How do entrepreneurs hone their pitches? Analyzing how pitch presentations develop in a technology commercialization competition. In Proceedings of the 33rd annual international conference on the design of communication (pp. 1-11).

 Spinuzzi, C., Pogue, G., Nelson, R. S., Thomson, K. S., Lorenzini, F., French, R. A., ... & Momberger, J. (2015, July). How do entrepreneurs hone their pitches? Analyzing how pitch presentations develop in a technology commercialization competition. In Proceedings of the 33rd annual international conference on the design of communication (pp. 1-11).

In this original research article, Spinuzzi et al study an accelerator in Korea. Startups at the accelerator aim to take their products to the US market. The startups are tech people, so they don't have any experience selling anything in general and in the US in particular. The study doesn't focus on cultural differences. 

The study focuses on revision. Unlike the Korea article from JBTC (I think?), this one doesn't get into Toulmin. Nor composition in the lit review. Not surprising since this is SIGDOC. All in all, the article shows how the startups revise their pitch decks (or not, in some cases) with feedback from judges. Unlike Tomlinson, Spinuzzi et al focus on the startups' value proposition and pitch deck--not the judges themselves. So not how the judges' feedback changes, but rather how the startups' design and argument changes. 

Spinuzzi et al asked three research questions:

RQ1: What kinds of feedback did presenters receive in the Quicklook® reports and training?

RQ2: What changes did they make to individual pitch arguments between training and final pitches?

RQ3: Do these changes correspond with favorable judges' scores?

Spinuzzi et al didn't like the way the accelerator gave feedback, so Spinuzzi et al asked them to change it. I think the accelerator got the market report (Quicklook®) back, which was completed by a third-party?, but then didn't share the results with the startup until the very end. "When discussing the pitch’s argument, the trainer referenced and showed portions of the Quicklook , which the presenters had not yet received. (The GIP recognizes this arrangement as a programmatic weakness and plans to rectify it in future years.)"; "More importantly, they had not yet received their Quicklooks, so they did not yet know which claims the market representatives had found unconvincing."

Technically, the word accelerator never appears. "the Gyeonggi-do Innovation Program (GIP), an entrepreneurship development program" Then later on: "This training program is structured as a pitch competition, training innovators how to make the complex arguments required by technology commercialization pitches in global markets."

Spinuzzi et al study four startup teams. "The 2013 competition involved over 200 applicants, of which 25 qualified to participate in the semifinals pitch. Here, we focus on four of the semifinalists (see Table 1)"

Spinuzzi et al collected many different types of data. "we collected primarily qualitative data for the sixth year of the GIP, including interviews, surveys, observations, and artifacts, and we analyzed a subset of these using an inductive coding scheme."

Open coding. "First, we closely compared presenters’ training and final pitches to detect changes in spoken words, slides, and visible actions. We coded these changes (Saldana [23]), using open coding to develop emergent categories. This process resulted in 42 codes. For this analysis, we focused on 21 codes clustered in three code types (Table 3)." Spinuzzi said this in an email: "Toulmin codes were starter codes (deductive). We figured that we could start by identifying Toulmin moves (a pretty safe bet) and explore the rest of the data inductively from there. But it was all still nonexclusive coding (ex: a feature could be coded for both claim and warrant), unlike the differential coding scheme you used for the pitch decks."

The startups went through the following training:

• Presentation skills training
• Practice pitch
• Trainer feedback
• Quicklook
• Final presentation

The results section is broken into three sections, each of which corresponds exactly with the above RQs. 

RQ1 (What kinds of feedback did presenters receive in the Quicklook® reports and training?) is itself broken into three parts.

• Structure. Interestingly, as Spinuzzi said, the pitch deck template they were working with seems very old, as evidenced by "technology" as a heading. Startups deviated or misunderstood the slides occasionally. 
• Claims and evidence. À la Toulmin, the trainers focused on trying to get the startups to focus more in qualifiers and rebuttals, on mitigating risks and blind spots in their arguments by saying things like, "We'll need to educated the public."
• Engagement. Demonstrations. Stories. Questions. 

RQ2 (What changes did they make to individual pitch arguments between training and final pitches?) is also broken down in the same way. 

• Structure. 
• Claims and evidence. 
• Engagement. 

The startups responded to feedback by adding more slides despite the fact that they had worries about time constraints, which is interesting. Seeing revision as "more" rather than "different."

From what it seems like, some added some, others deleted some; some responded to feedback more than others. 

RQ3 (Do these changes correspond with favorable judges' scores?) did NOT share the tripartite structure of the other two RQs. It is in this section that Spinuzzi et al talk about how the feedback is problematic insofar as it can only prompt the startup to hone the argument, not the product or design or technology itself--thus echoing THE concern of fields like disability and usability studies. 

The Quicklook report offered go versus no go suggestions. 

Having trouble locating the answer to RQ3. "The question is not as straightforward as it might appear." Yes, but what is the answer? I guess maybe the answer is hard to pin down because not all of the startups received that good of a score and in fact two were cautioned not to go on with the product (no go). In these cases, I think, they said you need a major overhaul of the product or value proposition, a pivot, but they couldn't pivot that quickly in the accelerator obviously. 

"each judge provided an overall score from 1 to 4 (with 1 being the best)" So in the case of the first one (K6015), the judges didn't even agree. The only one with high scores in which the judges agreed AND in which the startup got the go signal was "K6010": this was the one that changed their story via the Jaws example, that got all rhetorical by deleting the sound and making the audience wear the bracelets I think??? This was also the one in which there was a co-creation between audience and startup (see London et al. [16]; Lusch & Vargo [17]) AND which was horrified to get the advice to enter the kinky market. 

So to answer the question, it's yes and no. Yes because it did happen, but only 25% of the time in an obvious way.  

Gap. "Although many studies have examined how pitches are presented, few examine how such arguments are developed."

It's almost as if Spinuzzi et al is doing an consultant (ish) job on the pitch competition, studying it as though it were a telecommunications firm. "Based on our analysis, we believe that pitch feedback in this program has tended to conflate different categories of feedback. In particular, the trainer and judges address (but do not separate) aspects of Argument (structure, claims and evidence, and engagement), and judges additionally address (but do not separate) aspects of Design and Use. We believe that pitch training programs could provide more categorized feedback by developing heuristics for separating these aspects; providing metrics for rating them; and providing concrete written feedback for each. We are currently working on such heuristics for the program under discussion."

Approx 2hrs, started 7am finished 8:51

https://utexas.box.com/s/whsy1hx4dax3u1p0h9qqk7imixw26ln1