displaystyle \frac{1}{{\mathrm{n}}_{\mathrm{i}}}\sum _{\mathrm{j}=1}^{\mathrm{n}}{({\mathrm{daysdiff}}_{\mathrm{ij}}-{\mu }_{\mathrm{i}})}^2}},$$

(2)

where n_i is number of NPIs of each firm i, daysdiff_ij represents the number of days between product j's introduction date and the IPO date of firm i, and μ_i is the mean of daysdiff_j for firm i. Higher dispersion indicates that NPIs are more spread out before an IPO.

Asymmetry of NPI

We measured asymmetry using the skewness of the number of days between each NPI date and the IPO date, within five years before IPO (see Equation 3). ¹⁰ For ease of interpretation, with our hypotheses focusing on left-skewed asymmetry, we multiplied the skewness measure by −1 so that a negative sign indicates that new products are disproportionately introduced further from the IPO (i.e., right-skewed), whereas a positive sign indicates they are disproportionately introduced closer to the IPO (i.e., left-skewed).

$${\mathrm{asymmetry}}_{\mathrm{i}}=-{\displaystyle \frac{1}{{\mathrm{n}}_{\mathrm{i}}}\sum _{\mathrm{j}=1}^{\mathrm{n}}{\left({\displaystyle \frac{{\mathrm{daysdiff}}_{\mathrm{ij}}-{\mu }_{\mathrm{i}}}{{\sigma }_{\mathrm{i}}}}\right)}^3,}$$

(3)

where σ is the standard deviation of the NPI_ij of firm i and other symbols are the same as Equation 2.

Industry growth

We measured industry growth based on the five-year average growth in net income in each industry (Keats and Hitt 1988), as specified in Equation 4:

$${\mathrm{y}}_{\mathrm{mt}}={\lambda }_0+{\lambda }_1\mathrm{t}+{\mathrm{e}}_{\mathrm{mt}},$$

(4)

where y_mt is the log-transformed industry sales of industry m for year t, and e_mt is the residual term. We used the antilog of the regression slope coefficient from Equation 4 as the industry growth rate; all firms in industry m have the same value for this growth rate.

Product innovativeness

To measure the IPO firm's product innovativeness, we conducted text analysis of the firm's press releases of NPIs and relied on McKenny et al.'s (2018) dictionary that includes words and phrases such as “next generation,” “transformation,” and “revolutionary.” This process requires each press release to be first cleaned to remove all nontextual words and symbols. We then calculated the proportion of words and phrases in the press release included in the dictionary to the total number of words in the press release. Because phrases may not appear verbatim in a sentence, we allowed phrases (e.g., “technologically advanced” and “next generation”) to be separated by a maximum of three words. To minimize measurement error, words and phrases attached to negative words (e.g., “never,” “not,” and “no”) are not counted, and keyword stems (e.g., “transform”) were used to strip the search term to its root, enabling the algorithm to capture all morphological variants (e.g., “transformation,” “transforming,” and “transforms”). We averaged these proportions across press releases for each firm, to create the measure for each firm's product innovativeness.

Control variables

Following extant IPO literature, we included several IPO-specific control variables (e.g., Luo 2008; Xiong and Bharadwaj 2011). Because venture capitalists may reduce information asymmetry through certification and monitoring (Megginson and Weiss 1991), we used a dummy variable to represent whether a firm used venture backing before the IPO (1 if a firm used venture capital backing for its IPO, 0 otherwise). In addition, IPO firms with reputable underwriters are likely to reduce investor uncertainty because the underwriter's reputation represents how competent the IPO firms are to the underwriters. Thus, we compiled a list of underwriters for each IPO from the firm's IPO prospectus and utilized the ranking of underwriters (Carter and Manaster 1990; Loughran and Ritter 2004). The scores range from 0 to 9, with higher scores representing a higher underwriter reputation. Following Liu and Ritter (2011), we used a dummy variable to code underwriter reputation, measured by the presence of a reputable lead underwriter for an IPO (1 if the score of the lead underwriter is 8 or above, 0 otherwise). Moreover, because a hot IPO market indicates a greater number of IPOs competing for investor attention (Busaba, Benveniste, and Guo 2001), we included IPO market intensity, which is measured by the number of IPOs offered in the same month as the focal firm's IPO.

We also accounted for several firm-level variables. First, following Heeley, Matusik, and Jain (2007), we included patent stock, measured as the log-transformed total number of patents filed by the firm prior to its IPO. Second, we controlled for founder characteristics using founder education, a critical signal that investors use to assess new ventures (Ko and McKelvie 2018). Following previous research (Dencker and Gruber 2015; Ko and McKelvie 2018), we measure founder education based on each founding member's highest completed degree, which ranges from high school to postgraduate, by searching the founders on LinkedIn and the company's official websites. Specifically, this variable takes the value of 1 if high school, 2 if undergraduate degree (e.g., BA, BS), 3 if postgraduate degree (e.g., MA, MBA, PhD, MD), and 2.47 (which is the average founder education in our sample) if the highest education of the founder is unspecified. When multiple founders exist within a firm, we averaged the data across founders to measure this variable. Third, we account for a firm's risk-taking propensity. Using each firm's 424(b) statement (i.e., the final IPO prospectus) from EDGAR, we use text analysis on the firm's IPO prospectus, which best represents the IPO firm's core information, practices, and plans. We used McKenny et al.'s (2018) dictionary on risk-taking, which includes words and phrases such as “no place to hide,” “risk taking,” or “ambitious.” Fourth, we included firm age to control for the extent to which investors perceive an older firm as more attractive, with older firms having better efficiency and stronger relationships (Saboo and Grewal 2013). We used the log-transformed number of years between the year the firm was founded and the year of its IPO for firm age. Finally, we controlled for IPO years to account for year fixed effects.

