Slide Structure of Pitch Decks From an Accelerator Program

Slide Presentations

From a multimodal perspective, slide presentations have been studied from a variety of approaches, such as the result of a social semiotic practice (Zhao et al., 2014), as part of an oral presentation (Bucher & Niemann, 2012; Lindenberg, 2023a, 2023b, 2023c; Rowley-Jolivet, 2004; Ruiz-Madrid, 2021) relating certain aspects of the slides to properties of the speaker or the audience, and as multimodal artifacts in isolation to the oral presentation that usually accompany them. Within these latter studies of slide presentations as multimodal artifacts, different aspects of the presentations have also been explored, with some research focusing on the use of specific visual elements within the slides, for example, the use of fonts (Kmalvand, 2015), images (Feng, 2021, Zhao & van Leeuwen, 2014), and distribution of text (Garrett, 2015). Meanwhile, others have analyzed the slide as a complete unit (Bucher & Niemann, 2012; Rowley-Jolivet, 2002, 2004) or explored the use of slides in relation to the presentation as a whole, as well as the macrostructure of the presentation (Atai & Talebzadeh, 2012; Djonov & Van Leeuwen, 2013; Feng, 2021; Rowley-Jolivet, 2002, Zapletalová, 2014).

One of the most influential studies on the multimodal characteristics of slide presentations has been the one realized by Rowley-Jolivet (2002), who studied the visual characteristics of scientific conferences in the fields of medicine, geology, and physics. Her work not only pioneered the research of slide presentations but also serves as the basis for other researchers who have followed in her steps by analyzing the same genre, scientific conferences, and by using similar, and in some cases the same, categories (Atai & Talebzadeh, 2012; Bucher & Niemann, 2012; Kmalvand, 2015; Zapletalová, 2014). Later on, in her work of 2004, Rowley-Jolivet analyzed in more detail the same corpus of conference presentations according to their visual content, classifying that content as scriptural, numerical, and figurative (ordinary or technical photography) and graphical (conceptual constructed elements, in opposition to naturalistic ones). These categories correspond, at the same time, to three distinctive semiotic modes, as defined by Rowley-Jolivet (2004): linguistic for scriptural elements, mathematical for numerical, and visual for both figurative and graphical elements; these semiotic modes are also used by other authors as a way of simplifying the range of categories involved. Bucher and Niemann (2012), for example, classified the slides into text slides, pictorial slides, or mixed slides, following more closely the classification of semiotic modes proposed by Rowley-Jolivet (2004). The authors (Bucher & Niemann, 2012) analyzed the use of different elements on slides from scientific conferences in the field of economics, natural sciences, and humanities, although their work used this classification only as a starting point to later describe cognitive difference in the reception of the different slides and, therefore, did not expand on the analysis of the slides themselves.

Other studies, such as those by Kmalvand (2015), Zhao and van Leeuwen (2014), and Garrett (2015), instead of analyzing each slide as a whole, put their focus on the elements present inside the slides. Thus, even though Kmalvand (2015) used the same categories developed by Rowley-Jolivet (2002, 2004), the author identified individual visual elements present in the slides used in thesis defenses from the field of applied linguistics, together with other elements, such as tables, fonts, transitions, animations, and the use of color, both in the background and in the text. In most cases, these elements were measured according to their use in the slide presentations studied as a whole and did not consider the number of elements per slide or the structure of the slide presentation. Zhao and van Leeuwen (2014) also analyzed specific visual elements present in the slides of university lectures. However, their research limited itself to the analysis of images classified as photographs, maps, clippings or screen captures, paintings, graphs, cartoons, and others, and, once more, even though the authors researched individual elements present inside every slide, the results were shown only in relation to the presentation as a whole. Garrett (2015), instead, used fewer categories, distinguishing only between the use of images, words, and paragraphs, which he considered as any grouping of words. He also considered the number of these elements used per slide, although he only measured them as a way to explore the correlation within these variables and the number of downloads, or the amount of use, these presentations received by the students, and did not describe the composition of the presentations he analyzed.

