Methodology of my Film Industry Survey 2014

3 March '14 Comment

This post explains my ‘Film Industry Survey 2014’, sets out my aims of the project and the exact methodology. You can read the full survey as one 25-page PDF by signing up to my free mailing list or you can read each section as individual blog articles at the following links…


The film industry is heavily influenced by shifting opinions, so I thought it would be fascinating to take its temperature on a number of hot topics including piracy, the appeal of 3D, gender, and how optimistic industry professionals are for 2014. As with all my research, my aim is not to prove a particular point but to see what interesting results the data provides. Some of these results confirm the conventional wisdom while others challenge it. I am looking forward to seeing how film professionals, pressure groups and journalists respond to these results. If you would like to know more, offer help/advice on future research or to just drop me a line, I can be reached at It’s always nice to hear from people who enjoy or use my research. I am also open to new collaborations and commercial projects.


Using a few industry resources I built up a list of 22,940 film industry professionals who have attended at least one of the three major film markets (Cannes, Berlin or AFM) within the past five years. Each was emailed and asked to fill in an online survey. Of those emailed, 44% opened the email, 15.7% clicked on the link and 34% of those (1,235) completed the survey. Interestingly, over 100 people then emailed back to let me know that they had taken part. I feel that these rates are extremely high, showing how open and communicative the industry can be.

The Respondents

The surveys were anonymous, although the link was only emailed to suitable people and not shared publicly. Of those who completed the survey…

  • 39% were in Production
  • 21% in Development
  • 13% in Sales & Distribution
  • 9% in Post-Production
  • 9% in Other
  • 5% in Marketing
  • 4% in Exhibition

Respondents were permitted to select multiple sectors. 56% selected just one sector, 23% selected two, 14% selected three and 5% selected four or more. 86 countries were represented, which included the United States (20.5%), the UK (15.7%), Canada (6.9%), France (5.5%), Germany (5.1%), Spain (4.0%), Australia (3.6%) and 79 other countries (38.8%). I asked for the number of films they had in development, in production, finished but not released and distributed in different budget ranges. I then used this to assess which of three budget categories best describes their work to date.

Sneaky questions

In one section of the survey I wanted to illicit truthful answers on tricky topics and so I resorted to a sneaky technique. To one randomly assigned set of participants I presented three statements about the industry (such as ‘I prefer to watch films on DVD than in the cinema’). I then asked the respondent how many of the three statements they agreed with, but only asking for the combined total (i.e. “I agree with two of the three statements”). Then, to a different randomly assigned set I offered the same three statements with the addition of one of the following statements…

  • It’s harder for women to succeed in the film industry than men
  • A movie is more enjoyable in 3D than in 2D
  • Investing in films in 2014 is a sensible business investment
  • I have illegally downloaded a TV show or feature film

By subtracting the average number of agreed-with statements (for example 3.45) from the average of the control group (for example 2.87) I was able to calculate the percentage of people who agreed with the additional statement (in this case it would be 58%).


As with any survey, there are a few limitations to this project. The most significant which occur to me are…

  • Lying – I have no way of verifying that the answers given are correct. I did all I could to prevent this by making the survey anonymous, by asking key information in subtle ways and by providing no reward for particular answers. I hope that this reduced any pressure respondents may have felt to lie, exaggerate or downplay.
  • Language – My emails and survey were both in English, which will naturally exclude non-English speaking professionals.
  • Weighting – In order to calculate how the opinions altered at differing budget ranges I asked respondents for the number of films they had made in a variety of budget ranges. I then used this to calculate which their principle budget range was. I would have preferred to have had a more precise method of determining in which budget range to place people, as this cannot take account of the significance of each movie. A ‘low budget producer’ may now be making larger films but if their move is recent then they would still be regarded as ‘low budget’ for this survey.
  • Selection bias – By targeting people who have attended a film market I am already providing a certain skew on the answers.
  • Self-selection bias – Respondents were exclusively made up of people who were willing to open my email, click on my link and complete the survey.


This survey would not have been possible without the continued support of Mike Mindel, Sophie Lifschutz, Lucy Fazey and Edward L Dark – thank you all!


Leave a Reply

Your email address will not be published. Required fields are marked *

Stephen Follows