We also controlled for important industry-level variables, using dummy variables to control for industry-level effects indicating whether the IPO firm is in a high-technology, pharmaceutical, or service industry. Moreover, we controlled for industry dynamism, which reflects industry volatility and is measured by the antilog of standard error of each regression slope coefficient in Equation 4 (Keats and Hitt 1988).

Table 1 summarizes the measures, Table 2 provides descriptive statistics, and Table 3 shows the correlation matrix.

Table 1.

Summary of Measures and Sources.

Variable	Measure	Source
IPO value	Market capitalization at the end of the first trading day (i.e., closing price multiplied by the number of shares outstanding on the first trading day). For ease of interpretation, we divided this measure by 100,000,000. Log-transformed due to the right-skewness of the measure.	Refinitiv; CRSP
Recency	Number of days between the date of the last NPI before an IPO (within five years before the IPO) and the IPO date. For ease of interpretation, multiplied by −1.	Factiva; Refinitiv
Dispersion	Standard deviation of days of differences between each NPI (within five years before the IPO) and the IPO date.	Factiva; Refinitiv
Asymmetry	Measured on a skewness continuum of days of differences between each NPI (within five years before the IPO) and the IPO date. For ease of interpretation, skewness multiplied by −1.	Factiva; Refinitiv
Industry growth	Five-year growth of industry net income. Regressing industry net income for each year t on time t, the antilog of the resulting regression slope coefficient is used as industry growth in year t in each industry. The industry growth measure is based on the firm’s four-digit SIC code.	Compustat
Product innovativeness	The portion of words and phrases included in the press releases on NPIs that are included in the dictionary for innovativeness developed by McKenny et al. (2018). For each firm, product innovativeness proportions are averaged across press releases.	Factiva; McKenny et al. (2018)
Venture backing	Dummy variable; 1 if the IPO firm uses venture capital backing before the IPO, 0 otherwise.	Refinitiv
Underwriter reputation	Dummy variable; 1 if the lead underwriter is reputable (8 or above in Carter–Manaster ranking), 0 otherwise.	Carter and Manaster (1990); Refinitiv
Firm age	Log-transformation of number of years between the year founded and the year of IPO.	Refinitiv; manual web search
IPO market intensity	Number of IPOs offered in the current month when the firm's IPO is offered.	Refinitiv; CRSP
Patent stock	Sum of the number of patents that a firm filed before the IPO issue date from the date founded.	USPTO
Founder education	Average of founder(s)'s education, where 2.47 (average founder education score in our sample) indicates highest education of the founder is unspecified; 1 if high school; 2 if undergraduate degree (e.g., BA, BS); and 3 if postgraduate degree (e.g., MBA, PhD, MD).	LinkedIn; company website; online search
Risk-taking propensity	The portion of words and phrases included in the IPO prospectus that are included in the dictionary for risk-taking developed by McKenny et al. (2018).	EDGAR; McKenny et al. (2018)
Industry dynamism	Five-year instability of industry net income. Regressing industry net income for each year t on time t, measure is the antilog of the standard error of each regression slope coefficient in year t in each industry (based on firm's four-digit SIC code).	Compustat
High-tech industry	Dummy variable; 1 if the industry is high-tech industry (based on the high tech industry variable on Refinitiv), 0 otherwise.	Refinitiv
Pharmaceutical industry	Dummy variable; 1 if the industry is pharmaceutical industry (based on the Reference Data Business Classification business sector variable being equal to “Pharmaceuticals & Medical Research” on Refinitiv), 0 otherwise.	Refinitiv
Service industry	Dummy variable; 1 if SIC first-digit code is 7 or 8, 0 otherwise.	Compustat

Table 2.

Summary Statistics.

Variable	Mean	SD	Min	Max
IPO value	21.22	47.62	.28	542.65
Recency	−351.92	361.91	−1,630	−1
Dispersion	372.73	170.79	11.55	856.3
Asymmetry	.03	.73	−2.11	1.99
Industry growth	1.17	.31	.1	3.85
Product innovativeness	.01	.01	0	.04
Venture backing	.64	.48	0	1
Underwriter reputation	.72	.45	0	1
Firm age	18.11	24.14	5	186
IPO market intensity	15.87	9.11	1	55
Patent stock	19.01	72.04	0	700
Founder education	1.95	1.1	0	3
Risk-taking propensity	0	0	0	0
Industry dynamism	1.14	.34	1	4.85
High-tech industry	.73	.45	0	1
Pharmaceutical industry	.06	.24	0	1
Service industry	.39	.49	0	1

Table 3.

Correlation Matrix.

