Crowded Clouds

One of the interesting features of the Cloud that hasn’t been mentioned so far, is that as well as being able to reach out through the Internet across the world to others, there’s the potential to receive some form of data not just from a select few scattered about the globe, but from hundreds or even hundreds of thousands of people.

This can seem daunting, or even undesirable, at first glance; after all, there’s little need for input from so many strangers for what’s ostensibly a creative endeavour. However, for a few select situations, the act of “crowdsourcing”, or getting some form of involvement from other people, can be extremely beneficial.

Crowdfunding

Consider large-scale cash lotteries—everyone spends a small amount of money for the chance to win a huge amount of money. The odds are massively stacked against every individual, of course, but on the other hand the risk is small enough (at least in the short-term) that it doesn’t really matter. Spending a dollar for the possibility of a return of millions of dollars doesn’t sound like a bad deal, statistical probability notwithstanding. However, when participants inevitably lose, they’re unlikely to derive much satisfaction at the prospect of having made someone else a millionaire.

Now consider that same large-scale, low-risk contribution to something that does have a more positive outcome—funding a production. Traditionally, productions are funded by a small number of parties each contributing a large part of the budget, with the hope of getting a high enough return on investment on its eventual sale to be worthwhile. But with “crowdfunding”, instead a large number of people each contribute a much smaller amount of money, but typically don’t expect anything in return.

Figure 16.1 Traditional Funding

Figure 16.2 Crowdfunding

If you’re rolling your eyes at this point, consider that in 2014, $16 billion dollars was raised through crowdfunding, of which $1.97 billion was directed towards film and the performing arts. The World Bank expects the overall value of crowdfunding to increase to $90 billion by 2020, effectively tripling the value of venture capital globally. Crowdfunding has been used to raise an estimated $6 million from approximately 90,000 “backers” for the production of Veronica Mars (2014) and just over $6 million from approximately 50,000 backers for the production of a new series of Mystery Science Theater 3000. Perhaps even more crucially, it’s likely that neither of those projects would have ever been greenlit via more traditional means.

Clearly with a crowdfunded approach, there are vastly different considerations to be made. On one hand, with each individual contribution being so small, there’s no-one who can claim a stake in the production and start making demands or assuming some form of creative control. However, at the same time, few of the contributors will likely be happy to simply hand over a sum of money and walk away, even if it’s a comparatively insignificant amount, which means you have to provide them with something in exchange, be it a forum to voice their ideas and opinions, or some regular feedback to give them an insight into how things are going.

Not having external involvement is not always a good thing, however. There are a great many stories of high-profile crowdfunded projects that secured a lot of money in funding (to the tune of millions of dollars in some cases) but that subsequently went off the rails in terms of their development, having been unable to manage its funds or schedule correctly. There’s also the problem of distribution to consider—it’s one matter to raise money to shoot a film but quite another to get it out in front of paying people.

For these reasons, it seems that the optimal approach, at least for the time being, is to use crowdfunding first as a way to prove to a studio or television station that there’s significant interest in a project, and second as a way to jump-start the marketing process by generating excitement about a project. That’s unless you’re trying to raise funds for a low-budget production, in which case a successful crowdfunding campaign can mean all the difference in the funding of the production and support in getting it made (2013’s Academy Award for Best Documentary Short Subject went to Inocente, having been crowdfunded to the tune of $52,527, and after raising $406,237 on Kickstarter, 2016’s Anomalisa was nominated for Best Animated Feature Film).

There are plenty of challenges associated with raising money through crowdfunding, however. First and foremost is attracting enough attention to be viable. Crowdfunding depends on a significant number of people willing to give you even a small amount of money on the promise that you’ll use it to make something they’d be interested in seeing. To do that you have to either have a million followers on Twitter or you need to do some marketing to raise awareness for a project before you even start asking for money.

Furthermore, you need to have some way of handling the logistics of crowd-funding. Potential funders need a way to actually give you their money in a convenient and secure way, and there need to be enough controls in place in the process to ensure that both you and they are protected in the exchange. Then there needs to be a way for funders to receive things like progress reports, but also for them to be able to communicate, if not with the production directly, then at least with each other.

Fortunately several platforms already exist to make this as simple as possible. Sites like Kickstarter (kickstarter.com) and IndieGoGo (indiegogo.com) allow producers to easily create a project page that has the details of a proposed production, along with a deadline date and a minimum target to meet by then in order for the project to be viable. People can then “back” projects for any amount they choose, either having discovered it on the site directly or being referred there from somewhere else. The sites themselves handle all the logistics of managing backers’ pledges, providing any technical support with regards to using the site. Meanwhile, there’s an easy way for producers to get analytics data about the backers, as well as provide regular “updates” to backers on the project, and provide the ability for the backers to comment and discuss each of the updates. In return for all this, the sites take a small percentage of the total amount funded if the campaign is successful in reaching its funding goal (if unsuccessful, typically the campaign ends without backers being charged, though this depends on the specific project).

