research
at workes
BRAND2 180111

Archive for the ‘sxsw’ Category

Blog, sxsw

Facebook Twitter

SXSW – Research Twitter-style

  • Date March 15 2011
  • Posted by Job
  • Tagged with
0

Face’s Job Muscroft checked out Twitter’s Mark Tramell discussing Twitter’s research methods.

Tramell’s talk on day 2 of SXSWi was billed as “Stop Listening To Your Customers”, but a more accurate description of the Twitter approach would be Don’t Ask – Just Watch.

Tramell discussed Twitter’s methodology, and gave a five-stage process for understanding where their service could be improved:

1. Twitter have a single minded philosophy – to observe current user behaviour and then, as Mark puts it, “reduce friction”.

2. When they find a user friction point they will look for current user work arounds or hacks and investigate their validity on a larger scale.

3. Sprint to prototype, developing a live demo of the new service for users to trial in a matter of weeks.

4. The trial is a user clinic format where the design and development team come together and watch face to face to see how users interact with the new service.

5. In the final stage, iteration, they repeat the user clinic process up to 16 times before they go live with a new service.

The role of research at Twitter is crucial as pretty much all of their product development has come by learning from their users what works and what doesn’t. By watching how people use the site, Twitter have developed numerous functionalities including the now invaluable #hashtag, @reply and lists.

There is something very powerful in the way Twitter approach research and I think any researcher should ask what prototypes we can build (going beyond concept boards) to help consumers show us – rather than tell us – what they want.

Chris Poole gave a great key note presentation at SXSW talking about the evolution of the infamous meme making bulletin board 4chan - once described by The Guardian as “lunatic, juvenile… brilliant, ridiculous and alarming”.

4chan is one of the ugliest sites you will come across but it gets 12million users per month and has 1 million registers users. So what is the key to its popularity?

1. It is a simple concept. Upload an image and a comment and see if other people interact with your content.

2. Unlike social networks, users of the site are anonymous and have a freedom to play and express themselves in ways you just can’t on Facebook…hence some of the adult material uploaded but more significantly the large amount of art criticism on the board.

3. Only the most engaging content stays on the site meaning that people encourage others to play, comment on and adapt their content.

Poole’s next move is updating the 4chan concept by launching Canvas http://canv.as/. On Canv.as all users are given photoshop quality tools to encourage them to be more creative and it  also removes a lot of the barriers to mass participation associated with 4chan.

What struck me about the success of 4chan is how it has managed to create the perfect environment for innovation. A stripped back environment where ideas are more important than the creators and where those ideas can spread and grow without egos getting in the way. What I really like about the new Canvas concept is the potential it has to democratise crowd creativity and as someone who works in the field of innovation I find this very exciting.

Face Community Manager Chris Harding sat in on Aza Raskin’s talk on Behavioural Change and came away with the Mozilla Labs co-founder’s one secret for successfully modifying behaviours – both online and off.

The key to all behaviour modification is to produce feedback loops. In other words, if people understand that a certain behaviour produces a certain result when it’s performed, their inclination to carry out that behaviour is dramatically influenced.

Feedback loops come in various forms, but ultimately most loops fall into one of two categories – either positive or negative. Positive feedback loops reinforce a behaviour by providing reward each time the criteria are met. A great example of this came in a study involving kids left in a room with a marshmallow. They were told that if they didn’t eat that marshmallow for 15 minutes, then they’d be given another.

Quite apart from the weird behavioural ticks that the tortured-looking kids develop to deal with delaying their gratification, it’s noticeable that once the kids understand that their actions (or lack of action) will have short-term positive consequences, their usual tendencies (to wolf the marshmallow as soon as look at it) are modified.

However, these immediate feedback loops can be negative. Raskin brought up the example of smoking – a reflection of his interest in healthcare as founder of Massive Health – to show how some feedback loops are negative. The immediate positive feedback from smoking – the effect of the nicotine – influences behaviour much more than the incremental feedback loop of later health issues. Raskin believes that we can massively improve our own health by changing the feedback loop and, thereby, our behaviours. This has been proven to work by dieters. If one writes down everything one eats, the hard evidence of gluttony induces a negative feedback loop – guilt – and influences behaviour. Raskin would take it further, though.