Variable	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
1. IPO value	1
2. Recency	.11	1
3. Dispersion	.1	.44	1
4. Asymmetry	0	−.01	−.04	1
5. Industry growth	−.07	−.04	−.01	.09	1
6. Product innovativeness	.03	.06	−.04	.07	.02	1
7. Venture backing	−.19	.01	0	.05	0	.03	1
8. Underwriter reputation	.05	.02	−.02	0	−.08	−.05	0	1
9. Firm age	.25	−.01	.07	−.09	.03	−.06	−.46	.08	1
10. IPO market intensity	.21	0	.03	.16	.03	.01	−.08	−.08	.05	1
11. Patent stock	0	−.01	−.04	−.03	−.01	.05	.36	−.13	−.26	−.11	1
12. Founder education	−.13	.08	.1	.06	−.07	−.04	.34	.08	−.32	−.09	.17	1
13. Risk-taking propensity	.08	.05	.02	.04	−.06	.1	.11	.03	−.07	.11	−.07	.02	1
14. Industry dynamism	−.02	−.02	.05	.05	−.08	.01	.01	−.15	.02	−.04	−.01	−.01	−.03	1
15. High-tech industry	−.19	−.06	−.08	.04	−.01	.08	.34	−.05	−.31	−.13	.32	.17	−.13	−.05	1
16. Pharmaceutical industry	−.16	.08	−.09	.05	.05	.20	.08	.01	−.07	−.04	.06	−.05	.1	−.03	.13	1
17. Service industry	−.01	−.16	.03	−.02	−.04	.02	.11	−.02	−.03	−.12	0	−.03	−.12	.02	.24	−.21	1

Notes: Those with p < .05 are in bold.

Model Specification and Estimation

IPO value is only realized if a private firm goes public (Pagano, Panetta, and Zingales 1998). Therefore, we first account for the incidence of an IPO, which addresses the selection issue relevant to a firm's strategic decision to go public. Accordingly, we code the variable IPO as taking a value of 1 if the firm goes public and 0 otherwise. By endogenizing the decision to go public, we account for the potential bias in the regression parameters in the IPO value equation, which is observable only for an IPO firm.

Following Xiong and Bharadwaj (2011) and Gulati and Higgins (2003), we account for selection bias using a two-stage Heckman estimation (Heckman 1979). The selection model accounts for the incidence of IPO, conditional on which the post-IPO value outcome is observed. In the first stage, using a more extensive set of 281,184 firms that includes all private firms and the IPO firms from 2006 to 2023, we use a probit regression to estimate the probability of IPO incidence. Thus, we model a firm's decision to go public using a probit specification as follows:

$${\mathrm{IPOPROB}}_{\mathrm{i}}={\gamma }^{\mathrm{\prime }}{\mathrm{W}}_{\mathrm{i}}+{\mathrm{u}}_{\mathrm{i}},$$

(5)

$${\mathrm{IPO}}_{\mathrm{i}}=1\mathrm{if}{\mathrm{IPOPROB}}_{\mathrm{i}}>0,$$

(6)

where IPOPROB_i is the underlying probability of the observed outcome IPO_i for firm i.

Following Gulati and Higgins (2003) and Wies, Moorman, and Chandy (2023), we include firm age and geographical location (measured by states) in the W_i vector to estimate the probability specified in Equation 5. We include firm age because older, more established firms are more likely to go public (Chemmanur, He, and Nandy 2010). Additionally, we include industry fixed effects in the W_i vector because firms in particular industries may be more likely to go public. Heckman models tend to perform particularly well for estimation with at least one variable that works as an instrument, satisfying the exclusion restriction. We use the firm's geographical location (U.S. state) as an instrument that satisfies the exclusion restriction (see Winship and Mare 1992). A firm's geographical location is related to its decision to go public (e.g., Gulati and Higgins 2003; Wies, Moorman, and Chandy 2023), because regional differences in the availability of resources and capital can influence the decision to go public, thereby satisfying the relevance criterion. In addition, geographic location determines a firm's decision to go public because the costs of generating information will vary across different locations (e.g., Loughran 2008). Supporting this view, disclosure and regulatory incentives (e.g., Shi, Pukthuanthong, and Walker 2013) also impact firms’ decisions to file for an IPO in particular geographic locations. Moreover, because there is no conceptual reason to expect geography to influence the IPO firm's performance directly, it satisfies the exclusion restriction. In line with Wies, Moorman, and Chandy (2023), we compare the model fit between the model with and without geographic location to support the selection of geographic location as an instrumental variable. We find that the fit of our selection model of going public improves when geographic location is included (Akaike information criterion [AIC] = 4,836.36; Bayesian information criterion [BIC] = 5,715.80) compared to when it is excluded (AIC = 5,214.12; BIC = 5,750.97), satisfying the relevance criteria. There was no improvement in model fit when including geographic location in Equation 7 (AIC = 818.70; BIC = 1,040.53) compared with excluding it from Equation 7 (AIC = 809.08; BIC = 949.57).