There are some additional steps producers can take to make their pitch more appealing to potential backers. It’s common for campaigns to offer various “rewards” to backers should the funding be successful. These are usually tiered (backing for a higher amount of money qualifies the backer for a better reward), and can be, for example, a credit or acknowledgment somewhere (not necessarily on-screen, but typically on an accompanying website), a copy of the production to watch once complete, the opportunity to meet key people from the production, some memorabilia (such as a signed prop), or even an invitation to an event (such as a screening). Rewards like this can be a good way to encourage people to pledge a higher amount than they would otherwise.

Box 16.1 Tips for Running a Successful Crowdfunding Campaign

The following is a list of things that the most successful crowdfunding campaigns seem to have in common, as well as potential pitfalls to avoid when pitching a project:

• ▸ Set expectations clearly. Don’t let backers think that they’ll be contributing more than they actually will be, and don’t give the impression that the production will be finished sooner than is realistic.
• ▸ Give backers a voice. Conversely, make sure backers have a way to be able to express their opinions and feel like they are valued (something which becomes much harder with more backers).
• ▸ Beware offering physical rewards. Physical rewards require shipping and come with a whole lot of logistical issues to deal with. Some successful crowdfunded projects failed in the long-term due to the cost of providing such rewards.
• ▸ Build interest before the campaign starts. Ultimately the success of the campaign will be due to a large extent on people actually finding out about it. Once a campaign starts, success may lead to a “virtuous circle” effect whereby the success of the campaign actually draws in more interest and attention, but there’s no doubt you should aim to start off strong, as a project seen to be floundering will likely drive people away.
• ▸ Show that you’ve got what it takes. Many backers are cynical, particularly those who have previously backed projects that were mismanaged or ultimately failed in some way. Outline your plans as clearly as possible and describe the experience of everyone involved to instil confidence.
• ▸ Explain why crowdfunding is necessary. If a studio isn’t prepared to back your production, why should anyone else?
• ▸ Provide regular updates. Even though just running a campaign can be hard work, regular updates reassure backers (and potential backers) that you haven’t thrown in the towel. This should continue in some form even after the campaign has ended, as it can leave backers with the confidence to back any future campaigns you run.
• ▸ Know that timing is everything. Campaigns that launch during summer months tend to attract fewer backers.
• ▸ Hire a crowdfunding support person. For bigger campaigns, dealing with the crowdfunding process and related issues can be extremely time-consuming. It’s best to hire someone (or multiple people as needed) to help keep the campaign running and possibly keep everything going smoothly even after the campaign ends.

Creative Crowdsourcing

The concept of outsourcing is not new. If some work is needed but a production needs expertise they don’t have or (more often) they need to do it for less money, outsourcing becomes a viable option. With the Cloud, outsourcing in this way becomes even easier, provided you’re willing to forgo some secrecy.

Various sites act as agents for creatives available for work, and several different approaches are used. With the most familiar, as used on sites such as Freelancer (freelancer.com), you provide a brief about some work that needs doing, and then receive bids from available freelancers willing to do the work, or alternatively browse through the profiles of freelancers available to do work. This is a relatively intuitive and quick process, with the downside that there’s perhaps some risk involved. Although most reputable crowdsourcing services provide ways to prevent people taking money and not delivering the work (or conversely, protect freelancers from having submitted the work and then not getting paid), there’s still no guarantee that the quality of work will be to the standard required.

That’s where the “contest” approach comes in. Rather than selecting from submitted proposals and hoping that the final work submitted will be sufficient, some services instead allow holding a contest. An organisation submits a brief as before, but this time freelancers will do the work on spec and submit finished work in the hope they are selected as the winner. The organisation then has several finished options to choose between, from which they select one to be the “winner” of the contest. The winner then gets paid for the work, and everyone else goes away empty-handed.

It’s a system that’s undoubtedly skewed in favour of the organisation requesting the work; the majority of freelancers “competing” to get paid still end up doing the work. Indeed, AIGA, the professional association for design, strongly discourages professional designers from participating in work where there’s a risk of not getting paid. Still, if what you’re looking for is low-cost work of adequate quality, this could be a viable approach.

Services like this exist for practically every creative endeavour. Crowd Studio (crowdstudio.com) provides a way to get bespoke music and audio for projects, whilst Tongal (tongal.com) provides videography. Numerous sites exist for sourcing 2D design for logos, branding, and so on, and still others exist for translating content, and most of them are readily accessible and inexpensive.

Box 16.2 Coworking

The flip side of being a lone freelancer connected via the Cloud in order to do work is that you lose some of the benefits of working in an environment with other people, such as having a sense of community or being around others to bounce ideas off or just socialise. As an increasing number of people are working outside of a traditional office, so too has there been a rise in the number of people looking to reconnect to some extent with a working environment.

