Part 2: The Drilling Down This survey looks at 150 crowdfunding projects from websites such as Kickstarter and Indiegogo that focus on role playing game related material. The bulk of the projects were found through a forum thread on RPG.net, though a few came from other sources on the internet. They cover a time period from April 2011 to April 2012.
The first part of this report looked at many of the overall numbers of the survey sample. Part two will look at the behavior of the backers and the creators by examining in greater detail the number of backings at different funding levels, along with some of the key offerings in those funding levels.
A Look at the Median Part of the feedback from part one of this report was “what is the median of the data?” The median is similar to an average, however rather than adding all of the factors together and dividing, instead what is determined is the value of the middle factor in the set of numbers. This helps to mitigate the effect of extreme values at either end, such as a very ambitious but unsuccessful funding goal, or an extremely successful funding, both of which do not represent the more typical patterns.
Table 1: Overall Set of Data
Table 2: Successful and Unsuccessful Median
When measuring the median with the data the numbers go down compared to the averages, however the ratios are not drastically different. It is still clear that successful projects overall aimed for a more modest goal than unsuccessful projects and then proceeded to more than double their funding goal. Meanwhile the unsuccessful projects aimed higher but only raised 9% of their funding goal.
Backer Behavior Over a year long period there were 16,763 backing instances with the 150 projects covered in the survey. A total of $834,920 were raised from these instances and broken up into 96 different funding values. Let us look at the broad patterns that emerged.
Backing Instances
 As you can see from Figure 1 there are clear trends in the amounts that backers were willing to fund. While this is in part shaped by the options creators offered, it also helps us see the price points that people were most willing to commit to. The three largest ($10, $25, and $50) account for 6516 backings, or 39% of all backings. If you include the second tier of lower spikes ($15, $20, $30 and $40) then this rises to 61% of all backings. On the higher monetary end of Figure 1 there are still clear spikes. The $100 and $150 levels combined had 1109 backers. There were also smaller, but significant spikes at $200 (80), $250 (61), $300 (52), $500 (61), and finally $1000 had 33 backers. The highest backed funding level was $2500 with one backer for the Quantum Roleplaying Game project. It should be highlighted that the $10 (2063) and $50 (2018) level nearly match each other in terms of backers. People are either willing to pay a small sum, for what likely is a PDF, or they are willing to invest more in the project and get a printed copy (often a hardcover) plus the PDF, or some other mixture of rewards.
Dollar Value of Funding Levels 
With Figure 2 it can be seen that the value of individual backings has a real impact on the overall financial gain. While the $10 funding level has the most backers and raised $20,630, it was the $50 funding level that had the highest monetary yield of $100,900. That greater yield is also evident on the far end of the chart, with the $500 funding level yielding $30,500 and the $1000 level pulling in $33,000. Thirty three backers easily outspent the 2063 backers at the $10 level. The top five
monetary yields were respectively $50, $100, $25, $40, and $75.
These five brackets yielded $315,665 which is 39% of all the funding
raised in the study.
Big Spenders vs. the Little Guys While the amount raised on funding levels above the $100 bracket was not insignificant, totaling $218,256.00, this is only 26% of the overall money raised. The bulk of the money raised came from funding levels of $100 or below, which falls in line with the kind of spending a typical roleplayer might spent on a game system.
Successful vs. Unsuccessful Backing Behavior Figures 3 and 4 below show how overall backers behaved between these two groups:
 
Comparing Figure 3 with Figure 1 it is clear that the patterns are similar. Successful projects took in the bulk of backings and thus represent fairly well the overall trends. With Figure 4 the pattern flattens as the backings disperse across the span of funding levels. There are still spikes at $10 (90), $20 (49), $25 (200), $50 (107), and $75 (122).
Next is how these groups broke out at a median level.
 
As with total backings, the overall and successful spread of median backings reveals a similar pattern. However even with the median being calculated; there are extreme swings at work in these charts. The $9 and $11 levels (the first two lines in figures 5 and 6 which spike to near 40 and 60+) shoot up due to a very small sample at those funding levels where particular projects did well in that one slice of the funding. If the backings from $8 to $12 are combined into the $10 pool before the median is calculated, it yields an overall median result of just 10 backings. This fluctuation continues throughout the spread of samples, with spikes coming from isolated and oddly priced funding levels.
Creator Behavior Because the spread of funding levels causes many swings when looked at through the median, let us turn to how the creators behaved in these projects. What was being offered at what funding levels? With roleplaying games the primary products that are being created are books with rules or content to help support play. The three main formats are PDF, softcover and hardcover. Overall how were these priced?
