Every time a client asks us to conduct design research we’re faced with two significant and often conflicting challenges: limited time to run the study; and finding the right balance between breadth and depth of data collected.
Now, thanks to the efforts Jon Snydal, Celine Pering and Catherine Sun of our San Francisco design studio we now have a new, useful addition to the data collector’s toolbox that goes some way to addressing both of these issues and along the way, channels everyone’s inner design researcher.
We recognise that this is one method amongst many and that the ultimate value depends on how the data is used – for example our frogMobs have fed directly into our ideation processes.
Of course crowdsourcing is hardly new: one of the world’s most popular sites relies on contributions from the ethersphere; a number of trend agencies essentially rely on unpaid contributions from a network of ‘trend scouts’ with varying degrees of success. In practice it can overlap with user diaries and cultural probes. (My own dabblings in this space include engaging shanty town communities in the design process; and the fivedollarcomparison.org site). Aficionados may wish to read Clay Shirky’s Cognitive Surplus – the rich thread of review comments on that page underlining the point.
Perhaps closer to the spirit of frogMob is the popular and enduring What’s in Your Bag flickr pool – a wonderful example of a focussed, inclusive and interesting research topic. That the vocabulary of the What’s in Your Bag flickr pool moderators has evolved to describe inappropriate submissions is also instructive – it includes a wonderful subset of FLAS (four letter acronyms) that reflect the types of biases that you are likely to see with pure public crowdsourcing: PESO – Photographic Equipment Show Off (they have their own flickr pool); SOSO – Single Object Show Off (they do too); POP – Pet Or Person i.e. the cat in the bag photo; WIMP – What’s In Your Pocket; and the more focussed MUBS – Make-Up Bags. The main drawback of a crowdsourced flickr pool as a design research method is that it lacks context. Who is the owner? What are they like? Where where they going? At what time of the day?
Crowdsourced data collection also comes with a number of inherent biases: the process appeals a particular kind of person; without suitable moderation it is easy for contributers to wander off-topic or use it as a vehicle for their own ends; and (in this instance) it is limited to people with access to a camera and internet. People also tend to be self selective – choosing days when their activities are more ‘interesting’ or aligned with the character they wish to project. People also avoid the documentation of certain items. In the flickr pool you’ll struggle to find sanitary products, condoms, pornography, seditious literature or Grade A pharmaceuticals – yet we know from our own in-depth research that these are commonly carried.
The considerable upside of frogMob over traditional methods is that it allows us to engage a wider and more geographically distributed group of people that would be impossible to otherwise reach; it can be fun for the particpants (one might argue it needs to be fun) – and for many the process will help them reflect upon and enrich their own understanding of the world around them. It also enagages participants at just the right level – the crowd makes it possible to step away from commitment without it being socially awkward; a particpant can take 1 or any number of photos depending on their time and inclination. And with a deadline of only 1 week, it allows people to snap, upload and move on.
Some of the biases can be diluted or filtered out by a team that understands the strengths and weaknesses of what is being collected through mechanisms and tools to help the community moderate itself. For example our friends at Ushahidi have pushed the boundaries in this space in the way that they validate and filter submissions.
For me the current strength of frogMob lies not in crowd-sourcing per se, but in the way it reaches out to our biggest asset – a multi-cultrual and globally distributed mix of ~500 employees – we have a primed, willing and motivated creative community with a vested interest in making the process a success. It’s far easier to move to more public crowdsourcing when the content is seeded by the enthusiasm of an existing community.
Like all good methods – it’s a work in progress, and over time we’ll share more of what we’ve learned.
Photo: taken as part of the ‘Power‘ frogMob – with a 7 month old daughter anything left at floor level will end up being chewed or eaten – furniture is positioned to cover electricity plugs, and cables are tucked out of the way.