This section encompassed eight questions, designed to measure respondents’ knowledge of ebook production processes and the extent to which small publishers have invested ‘human capital’ (Thompson, 2010, p. 4) in ebook production:
Interestingly, only 12% of publishers have been making ebooks for over ten years – in other words, since before the launch of the Kindle and ebooks’ commercial breakthrough in 2007. Only a small group of publishers started making ebooks within the last two years, perhaps indicating that most publishers who intend to make ebooks are already doing so. This conclusion is supported by data from Aptara (2012, p. 6) and BookNet Canada (2015, p. 9), both of which showed decreasing numbers of publishers who planned to begin selling ebooks in the near future.
There does not appear to be any relationship between the length of time a publisher has been making ebooks and the proportion of revenue those ebooks represent. In fact, no publisher making ebooks for more than ten years reported making more than 30% of its revenues this way – perhaps because ebook-only publishers were not economically viable at the time, and more nimble entrants have since overtaken the early adopters. Interestingly, most publishers earning over 50% of their revenue from ebook sales have been operating for between four and six years.
It is worth noting here that there was no way to control for the length of time a publisher has been in operation, so the predominance of publishers reporting selling ebooks for just a few years may simply be a function of the relatively short lifespan of many small publishers.
Less than a third of small publishers produce their ebooks entirely in-house. This question is somewhat comparable to Aptara’s question ‘to what extent are/will your eBooks be produced by an external partner/vendor?’, although the possible answers are phrased differently (2012, p. 30). Aptara’s respondents reported a similar rate of externalisation: 71% produced at least some ebooks externally, compared with 70% in my survey. The pattern is similar in South Africa, with seven of the ten publishers surveyed by Gaigher, Le Roux, and Bothma reporting outsourcing at least some aspects of their ebook production (2014, p. 266).
There does not seem to be any clear production pattern when the results are broken down by country.
In a survey of information technology (IT) managers by Aubert, Rivard, and Patry, it was found that IT activities were more likely to be outsourced when they had low ‘uncertainty’ (2004, p. 929). Two types of uncertainty were identified in the context of outsourcing: ‘demand uncertainty’, where the outsourcing business cannot initially determine its requirements, and ‘measurement uncertainty’, where the outsourcing company is unable to accurately evaluate the product or service provided in terms of its quality, timeliness, or price (2004, pp. 922-923).
It seems likely that ebook production is a field vulnerable to both demand uncertainty and measurement uncertainty: both side effects, to some extent, of less than comprehensive in-house knowledge (or ‘human capital’). Indeed, Aptara’s survey found that 55% of publishers outsourced production due to a ‘lack of internal capability and/or resources’ (2012, p. 31).
This question was intended to explore anecdotal evidence gathered from informal discussions with publishers and ebook producers that suggested fixed layout ebooks are expensive to create and generate low revenue. Generally, publishers do not produce fixed layout ebooks except for special titles in genres like cookery, and they are only accessible on certain devices like the Apple iPad, Kobo tablets, and the Kindle Fire (eBook Architects, 2015). Only 78% of respondents answered this question – a unusually low response rate when compared with the rest of the survey, with every other question achieving a response rate above 95%. It seems reasonable to assume, too, that respondents may not have realised they could skip this question, and thus some of the ‘neither agree nor disagree’ responses may in fact have been intended to indicate that the respondent publisher did not produce fixed layout ebooks at all. It can be inferred from this unusually low response rate that a significant proportion of the respondents do not produce fixed layout books. The result indicates that publishers are not deeply interested in fixed layout ebooks, with only 4% of those surveyed expressing a strong opinion in one direction or the other. The Canadian BookNet survey supports this hypothesis, with 52% of publishers reporting that they do not produce fixed layout ebooks, with 12% unsure (2015, p. 14).
This provides an interesting contrast with the Aptara study, which found that 58% of the respondent publishers produced fixed layout books (2012, p. 22). It should be noted, however, that Aptara sells fixed layout conversion services, and is likely to have targeted this survey to its customers, so the results may not have been representative.
The majority of small publishers are satisfied with their in-house ebook production skills. In fact, given the extent to which publishers externalise ebook production (see Figure 10 above), the level of satisfaction is surprisingly high. If scores are weighted (with ‘very dissatisfied’ equivalent to a score of 1 and ‘very satisfied’ equivalent to a score of 5), the mean satisfaction score is 3.56.
Perhaps unsurprisingly, the greater the proportion of revenue ebooks represent at a publisher, the more likely that publisher is to be satisfied with its in-house production skills. No publisher making more than 10% of its revenues from ebook sales reported any degree of dissatisfaction with its in-house ebook production skills.
It is possible that publishers may be increasing their reliance on outsourcing: a recent Canadian survey found that 39% of respondents had dedicated digital staff in 2014, a decline from 44% in 2013 (BookNet Canada, 2015, p. 7).
Publishers are marginally less satisfied with ebook knowledge among non-production staff, but still satisfied overall. If the satisfaction score calculated above for production staff is repeated here, the mean satisfaction for non-production staff is a marginally lower 3.26.
Given the respondent publishers’ small size, this data should be treated with caution. It is likely that the respondents’ size necessitates some level of job sharing between their employees – if they have any! Many publishers may not have any staff members working purely in a ‘production’ or ‘non-production’ role.
Over a third of the respondents reported spending between USD$50 and $100 on the production of an average ebook. Worryingly, a number of publishers were not aware of their ebook production costs. Aptara found that cost was not a major factor in outsourcing decisions, with only 11% of respondents using external partners with the aim of lowering their costs (2012, p. 31).
However, as it is expressed in this survey the question may have been overly reductive. Susan Hawthorne of Spinifex Press points out that when calculating the cost of producing ebooks it is important to consider other externalities such as training, legal fees, the time spent creating metadata and checking files, distribution costs, and other incidental and future costs (Hawthorne, 2015, no page). It is hoped that future research will investigate these less tangible costs in greater depth.
Interestingly, when questioned about production costs, publishers skewed towards feeling they were low: 39% of respondents felt ebook production costs were ‘low’ or ‘very low’. There was no clear relationship between publishers’ spend on ebook production and their opinions about that cost: in fact, the only publisher to regard ebook production costs as ‘very high’ paid less than $50 per ebook. The consensus that ebook production costs are not ‘high’ – and hence lower than the market can bear – may indicate there is some room in the ebook production market for higher-priced specialist freelancers.
Ebook production is a speedy business: two thirds of respondents reported their ebook production taking two weeks or less. In-house production was marginally faster than outsourced production, a conclusion which fits with Aptara’s finding that after poor quality, long turnaround time was the external production issue of most concern to publishers (2012, p. 32).