Crowdsourcing at Amazon: Democratization of TV / Film Content
“Hopefully we can avoid big bombs. Our notion for what the world needs may be a roller-skating movie or a battleship film, but that could be wrong. We can do tests and find out that, actually, no one cares about this project or that one. If you do that before you spend $200 million on it, that would be good. Good for customers and good for the business.”
– Roy Price, Former Head of Amazon Studios(1)
Introduction to Amazon: A Pioneer in Content Crowdsourcing:
Amazon launched Amazon Studios in 2010 to begin producing original content for their Amazon Prime streaming platform(9). Streaming platforms, unlike traditional linear media formats, allow for more programming “shelf space”, given the lack of primetime TV slots or prime movie opening days. While TV producers often seek out formal customer feedback before greenlighting a show, Amazon Studios spearheaded a fully crowdsourced model of content creation and feedback. This crowdsourcing approach aims to increase the success rate of new creative content(8), compared to the traditional Hollywood practice of sourcing and filtering early-stage ideas through a board of media executives.
Amazon Studio’s Crowdsourcing Across The Production Funnel:
Amazon Studios has attempted to introduce crowdsourcing into multiple aspects of their content production funnel:
Example 1 – Open-Submission of Script and Concept Videos:
Amazon Studios has an open-submission policy for scripts and concept videos, meaning it will accept concepts from anyone, regardless of their level of experience. This approach dramatically increases the size their production funnel, which may be a significant advantage in the streaming video economy (as described above). Through open-submission Amazon has produced a number of successful products, such as “Zombies versus Gladiators” and the children’s TV series “Magic Monkey Billionaire.” (1)
Example 2 – Customer Feedback on Concepts / Footage via Amazon Preview:
Amazon Studios has also implemented Amazon Preview, an invite-only community of top customers who are encouraged to provide feedback on various concepts, storyboards, and test footage of submitted content. These customers provide valuable feedback, ratings, commentary, and hashtags that serve to categorize and analyze content(4). This system allows Amazon to collect information on new content earlier than other studios and at a much lower cost. (3)
Example 3 – Public Voting on Pilots via Pilot Season:
Amazon implemented “Pilot Season” in 2012. This program encouraged customers to watch pilots and vote on which should have a full season commissioned. Former Director of Amazon Studios, Roy Price, described the value of this process by saying, “we’re trying to develop or find shows that will be popular with Amazon customers, so who better to ask than Amazon customers themselves.” Greenlight decisions are made not only based on voting data, but on actual customer usage behavior(1). Two of Amazon’s biggest hits, “Transparent” and “Mozart in the Jungle”, were selected through the Pilot Season program. (2)
Amazon Studio’s Impediments and Avenues for Improvement:
Amazon is the only player to utilize a crowdsourcing model. At the end of June 2018 Amazon closed their open-submission portal on June 30th, 2018(6). Creators will still have access to the companion Storybuilder and Storywriter tools (Amazon-provided digital tools aimed at helping under-resourced creatives) (6). In addition, Amazon announced in July 2018 that it would end its Pilot Season program as well. As of November 2018, Amazon still offers Amazon Preview, which it views as value additive to both users and to the company’s market research efforts. One reason for the closure of these programs, according to Amazon’s co-head of TV, Albert Chang, was that “you end up taking way too long to get the actual season done.” (3)
Impetus for Change:
The company hasn’t explained much more regarding the reasoning behind the shutdown besides “looking for ways to become even more efficient.” (6) In part, this transition was likely due to a change in Amazon’s Prime Video strategy. As the streaming video category has matured, Amazon is increasingly focused on blockbuster productions with very large title-level budgets (Amazon spent an estimated $4.5B on content in 2017 — twice HBO’s budget(5)). As the required investment of each individual title increases, studios are increasingly incentivized to mitigate the risk of new projects through the utilization of existing IP (e.g., licensing known entities such as comic books) or through partnerships with established talents (e.g., writers, directors, actors) (6).
Suggestions for Management:
It is difficult to maintain crowdsourcing as a company scales and matures, especially in markets where competition for new projects is competitive and requires a high level of expertise. Amazon’s crowdsourcing model failed because A) it created a delay between pilot and production that was unacceptable to the most in-demand content creators(3), and B) it sometimes created data biased toward a non-representative demographic of consumers (those who view and vote on new pilots). However, there are ways for Amazon to address these concerns and re-implement crowdsourcing. I suggest that the company explore the use of computer assisted tools to sort and filter user-submitted scripts, which could and minimize labor requirements and increase speed-to-market. Additionally, Amazon could better integrate voting mechanisms into their core Prime Video Platform, utilizing “painted door tests” or showing short previews of potential projects before shows, which would widen the sample of consumers to a more representative sample. This would enable Amazon to collect a greater and more representative scale of data much faster than their initial model allowed(7). Crowdsourcing was an innovative attempt at gaining a competitive edge in content creation, and Amazon should continue investing in related models.
- Can we utilize machine learning or machine-automated selection to improve crowdsourcing efficiency?
- What part of the funnel should Amazon reconsider utilizing user input – idea generation or selection?
(Word Count: 799)
 Reuters. (2018). Crowdsourcing goes to Hollywood as Amazon makes movies. [online] Available at: https://www.reuters.com/article/us-amazon-hollywood-crowd/crowdsourcing-goes-to-hollywood-as-amazon-makes-movies-idUSBRE8990JH20121010 [Accessed 13 Nov. 2018].
 Business-standard.com. (2018). Amazon stops putting ‘pilot episodes’ online for viewer feedback. [online] Available at: https://www.business-standard.com/article/news-ians/amazon-stops-putting-pilot-episodes-online-for-viewer-feedback-118073000713_1.html [Accessed 13 Nov. 2018].
 Fortune. (2018). Amazon’s ‘Pilot Season’ Practice Officially Ends. [online] Available at: http://fortune.com/2018/07/29/amazon-pilot-season-viewer-feedback/ [Accessed 13 Nov. 2018].
 Page, V. (2018). Amazon Studios vs. Hollywood: Is Netflix the Better Pick? (AMZN, NFLX). [online] Investopedia. Available at: https://www.investopedia.com/articles/personal-finance/111815/amazon-studios-vs-hollywood-netflix-better-pick.asp [Accessed 13 Nov. 2018].
 CB Insights Research. (2018). Amazon Strategy Teardown. [online] Available at: https://www.cbinsights.com/research/report/amazon-strategy-teardown/ [Accessed 13 Nov. 2018].
 Engadget. (2018). Amazon Studios stops taking amateur scripts on June 30th. [online] Available at: https://www.engadget.com/2018/04/14/amazon-studios-ends-open-call-for-scripts/ [Accessed 13 Nov. 2018].
 Tucker, R. (2018). How Crowdsourcing Can Enhance Innovation Performance. [online] Forbes. Available at: https://www.forbes.com/sites/robertbtucker/2017/06/16/how-crowdsourcing-can-enhance-innovation-performance/#75695dae54df [Accessed 13 Nov. 2018].
 K. Boudreau and K. Lakhani. Using the crowd as an innovation partner. Harvard Business Review 91, no. 4 (April 2013): 61–69.
 Wallenstein, A. and Wallenstein, A. (2018). What the TV Biz Could Learn from Amazon Studios. [online] Variety. Available at: http://variety.com/2013/biz/news/amazon-viewer-data-for-development-process-1200429921/ [Accessed 13 Nov. 2018].