Based on the estimates of parameters from Equation 5, we calculate the inverse Mills ratio $\lambda =\frac{\phi ({\gamma }^{\mathrm{\prime }}{\mathrm{W}}_{\mathrm{i}})}{\varphi ({\gamma }^{\mathrm{\prime }}{\mathrm{W}}_{\mathrm{i}})}$ , where φ is the probability density function, and ϕ is the cumulative density function of the normal distribution. We include the inverse Mills ratio in the second stage, in Equation 7, to control for potential self-selection bias. If a firm goes public (i.e., the selection occurs), IPO value is observed; if a firm remains private, it is not observed. We model IPO value conditional on the occurrence of IPO as follows:

$${\mathrm{IPOV}}_{\mathrm{i}}=\alpha +{\mathrm{X}}_{\mathrm{i}}^{\mathrm{\prime }}\beta +{\mathrm{Z}}_{\mathrm{i}}^{\mathrm{\prime }}\mathrm{K}+{\epsilon }_{\mathrm{i}},\mathrm{if}{\mathrm{IPO}}_{\mathrm{i}}=1,$$

(7)

where IPOV_i is the IPO value of firm i, α is the intercept, β is the vector of regression coefficients, X_i is the vector of independent and moderating variables along with associated interaction effects, Z_i is the vector of control variables while K is the vector of control variable coefficients, and the error term ɛ_i is assumed to be independently and normally distributed. We standardize the measures to enhance interpretability and to account for potential multicollinearity. We estimate the model using robust standard errors.

Furthermore, a firm could be setting its temporal NPI patterns and product innovativeness to affect the IPO's success, making the effects of the primary variables endogenous. Thus, we account for the potential endogeneity of recency, dispersion, asymmetry, and product innovativeness using a two-stage control function approach following previous research (e.g., Petrin and Train 2010). In the first stage, we regress the endogenous variables on a set of exogenous variables to estimate the residuals. We use two sets of exclusion variables: industry-level averages of NPI recency, dispersion, asymmetry, and product innovativeness of all firms within the same industry excluding the focal firm, as well as the averages of the firms with similar resources excluding the focal firm (e.g., Germann, Ebbes, and Grewal 2015; Sharma, Saboo, and Kumar 2018).

These exclusion variables meet the relevance criterion as industry peers often look to each other to guide their competitive strategies (e.g., Sridhar et al. 2016). Furthermore, because optimal levels of NPI temporal patterns before IPO are unknown, firms may imitate or act opposite to similar firms. Similarly, there is no set standard for the optimal level of product innovativeness before IPO. Indeed, firms tend to resolve uncertainty in their decision-making by either imitating or contrasting their actions to their peers (Spender 1989), as suggested by mimetic isomorphism (DiMaggio and Powell 1983) and literature on competitive actions and responses (e.g., Montgomery, Moore, and Urbany 2005). Therefore, these variables also satisfy the exclusion criterion because it is unlikely that comparable firms (relative to the focal firm) would collectively coordinate in making decisions about their NPI recency, dispersion, asymmetry, and product innovativeness in a manner that influences the focal firm's IPO value (e.g., Germann, Ebbes, and Grewal 2015).

Thus, for the first set of exclusion variables, we include industry-level averages for recency, dispersion, asymmetry, and product innovativeness to account for industry norms. Specifically, all firms in the same industry (based on a two-digit Standard Industrial Classification [SIC] code), excluding the focal firm, are included in calculating the industry averages. For the second set of exclusion variables, we include similar-firm averages using patent stock to account for resource similarity. We rely on patent stock because patents provide a technological foundation for innovation. Firms without patents cannot follow similar innovation strategies as firms with patents; therefore, firms with patents are in one group, while firms without patents are in another group. With this approach, all firms that have patent stock, excluding the focal firm, are included in the measure of similar-firm averages when the focal firm has patents, whereas all firms without any patents, excluding the focal firm, are included in similar-firm averages when the focal firm does not have any patents. Our approach aligns with using multiple sets of instruments that account for multiple sources of variation (e.g., Mallapragada et al. 2025; Moon, Tuli, and Mukherjee 2023), and it is similar to other approaches that rely on industry-level and similar-firm averages used in previous research (e.g., Sharma, Saboo, and Kumar 2018). Thus, we accounted for the endogenous nature of the four variables using the following equations:

$${\mathrm{Recency}}_{\mathrm{i}}={\tau }^{\mathrm{Rec}}+\mathrm{IND}\mathrm{\_}{\mathrm{REC}}_{\mathrm{i}}^{\mathrm{\prime }}{\omega }^{\mathrm{Rec}}+\mathrm{SIM}\mathrm{\_}{\mathrm{REC}}_{\mathrm{i}}^{\mathrm{\prime }}{\theta }^{\mathrm{Rec}}+{\mathrm{Z}}_{\mathrm{i}}^{\mathrm{\prime }}{\pi }^{\mathrm{Rec}}+{\xi }_{\mathrm{i}}^{\mathrm{Rec}},$$

(8)

$${\mathrm{Dispersion}}_{\mathrm{i}}={\tau }^{\mathrm{Disp}}+\mathrm{IND}\mathrm{\_}{\mathrm{DISP}}_{\mathrm{i}}^{\mathrm{\prime }}{\omega }^{\mathrm{Disp}}+\mathrm{SIM}\mathrm{\_}{\mathrm{DISP}}_{\mathrm{i}}^{\mathrm{\prime }}{\theta }^{\mathrm{Disp}}+{\mathrm{Z}}_{\mathrm{i}}^{\mathrm{\prime }}{\pi }^{\mathrm{Disp}}+{\xi }_{\mathrm{i}}^{\mathrm{Disp}},$$