Studies that have put a stronger emphasis on the composition of a slide presentation as a whole, and not of the individual slides, have also drawn, in many cases, inspiration from the work of Rowley-Jolivet (2002, 2004). In particular, in her work of 2002, the author (Rowley-Jolivet) analyzed the use of slides in 90 conference presentations, measuring the number of slides per presentation and the average periodicity of the projection of every slide, as well as describing the use of color, sequences, and type of slides—this last element being explored in more detail in her later work (Rowley-Jolivet, 2004). The presence of sequences, which are defined by Rowley-Jolivet as “semi-autonomous segments from the prior and subsequent discourse” (2002, p. 37), marks a step toward the description of the structure of a slide presentation, instead of describing the slides or their internal components as isolated elements. These sequences were then analyzed in terms of their recurrence as well as by the use of specific visual elements or by the change of certain linguistic or visual characteristics, such as the use of discourse markers. One sequence described, for example, was Patient and tumor characteristics, which consisted of two tables, one displaying the information of the patients followed by a second one with the information of the tumors. However, the author never explicitly states if these tables are presented in individual slides or side by side, something that limits our understanding of these sequences. A second limitation, as even Rowley-Jolivet herself (2002) notices, is not having considered the field of the presentations, something that Rowley-Jolivet hypothesizes might be linked to the presence of these sequences. The exploration of sequences has been left unexplored by Rowley-Jolivet in her later work, as well as by the studies by other authors who expanded on her work. Instead, other authors (Atai & Talebzadeh, 2012; Djonov & van Leeuwen, 2013; Feng, 2021; Zapletalová, 2014) placed their attention on either the recurrent function of certain slides or the macrostructure of the presentations as a whole, not considering if that structure is materialized through the use of certain sequences of slides or specific elements.

Other studies, meanwhile, have reported on specific structural characteristics of slide presentations, such as the common use of a title slide to start a presentation as well as the use of specific types of final slides in presentations from academic and professional settings, which is mentioned by Atai and Talebzadeh (2012) as well as Djonov and Van Leeuwen (2013), Zapletalová (2014), and Zhao et al. (2014). First, a title slide is an introductory slide that presents the title of the presentation as well as information regarding the presenter such as name and affiliation (Atai & Talebzadeh, 2012). As Zhao et al. (2014) notice, a title slide is automatically generated as a starting point when creating a new slide presentation using PowerPoint 2003 or 2007, something that continues to be the case with more recent versions of the software and that could be a factor in its widespread use. Meanwhile, the use of the final slide of a presentation has been described as taking diverse functions by different authors and, contrary to the title slide, there has been no particular template linked to it. One such use is as a thank-you slide described as a slide that presents a message of gratitude toward the audience, as well as repeating the name and affiliation of the presenter in the case of conference presentations (Atai & Talebzadeh, 2012), although they have also been described as a part of both academic lectures and corporate presentations (Djonov and Van Leeuwen, 2013). Another type of final slide has been described as presenting a goodbye phrase (Zapletalová, 2014) that is described to be also present in conference presentations. Finally, Djonov and Van Leeuwen (2013) also described the use of a final slide to present credits, in the case of corporate presentations, or references to key readings, in the case of lectures, although no further characteristics were mentioned.

One aspect that these articles have in common, regardless of the function described for the last slide or the genre of the slide presentation, is that they all have described the initial and final slides in isolation to the overall structure of the presentation and, in some cases, by only mentioning them instead of describing the overall structure of the slide presentation (Zhao et al., 2014), while in other cases making an explicit distinction between them and the rest of the structure of the slide presentations (Atai & Talebzadeh, 2012; Djonov & Van Leeuwen, 2013; Zapletalová, 2014). Thus, Djonov and Van Leeuween (2013) make the distinction between the title and end slides and what they call content slides, which are all the slides that form the presentation. Meanwhile, Zapletalová (2014) distinguishes between scientific and nonscientific slides in her analysis of conference presentations, with the title and final slides being part of the latter and, therefore, being excluded when comparing the structure of the slide presentation to that of a research article. Although highlighting the use of these initial and final slides allows us to understand the similarities of the use of slide presentations in different genres, their description as isolated elements and exclusion from the macrostructure of the presentation is also problematic as it gives a distorted image of how these genres are built.

The macrostructure of slide presentations has been studied only in conference presentations as a reflection of the research article (Atai & Talebzadeh, 2012; Rowley-Jolivet, 2002; Zapletalová, 2014). As already mentioned, Rowley-Jolivet (2002) studied the used of specific sequences and not the macrostructure of conference presentations, but the author was the first to present the possibility of a conference presentation having a similar structure of that of a research article, something that served as a jumping point for other authors. Atai and Talebzadeh (2012) and Zapletalová (2014), for example, continued the work of Rowley-Jolivet (2002) by analyzing the same genre, conference presentations, as well as by using similar categories, and expanded it by including a macro- and micro-structural analysis of these presentations. First, Atai and Talebzadeh (2012) were able to describe the macrostructure of conference presentations as following the same structure as a research article (introduction, methods, results, and discussion), as described by Swales (1990), noting that introduction and results were the sections with the largest number of slides, representing together more than half of the total number of slides of the presentations. Later on, Zapletalová (2014) would also confirm the use of the same structure, although she did not add more details regarding the distribution of the slides, focusing instead on the use of specific types of slides across the presentation.