“Coworking” groups are intended to cater to those needs. These can be informal groups of people who work for different organisations but work out of a single location (such as one person’s home or a cafe). In this way, each member gets the social connection they’re otherwise missing but with a greater degree of flexibility (if you don’t get on with a particular group, you can just leave).

There are also more formal variations of this, whereby companies such as WeWork (wework.com) rent out shared office space to people coming from a variety of different organisations. Although these tend to be expensive compared to having a single company leasing space for all its employees, they prove popular due to the accessibility, security, and psychological benefits they provide.

Microtasking

The options for getting important tasks done through crowdsourcing as discussed so far are referred to as “macrotasking”, whereby a small number of people are selected from potentially hundreds in order to carry out the work. The opposite is also possible—sourcing hundreds of people to perform numerous basic manual tasks—and is referred to as “microtasking”.

Microtasks can range from things like looking at images of receipts and working out which company each of them was from to fact-checking data from a spreadsheet. They are things that are too complicated or too impractical to be able to get a computer to do but that a person can do without much effort. When faced with having to get a significant number of micro-tasks done in a specific timeframe, you can either hire someone to spend time completing them all or you can turn to the Cloud.

Services such as Amazon’s Mechanical Turk (mturk.com) aim to make this process as accessible as possible. Anyone with Internet access can opt to be a worker on the services (though having specific skills, or having been a worker for a certain period of time, allows them to be able to apply for a greater variety of work), and anyone can submit “Human Intelligence Tasks” (HITs) that they need doing. Requesters submit a short brief about what’s needed, an allotted time to complete each task, any qualifications that workers should have, and then a budget for getting the work done.

There are numerous applications for this approach. It can be an easy way to get reams of data verified in a particular way (or, for example, dealing with duplication issues in data). It’s possible to get images classified and categorised (for example, from a pool of 5,000 images, determine whether the main character is wearing a blue dress or a red one). It can even be used for research purposes (for example, to see if people prefer one particular piece of music over another). If you really wanted, you could conduct test screenings in this way (though you’d probably want set up the request such that workers would need to sign an NDA first).

The headline benefit of this system is that it is cheaper perhaps than any other approach. Workers typically earn around $1–$1.50 per hour doing the work, which is notably lower than the minimum wage (a fact that has caused these services to be heavily criticised), although that amount increases slightly for workers who are more “recognised”, having performed more tasks successfully for the service in the past.

The other benefit is the time it takes to complete the work. By splitting the tasks up amongst potentially hundreds of people, you can get them all to work in parallel, which means it will get done a lot faster than one person doing them all in sequence (and when you consider that person will not work 24 hours a day).

Box 16.3 Physical Microtasking

All of the microtasking approaches covered so far involve manipulating data of some sort, whether in the form of a video file or a spreadsheet of data to check. However, there are also options for crowdsourcing things that need to be done offline, like collecting packages or booking a hotel reservation over the phone.

Services like TaskRabbit (taskrabbit.com) follow a similar model to Mechanical Turk but instead allow people to be recruited to perform oneoff physical tasks. These can involve, for example, waiting in line, running errands, cleaning, or doing some research. Likewise, Uber (uber.com) uses crowdsourcing to provide private transportation services to people.

Crowd Voting

The final aspect of crowdsourcing is crowd voting. Put simply, crowd voting is getting a large number of people to vote on a topic. This approach has been popularised by sites like Yelp (yelp.com) and TripAdvisor (tripadvisor.com), which collate and aggregate user-submitted reviews for local business and travel destinations respectively. More subtle implementations can also be seen across numerous commerce-based sites where users submit ratings and reviews for products. Perhaps even more prominently, crowd voting is put to use on a number of high-profile televised performance competitions such as The Voice (nbc.com/the-voice), wherein viewers collectively get to vote on who they want to win.

As useful as crowd voting might be, there’s not actually any easy way to tap into that sort of power unless you have your own established, large community to draw upon as needed (unless you just happen to be making the next series of a hugely popular TV series or a sequel to a blockbuster with a large fanbase). The alternative is to leverage something like Mechanical Turk, and pay people to vote on various topics, or else get people to take part in an online survey somewhere.

Box 16.4 The Wisdom of the Crowd

A novel experiment taking place in the Cloud is “A melody written by a crowd” (crowdsound.net). Visitors to the site are offered the opportunity to help in the creation of an entire melody, written note by note, using votes to determine what each note should be, with each note in the sequence based on 50 submitted votes on a number of options.

Following on from this, the site aims to run another experiment, “Lyrics Project”, whereby a similar setup will allow visitors to vote on each word used as lyrics for a pre-composed melody. With both experiments, the idea is to create something not only unique but also based on a diverse range of input, something that might considered extraordinarily difficult to accomplish any other way.

Popular Crowdsourcing Services