(9)

$${\mathrm{Asymmetry}}_{\mathrm{i}}={\tau }^{\mathrm{Asym}}+\mathrm{IND}\mathrm{\_}{\mathrm{ASYM}}_{\mathrm{i}}^{\mathrm{\prime }}{\omega }^{\mathrm{Asym}}+\mathrm{SIM}\mathrm{\_}{\mathrm{ASYM}}_{\mathrm{i}}^{\mathrm{\prime }}{\theta }^{\mathrm{Asym}}+{\mathrm{Z}}_{\mathrm{i}}^{\mathrm{\prime }}{\pi }^{\mathrm{Asym}}+{\xi }_{\mathrm{i}}^{\mathrm{Asym}},$$

(10)

$${\mathrm{Innovativeness}}_{\mathrm{i}}={\tau }^{\mathrm{Innov}}+\mathrm{IND}\mathrm{\_}{\mathrm{Innov}}_{\mathrm{i}}^{\mathrm{\prime }}{\omega }^{\mathrm{Innov}}+\mathrm{SIM}\mathrm{\_}{\mathrm{Innov}}_{\mathrm{i}}^{\mathrm{\prime }}{\theta }^{\mathrm{Innov}}+{\mathrm{Z}}_{\mathrm{i}}^{\mathrm{\prime }}{\pi }^{\mathrm{Innov}}+{\xi }_{\mathrm{i}}^{\mathrm{Innov}},$$

(11)

where Recency_i in Equation 8 is the NPI recency of focal firm i, Dispersion_i in Equation 9 is the NPI dispersion of focal firm i, Asymmetry_i in Equation 10 is the NPI asymmetry of focal firm i, and Innovativeness_i in Equation 11 is the product innovativeness of focal firm i. τ are model intercepts; IND_REC_i, IND_DISP_i, IND_ASYM_i, and IND_INNOV_i are vectors of the exclusion variables that include average (excluding the focal firm) industry-level NPI recency, dispersion, asymmetry, and innovativeness; SIM_REC_i, SIM_DISP_i, SIM_ASYM_i, and SIM_INNOV_i are vectors of the exclusion variables that include averages (excluding the focal firm) of recency, dispersion, asymmetry, and innovativeness of firms in the same patent stock group (i.e., with vs. without patent stock). ω and $\theta$ are the vectors of the regression coefficients, Z_i is the vector of control variables, and error terms ξ_i are assumed to be independently and identically distributed. We use the residuals from this estimation stage as additional control variables in Equation 7.

Results

Results for Equation 7 (i.e., IPO value) are presented in Table 4, whereas Web Appendix C presents first-stage results of the control function approach (i.e., Equations 8–11). Results show that NPI recency is negatively associated with IPO value (b = −1.477, p = .003), supporting H₁ and indicating that the closer the last NPI before the IPO date, the lower the market capitalization for the IPO firm. Results also show that NPI dispersion is positively related to IPO value (b = 1.425, p = .007), revealing that the more the NPIs are spread out preceding the IPO date, the higher IPO firm's market capitalization, supporting H₂. Finally, a left-skewed asymmetric pattern of NPIs before an IPO, where new products are introduced in a clustered pattern close to the IPO date, is negatively related to IPO value (b = −1.089, p = .044). This supports H₃, indicating that an IPO firm's market capitalization is lower when the firm's NPIs occur disproportionately closer to the IPO (than when they are introduced disproportionately further from the IPO).

Table 4.

Temporal Patterns of New Product Introductions on IPO Value.

Variable	Control Variables Only		Main Effects Added		Moderation Effects Added
	Coeff.	p-Value	Coeff.	p-Value	Coeff.	p-Value
Recency			−1.477 (.493)	.003	−1.393 (.486)	.004
Dispersion			1.425 (.521)	.007	1.280 (.501)	.011
Asymmetry			−1.089 (.539)	.044	−.945 (.548)	.086
Recency × industry growth					.151 (.072)	.037
Dispersion × industry growth					−.231 (.084)	.006
Asymmetry × industry growth					−.071 (.053)	.187
Recency × product innovativeness					.077 (.046)	.096
Dispersion × product innovativeness					−.09 (.048)	.062
Asymmetry × product innovativeness					.021 (.045)	.637
Industry growth			−.111 (.086)	.199	−.063 (.082)	.444
Product innovativeness			−.041 (.054)	.452	−.057 (.054)	.293
Venture backing	−.213 (.15)	.157	−.11 (.151)	.468	−.08 (.145)	.582
Underwriter reputation	.193 (.14)	.17	.316 (.145)	.03	.325 (.152)	.34
Firm age	.201 (.098)	.041	−.097 (.14)	.488	−.077 (.144)	.591
IPO market intensity	−.054 (.089)	.541	.008 (.101)	.941	.002 (.101)	.987
Patent stock	.211 (.081)	.009	.203 (.073)	.006	.206 (.073)	.005
Founder education	−.077 (.065)	.241	−.067 (.092)	.47	−.096 (.095)	.315
Risk-taking propensity	.092 (.067)	.176	.036 (.074)	.627	.04 (.073)	.587
Industry dynamism	.018 (.052)	.735	−.029 (.062)	.645	−.063 (.069)	.368
High-tech industry	−.167 (.164)	.309	.195 (.218)	.373	.061 (.217)	.778
Pharmaceutical industry	−.388 (.198)	.051	.868 (.429)	.044	.758 (.421)	.073
Service industry	.115 (.117)	.325	−.486 (.233)	.038	−.438 (.223)	.05
Inverse Mills ratio	.03 (.095)	.75	.207 (.104)	.047	.205 (.107)	.056
Control function residual for recency			1.576 (.5)	.002	1.498 (.492)	.003
Control function residual for dispersion			−1.435 (.545)	.009	−1.301 (.521)	.013
Control function residual for asymmetry			1.043 (.532)	.051	.884 (.539)	.102
Control function residual for product innovativeness			.073 (.049)	.132	.068 (.047)	.148
Constant	1.906 (.251)	< .001	2.437 (.298)	< .001	2.433 (.288)	< .001
IPO year fixed effects	Yes		Yes		Yes
R-squared	.358		.424		.444