Feng (2021), instead of describing the macrostructure of slide presentations, identified the structural moves of what he called instructional PowerPoints used in linguistic courses from Peking University. The author studied 33 slide presentations created by 6 award-winning professors of a contest for best use of slide presentations hosted by the university. He identified 8 moves: Headline, Content, Lead-in, Concept-Definition, Explanation, Exemplification, Exercise-Answer, and Assignment. Similar to the use of a title slide described by other authors (Atai & Talebzadeh, 2012; Djonov & Van Leeuwen, 2013; Zapletalová, 2014; Zhao et al., 2014), Feng (2021) describes the use of the Headline move on the first slide of the presentations to present the topic of the lecture, although it can also be seen at the top section of a slide, in which case it is described as a second-level Headline. In this second case, Headline is also described as being commonly used together with the Content move in a single slide, but the distribution among slides is not further explained for the other moves. The author also analyzed the use of images and their relation to the text, but did not relate them to any specific move. Moves were also not organized into any particular macrostructure; instead they were only described in accordance to their presence or absence within the entire presentation.

The Pitch Deck

The pitch deck, as described by Spinuzzi, Nelson, et al. (2015), is a slide presentation that usually accompanies a business pitch. Yet, even though the pitch has been widely researched (Sabaj et al., 2020), the pitch deck has been the focus of only a few studies, which have explored its evolution (Spinuzzi et al., 2014; Spinuzzi, Nelson, et al., 2015; Spinuzzi, Pogue, et al., 2015) or its reception (Williams et al., 2019), describing the pitch deck only in relation to them. Thus, Spinuzzi et al. (2014, Spinuzzi, Nelson, et al., 2015, Spinuzzi, Pogue, et al., 2015) investigated the role of the different genres used throughout an accelerator program in the making and remaking of the pitch deck. Through an ethnographic approach, they were able to document the different genres involved in the creation of a pitch and pitch deck (Spinuzzi et al., 2014), the reuse of text from other documents in the editing of the pitch deck (Spinuzzi, Nelson, et al., 2015), and the motivation behind changes in the pitch decks throughout the training process (Spinuzzi, Pogue, et al., 2015). Meanwhile, Williams et al. (2019) analyzed the reception of the visual characteristics of a small group of pitch decks. Attributes, such as color, typography, iconography, and organization, were not described, but scored by a group of experts through a rubric to determine the level of presence and success they had in every slide presentation.

Spinuzzi, Pogue, et al. (2015), in particular, mentions the structure of the pitch decks when describing its changes in four pitch decks that were created and edited through a South Korean training program. In this study, two versions of a pitch deck were compared and described in accordance to the explicit feedback given by the trainers, which related to the structure, claims and evidence, and the engagement of the decks. Slides were identified by the title used on each one of them, but the structure of the pitch deck was only described in terms of its changes, which were considered as additions, cuts, or moves of the slides. Results showed participants addressed feedback only partially. In the specific case of structural changes, three of the four presentations examined added new slides, making addition the most common change. In contrast, only a few slides were moved, making it the less common of the changes. When describing the overall structure of the pitch decks, the authors only mentioned that participants followed very closely the template given by the training program they were part of, but no further details were given other than the names of the slides that had some change, and the structure of the template was never presented.

The structure of pitch decks from the biotechnology field has been described in detail by Shimasaki (2020) in his manual Biotechnology Entrepreneurship: Leading, Managing, and Commercializing Innovative Technologies, although no research is linked to this structure. According to his book, a pitch deck should be composed of 18 topics, with each topic being at least one slide long. This structure is preceded by a title slide, just as mentioned already in other studies (Atai & Talebzadeh, 2012; Djonov & van Leeuwen, 2013; Zapletalová, 2014; Zhao et al., 2014). The oral presentation is then meant to begin with a slide stating the company purpose, history, and mission, and finish with a thank you and Q&A slide, also reminiscent of what has been described by other authors (Atai & Talebzadeh, 2012). After the oral pitch is over, the pitch deck could continue with two final topics meant to foresee and respond to likely questions. One of these topics could be a slide presenting potential risks; meanwhile, the other could be one or more slides supporting the claims throughout the presentation. This structure is defined by Shimasaki (2020, p. 326) as a “Master Slide Deck,” which could then be adapted to better fit the audience and time constraints of the different situations where it might be used. A shorter pitch deck, for example, is described as being composed of 10-12 slides and presented in up to 10 minutes, although the author does not explain which topic, of the 18 presented, should remain and which ones should be skipped.