   Caveats It should be noted that the data collected with these book formats was assessing the first instance in a project where that particular format was being offered. Not every project was clear in exactly what type of print book was offered. Determining if it was softcover or hardcover sometimes required guesswork or additional research from this author. Sometimes this was fairly easy due to the funding level, such as a print book being offered for under $20, which likely meant a softcover book. The assumption that was taken when there was no clear way of determining the difference was that the book was a softcover. Some projects did not offer any book as these were RPG related but were producing maps, cards and other alternative methods of play. This was only a small percentage of the entire survey. There was a great deal of variance in what formats were being offered. While most projects did offer a PDF once it came to a print product there was a decent degree of either offering a softcover or hardcover. A much smaller group offered all three formats.
Table 3: Overall Book Format Offerings
Table 4: Successful and Unsuccessful Format Offerings
Conclusions on Book Formats There were clear clusters of format prices. PDFs were primarily presented at $10, softcovers had a range of $20 to $30 with a solid peak at $25, and hardcovers peaked at $50, but showed decent representation from $40 to $60. When the $10/$25/$50 peaks are compared to Figure 1 it is clear how much these price points drive backer behavior, funneling their selection into specific levels. While these price points frame what backers can pay, it also shows how responsive backers are to these prices and gives a solid target that future crowdfunding projects should consider for their campaigns.
Custom Artwork and Content There was a huge range of special rewards offered in the survey. Two that were fairly common and this author tracked were custom artwork and custom content. Custom artwork is defined as artwork that a backer would provide direct input into the creation of the art. Some examples would be sending in a photo of yourself and having an artist use that to create artwork with your likeness to be put into the book. Another would be the backer defining the features of a character which an artist then uses to create the image. Overall there were 44 offerings of custom artwork in the survey, of which 31 were involved with successful projects. Custom content is similar in that the backer is able to provide direct input into the development of content, or even hand in their own content to the creators, who then integrated it into the final work. One approach that was not tallied in this category was naming rights. It was rather common for a creator to offer a backer the chance to either invent a name, or have an NPC or other game setting feature named after the backer. This report only focused on offerings that involved some real degree of collaboration with a backer. Overall there were 100 offerings of custom content in the survey, of which 72 involved successful projects.
  On a purely observational level this author noted while going through the data that both of these categories were almost always being backed within the successful pool of projects, with only some of the highest funding levels being passed over some of the time. Even in the unsuccessful pool it was observed that among the higher priced funding levels these categories were being funded even if the lower tier offerings (such as just a PDF or print copy) were lacking support. The real spikes for both of these offerings occur in the $150, $200, $250, $500 and $1000 levels. This lines up well with Figure 1 where it was shown that these price points showed healthy activity. Overall $150 had 347 backings, $200 had 80 backings, and $250 had 61 backings, $500 likewise had 61 backings and $1000 had 33 backings.
Part Two Conclusions The performance of a crowdfunding project is impacted by many factors. A primary one is how the funding levels are priced and what is being offered within those levels. One mistake that many unsuccessful projects seemed to make was to aim for a funding level that would require an unrealistic number of backers. The median raised per backer for successful projects is $41.51. The median number of backers is 110 for successful projects. So at the broadest of levels, the most that one should expect to raise is around $4566. However, that should not be the goal, but rather the unspoken (at least to the public) hope. Many unsuccessful projects aimed to this level or much higher and thus reduced the psychological momentum that can come from a project that looks strong and healthy. If a project is far off from its target, especially in the final weeks, then people are less likely to back the project simply because it looks as if it will not succeed. Backing is in part an emotional commitment and many people will avoid voting with their dollars on what is perceived to be a failing prospect. Another thought experiment would be to assemble a theoretical crowdfunding project with the various funding levels and offerings detailed in this report.
While $2250 is
only about half of what the median raised for successful projects,
this might be a modest view of what a crowdfunding project could aim
for in terms of results. It is after all close to the $2000 median
goal set for successful projects. By keeping to a realistic vision
of what is possible it is far easier to reach the funding goal, and
increases the chance that the project will take off and perhaps
double or more what was aimed for.
Part 3 The next segment will shift from raw numbers to compiled notes taken
by this author while going through the projects, highlighting project
offerings and what seemed to work or not work. - Neil Carr Join the expanded discussion at: theRPGsite, RPGnet, RPGgeek, EnWorld You can get a PDF version of this report here. I wanted to thank Annick Jean, Brian Leet and Tyler Dion for their assistance with this report.
 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Articles >