Notes: n = 298. Robust standard errors in parentheses.

Regarding the moderating roles of industry growth and product innovativeness, we find the negative relationship between NPI recency and IPO value is weakened by industry growth (b = .151, p = .037) as well as product innovativeness (b = .077, p = .096), thereby supporting H₄ and H₅. For the positive relationship between dispersion and IPO value, industry growth weakens the relationship (b = −.231, p = .006) as does product innovativeness (b = −.09, p = .062), supporting H₆ and H₇. However, neither industry growth nor product innovativeness significantly moderates the relationship between NPI asymmetry and IPO value (industry growth: b = −.071, p = .187; product innovativeness: b = .021, p = .637), failing to support H₈ and H₉.

Using slope analysis, with low (10th percentile) and high (90th percentile) levels of recency and industry growth, the interaction of NPI recency and industry growth on IPO value reveal the simple slopes are significant for low (t = −3.058, p = .002) and high industry growth (t = −2.6, p = .01). The interaction between NPI dispersion and industry growth on IPO value reveals significant simple slopes of the interaction for both low (t = 2.858, p = .005) and high industry growth (t = 2.154, p = .032). For the interaction between NPI recency and the firm's product innovativeness on IPO value, with low and high levels of recency and product innovativeness, the simple slopes of the interaction are significant for low (t = −2.963, p = .003) and high product innovativeness (t = −2.725, p = .007). Finally, the interaction between NPI dispersion and product innovativeness on IPO value reveal significant simple slopes of the interaction for both low (t = 2.749, p = .006) and high product innovativeness (t = 2.315, p = .021). We illustrate these interactions in Figures D1–D4 in Web Appendix D.

Robustness Checks and Additional Analyses

Alternative measures of IPO value

Consistent with extant research, we measure IPO value as the market capitalization on the first day of IPO. To capture longer-term implications, we use market capitalization at a more distant time after the IPO as an alternative outcome. In line with Xiong and Bharadwaj (2011), we tested the robustness of our results using market capitalization on the 30th and 90th days after the IPO date. Results are generally consistent with our main findings (see Web Appendix E).

Incorporating nonlinear relationships of the explanatory variables with IPO value

To test potential nonlinear relationships between the temporal NPI patterns and IPO value, we added quadratic terms for recency, dispersion, and asymmetry to our model. Across all three temporal patterns, this extended model does not demonstrate any nonlinear relationships, and our focal results remain consistent in significance and direction (see Web Appendix F).

Simultaneous estimation

As an alternative to the two-step approach, we also aimed to simultaneously estimate the equations in our model (e.g., Mallapragada, Chandukala, and Liu 2016). We jointly modeled the endogeneity correction for four potentially endogenous variables (i.e., NPI recency, dispersion, asymmetry, and product innovativeness) and the main regression that involves the outcome (i.e., IPO value) using the conditional mixed process routine in Stata (Roodman 2011) to estimate the model. Results are overall consistent with those obtained using the control function approach (see Web Appendix G).

Alternative endogeneity corrections

In addition to the control function approach and the simultaneous estimation method we use to correct for endogeneity of recency, dispersion, asymmetry, and product innovativeness, we used an alternative method to correct for the endogeneity of product innovativeness using the copula-type nonparametric control function approach, to offer triangulation (Breitung, Mayer, and Weid 2024) inspired by Park and Gupta (2024). ¹¹ We use the standard control function residuals for the three NPI temporal pattern independent variables and employ the method described in Breitung, Mayer, and Weid (2024) to account for the endogeneity in product innovativeness. The results are consistent with our results from Table 4, in terms of both directions and significance. We report these results in Web Appendix H (Table H1). Additionally, as an alternative method, we employ the Lewbel (2012) method to account for the endogeneity in all four variables, which has been extensively used in marketing research (e.g., Ghazimatin, Mooi, and Heide 2023; Srinivasan, Wuyts, and Mallapragada 2018). Lewbel's (2012) approach leverages higher moments, particularly heteroskedasticity, to construct synthetic instruments. We provide the results, which are generally consistent with results from our primary estimation, in Web Appendix H (Table H2).