The macrostructure of the pitch, in contrast, has been described by Daly and Davy (2016) through the analysis of 13 winning pitches from the reality TV program Dragons’ Den. The authors described the macroorganizational structure of these pitches as consisting of 10 stages: Greeting the audience; Identifying oneself and one’s company, product, or service; Announcing the amount of investment required; Explaining/presenting the product or service; Presenting the history/achievements of the company and/or founder; Describing future plans; Presenting the target customer; Recapitulating and developing; Thanking the audience; and Expressing a willingness to answer questions. The authors found that most of the pitches analyzed used similar sections, although they appeared in different positions. Of these 10 sections identified, 4 were described as obligatory and the other 6 as optional. The obligatory sections were: identifying oneself and one’s company, announcing the amount of investment required, explaining/presenting the product or service, and describing future plans. These results, although an important approximation to the study of the pitch and the structure of the pitch deck, have some limitations related to the origin of the sample, which represents a very specific use of the pitch as a televised spectacle and, therefore, reflects the demands and restrictions of a TV show. Furthermore, the pitches presented in Dragons’ Den do not use a slide presentation, which differs greatly from the pitches usually seen in accelerator programs or pitching competitions. Thus, although this work is an important part of understanding the structure of the pitch, it might not be representative of the structure of pitches used in other contexts.

Data collection

Data were collected directly from Start-up Chile, a state-funded accelerator program. Start-up Chile was launched in 2010 (Gonzalez-Uribe & Leatherbee, 2018) and, at the point of contact, consisted of three programs: S Factory, Seed, and Huella. Data came from the 17th and 18th generation of the Seed program, Start-up Chile’s main program at the moment. These generations were composed of 153 companies from a variety of sectors and countries that participated between 2017 and 2018. Because of the international nature of the program, all documents produced throughout the program were in English.

The corpus of this study consists of 96 pitch decks, with a total of 1513 slides. The pitch decks collected correspond to the final version of pitches produced by participants. Each final version is presented in 3 minutes at Pitch Day, an event realized after participants receive different forms of training and feedback, and where a group of finalists are selected to be part of Demo Day, an event that grants companies media coverage, as well as the opportunity to receive further funding and mentoring.

As seen in Table 1, presentations varied, with the shortest pitch deck being composed of only 7 slides, whereas the longest deck used 42 slides. Even in the extreme cases, the presentation could not last longer than 3 minutes, which meant an average time of 4 seconds per slide in the case of the longest presentation.

OPEN IN VIEWER

Table 1.

Corpus.

Number of presentations	96
Total number of slides	1513
Average number of slides per presentation	15.75
Minimum number of slides	7
Maximum number of slides	42

Data Analysis

The analysis model was created through a mixed approach that considered top-down and bottom-up stages as explained by Biber et al. (2007). This meant the use of preexisting categories to create an initial model that was later tested and adjusted by modifying, adding, or deleting categories to better reflect the characteristics of the pitch decks being analyzed. A first step consisted, then, in using as a starting point the topics taught by Start-up Chile in what they call SUP Academy, which consisted of 11 main topics and 5 optional ones. Afterward, these categories were tested and adjusted by two independent coders who analyzed a third of the corpus. The final model consisted of 18 topics described in Table 2. These final categories were later applied by four coders to the entirety of the corpus.

OPEN IN VIEWER

Table 2.

Coding Model.

Topic	Description
Company	The slide presents company name, logo, or contact information
Team	The slide presents team members
Problem	The slide presents problem or pain to be solved
Product	The slide presents the products or services of the company
Achievements	The slide presents the current achievements of the company
Market	The slide presents a description of the current market or possible client groups for the products or services
Advantages	The slide presents the positive attributes that distinguish their products or services from others
Business model	The slide presents a general scheme of how the company operates or plans to operate, or explains the ways how they expect to generate income
Next steps	The slide presents the milestones expected to be achieved by the company in the nearby future
Gratitude	The slide expresses gratitude toward the audience
Competition	The slide presents specific competitors or the advantages or disadvantages related to them
Call to action	The slide gives a directive to the audience
Value proposition	The slide presents a short statement with the mission or main characteristic of the company
Clients	The slide presents specific clients
Solution	The slide presents a general solution to a problem, but does not relate it to a specific product or service of the company
Funding	The slide explicitly asks for money
Partners	The slide presents specific partners or investors
Other	The slide presents no clear topic

To ensure the reliability of the results, the four coders were first trained on the model and later tested by analyzing eight presentations. The analysis consisted of 66 observations that were used to triangulate the level of agreement between the coders through Fleiss’s kappa (Torres Gordillo & Perera Rodríguez, 2009). The coders achieved a kappa of 0.75, right at the limit between good and excellent. Because the level of agreement was high, the corpus was analyzed by the four coders independently, and to ensure further reliability, two meetings were conducted where the four coders would present difficult cases and come to agreement.