Discussion

Given that a successful IPO can substantially affect a firm's growth trajectory, it is no surprise that the phenomenon of going public has captured the attention of marketing scholars. Building on research that shows the value of innovation prior to an IPO, we aimed to determine how the NPI temporal patterns prior to an IPO are associated with a firm's IPO value and whether industry growth and product innovativeness shift these relationships.

Even after accounting for the IPO firm's patent stock and other firm-, industry-, and IPO-related variables, we find that NPI recency, dispersion, and asymmetry before a firm's IPO play pivotal roles in investors’ assessments of IPO value. Specifically, we find that a firm's NPI recency and (left-skewed) asymmetry prior to its IPO reduce IPO value, whereas dispersion tends to enhance it. Our findings also demonstrate that when an IPO firm operates in an industry with higher growth or has greater product innovativeness, the positive relationship between dispersion and IPO value is mitigated. Additionally, we find tentative evidence that the negative relationship between recency and IPO value is weakened under these conditions.

Theoretical Implications

Our research offers several implications for theory. First, this study extends the marketing literature on IPOs by highlighting the three important NPI temporal patterns: recency, dispersion, and asymmetry. These timing-related variables capture the nuance of when new products are introduced and help characterize a firm's NPI strategy when the firm plans for an IPO. We build on existing research on innovation (Cao et al. 2023; Heeley, Matusik, and Jain 2007; Wies, Moorman, and Chandy 2023) to highlight the importance of considering NPI timing before an IPO. Importantly, we show that a firm's pre-IPO NPI timing (i.e., recency, dispersion, and asymmetry) influences its IPO value—a key contribution to extant research.

Second, our research contributes to the marketing literature on innovation by highlighting important temporal patterns of innovation. Research has highlighted the pace and irregularity of innovation (Sharma, Saboo, and Kumar 2018), particularly in the absence of a specific deadline. In our research, the date of IPO is an important deadline, and we show the importance of recency, dispersion, and asymmetry of NPIs prior to this IPO deadline as critical to the IPO firm's value. These three NPI temporal patterns correspond to an important theoretical component of the dispositional perspective of time (e.g., Chen and Nadkarni 2017; Mohammed and Angell 2004), providing a comprehensive understanding of innovation timing. Unlike previous research that acknowledges deadlines and has shown positive effects of delaying NPIs closer to the end of the calendar year (Moorman et al. 2012) or to the competitors’ product launches (Harutyunyan and Narasimhan 2024), we demonstrate that delaying an NPI close to the IPO hurts IPO value.

Third, our study reinforces the temporal aspects of signaling theory. While the extant literature on signaling theory emphasizes a single temporal aspect (e.g., a point in time), our research highlights the importance of accounting for multiple temporal aspects that span across immediate (e.g., recency) and cumulative (e.g., dispersion and asymmetry) time to provide a more comprehensive understanding of signal timing.

Managerial Implications

Our results also offer important direction for managers. Foremost, we show that managers can signal the firm's prospects and communicate confidence about the IPO firm's future with a concerted interest in when the firm launches its NPIs. We show that when a privately held firm considers an IPO, it will also need to consider the recency, dispersion, and asymmetry of its NPIs prior to the IPO.

In terms of economic significance, paying attention to NPI timing is paramount. Our findings indicate that a 1% increase in NPI recency (i.e., NPI is closer to the IPO date by approximately 3.5 days) results in a 1.34% decrease in IPO value. Thus, when an average IPO firm introduces its last product only 3.5 days closer to its IPO, it reduces the firm's market capitalization by $30,000,000. Further, a 1% increase in NPI dispersion leads to a 3.65% increase in IPO value for an average firm, which translates to $80,110,300 increase in IPO value. Finally, a 1% increase in asymmetry of NPIs (i.e., NPIs are more disproportionately clustered closer to the IPO) is associated with a 1.03% decrease in IPO value for an average firm, which translates to $22,886,600 decrease in IPO value.

These findings suggest that, in general, when preparing for an IPO, firms benefit from emphasizing a consistent and dispersed nature of their innovations over time, rather than focusing solely on product introductions that occur right before the IPO. Managers can showcase a timeline of NPIs to demonstrate a sustained commitment to innovation. Our results point to the value of allocating resources to maintain a steady stream of innovation activities well before an IPO rather than concentrating resources on short-term product development efforts immediately preceding the IPO. In line with this, firms may consider incorporating metrics on the timing and distribution of NPIs into their performance evaluation systems for R&D and product development teams, especially as they consider an IPO.

We also offer actionable suggestions for firms in particular industries based on post hoc descriptive insights. Managers will first need to be aware of any industry norms regarding the timing of NPIs. Specifically, the descriptives in Table 5 demonstrate that the temporal patterns of NPIs are heterogeneous across industries.

Table 5.

Descriptives of New Product Introduction Timing Across Industry.

Group	Industry	Recency	Dispersion	Asymmetry
Group 1 (n = 298)	Pharmaceutical industry (n = 19)	−240.842 (276.784)	311.721 (148.728)	.163 (.784)
	Nonpharmaceutical industry (n = 279)	−359.48 (366.159)	376.887 (171.638)	.019 (.722)
Group 2 (n = 298)	High-technology industry (n = 217)	−365.871 (365.793)	364.283 (168.917)	.045 (.723)
	Non-high-technology industry (n = 81)	−314.531 (350.788)	395.367 (174.764)	−.015 (.738)
Group 3 (n = 298)	Service industry (n = 116)	−424.966 (432.635)	378.454 (178.243)	.015 (.679)
	Nonservice industry (n = 182)	−305.357 (300.672)	369.085 (166.266)	.037 (.756)

Notes: Mean is presented in each cell, with standard deviation in the parentheses. The three groups are mutually nonexclusive.

Finally, reinforcing the importance of the industry's growth and the IPO firm's product innovativeness, we elaborate on critical implications for managers when determining the importance and roles of their NPI recency and dispersion prior to IPO in Table 6. Given that our results do not reveal significant moderating roles of industry growth or product innovativeness on the main effect of NPI asymmetry, we discuss this in terms of potential avenues for future research.

Table 6.

Managerial Implications.

	High Industry Growth	Low Industry Growth	High Product Innovativeness	Low Product Innovativeness
Recency	IPO firms can afford to introduce products closer to IPO. For example, firms in industries that are susceptible to the growing influence of artificial intelligence could focus on launching product updates or versions closer to potential IPO dates because investors will appreciate the rapid development cycles.	IPO firms that strictly avoid launching their final pre-IPO new product near the IPO date may increase their IPO value.	IPO firms can benefit from launching a new product near the IPO date because the firm's innovativeness signals substantive innovation rather than superficial releases. Managers of such firms can thus confidently communicate the novelty of their NPIs in their press releases, emphasizing how these products reinforce the firm's long-term innovation trajectory. However, as the interaction of recency and product innovativeness demonstrates mixed findings across model specifications, we emphasize its managerial implication as tentative.	IPO firms may avoid introducing products too close to the IPO date because investors may discount such launches as opportunistic from less innovative firms. However, we note that the interaction of recency and product innovativeness demonstrates mixed findings, and we emphasize that this managerial implication is tentative.
Dispersion	IPO firms may not need to reinforce a dispersed pattern of NPIs as much as those in industry with low growth because this signal becomes more clouded by environmental noise.	IPO firms may reinforce a dispersed pattern of NPIs. These firms would benefit from planning a product road map well before IPO considerations and maintaining steady R&D investments and a robust innovation pipeline from which they can launch new products in a dispersed pattern. They can ensure a sufficient history of dispersed new products before IPO to highlight this pattern.	IPO firms may avoid overrelying on the temporal spread of their product pipeline as a standalone signal of strength and instead focus investor attention on the breakthrough quality and market potential of selected innovations, ensuring that dispersion is not misinterpreted as fragmented or unfocused innovation efforts.	IPO firms may strongly reinforce a dispersed pattern of NPIs, showcasing a stable and sustained record of product deliveries to the market over time. By strategically highlighting a dispersed pattern of NPIs, even if less innovative, managers can reassure investors about the firm's strategic pacing, operational reliability and consistent presence in the market.

Limitations and Further Research

Our research has some limitations that could pave the path for further research. First, we focus on the temporal patterns of NPIs, gauged by recency, dispersion, and asymmetry, whereas future research could investigate other temporal aspects associated with NPIs. For example, the agglomeration of NPIs, which can be measured by the kurtosis of the differences in time elapsed between each NPI and the incidence of IPO, may offer additional insights. Second, we find evidence that three NPI temporal patterns impact IPO value, whereas future research could investigate whether and how the IPO firm's reputation may shift the relationships we study. Although we control for firm age, extant research demonstrates that consumers have varying responses toward underdogs’ marketing strategies (Kirmani et al. 2017). This reputational effect may sway investor perceptions within the IPO context. Third, we focus on IPO value, whereas future research could consider how other IPO-related outcomes may shift in response to NPI temporal patterns. Understanding the drivers of NPI timing and IPO timing would also offer a rich avenue for further research. Fourth, given the potential for different innovation norms across industries, future research could investigate a specific industry, offering a deep dive into understanding the effects of a firm's NPI temporal aspects over its lifetime. Exploring the potential complementarity between a firm's product innovativeness and the industry's characteristics may also offer further nuanced insight into the role of NPI patterns for a firm's IPO. Finally, while the role of dispersion and the interaction of recency and product innovativeness demonstrate mixed support based on an alternative endogeneity correction method, we highlight that future research could further explore these roles.

Conclusion

An IPO is a significant milestone for a private firm. Our research highlights the importance of a firm's NPI timing prior to its IPO in driving IPO value—a significant outcome relevant to this milestone. We demonstrate how and when the recency, dispersion, and asymmetry of NPIs impact the IPO's value, shedding new light on an overlooked aspect of NPIs and their influence on firm performance. We hope our research paves the way for additional research in this domain.

Supplemental Material

sj-pdf-1-jmx-10.1177_00222429251382272 - Supplemental material for Temporal Patterns of New Product Introductions and IPO Value: The Importance of Recency, Dispersion, and Asymmetry

Supplemental material, sj-pdf-1-jmx-10.1177_00222429251382272 for Temporal Patterns of New Product Introductions and IPO Value: The Importance of Recency, Dispersion, and Asymmetry by Suyun Mah, Rebecca J. Slotegraaf and Girish Mallapragada in Journal of Marketing