For example, I imagine almost everyone reading this blog has been guilty at some point of failing to finish a course of antibiotics. As soon as we’re on our feet again, they tend to slip into the background as our lives (and beer) beckon. This is a risky business and slows recovery. So Raskin wants every course of antibiotics to prescribed with a shot in the finger which produces a large brown splodge. The mark reacts to the chemicals in the antibiotics and will only fade once the entire course has been taken – this short-term feedback loop reinforces the positive behaviour of finishing the antibiotics by providing reward.

In an online business sense, the instant feedback that a customer receives directly affects their tendency to conduct transactions on your site. Amazon released data revealing that for every 100ms of reduced load time (less than half human reaction speed) sales increased by 1%. The annoyance of longer loading times, even at a sub-reaction speed level, directly modified behaviour and reduced the tendency to spend at Amazon.

For Face, this reinforces the knowledge that our platforms need to be as intuitive and user-friendly as possible. Feedback loops are also relevant in how we engage our users. To encourage the best possible content, there needs to be an ongoing system of reward for consistently high-quality input. These rewards needn’t necessarily be monetary – in the online world kudos and respect is almost as desired as renumeration, hence the presence of “like” and “recommend” buttons on nearly all of the most engaged online communities.

In the second of our SXSW Interactive blogs, Face digital project manager Marion Renoux digests her first day of conferences. She dragged herself away from the eat-as-much-as-you-like breakfast buffet for “How not to design like a developer” with Chrissie Brodigan @ Mozilla & “The New Frontiers of Social Gaming” with Brian Reynolds, chief game designer @ Zynga. First up, “how not to design like a developer” and, after the stars, “the new frontiers”.

In “How Not to Design Like a Developer”, Chrissie was discussing the challenges of open source projects and how to facilitate a better collaboration between developers and designers, especially crucial in the context of open source projects.

Although the Face digital team is not working on an open source model, Chrissie made a pertinent point about the best practice to follow in order to reconciliate feature development and user experience:

1. Be careful with workaround: developers think about workarounds as a way to speed up the project completion, but it’s completely wrong from a UX point of view.

2. Even if you give the best layered Photoshop file, without proper documentation and walk through there’s a high chance the design will be misinterpreted in development.

3. Designers are guilty of being trendy, while developers build for the future: better find a good compromise between both directions.

4. Designers must practice a team friendly version control – like developers do with version control and code review – because, for example, you may want to rollback to an earlier version of the design, or get new designers to pick up design and make it better

***

Brian Reynolds was happy to surrender secret of a successful social game – for instance the worldwide known Farmville.

To be honest, I’ve always abhorred this game and all the stupid updates I would get in my Facebook feed. Between you and me, I’ve quite frankly considered defriending people playing too much of a farmer for my taste…

However, this presentation made me reconsider my opinion on the game. After some polite rambling on how Facebook changed the face of the digital world (I don’t know why every talk has a part dedicated to how The World has been revolutionised by The Social Network) Reynolds presented Social Gaming with another angle: basically it’s not the gaming aspect but the socialising aspect that matters.

So here are the secrets he revealed for a good social game:

1. Give the game away for free. The idea is that because the social game is based on people interactions, you must not restrict its access and everybody should be able to participate/socialise/engage with no monetary barriers. However, upgrades are paid: give away the blades, sell the razors (a bit dark but quite an efficient metaphor)

2. Let people express themselves and don’t block their creativity. Even better, give them the tools to create.

3. Obviously always let the players share and socialise because what they really want to do on Facebook is to learn about their friends and secretly shout “Remember me! Start a conversation, I’m lonely!”

4. Learn from your users. They may interpret and appropriate themselves with your game in a different way to that you’ve initial envisioned and designed it for. By observing their online behaviour you can improve the game, taking it in the right direction for the pleasure of your audience

MD of Face Job Muscroft reports from his first two sessions at SXSW Interactive in Austin, Texas. Job tore himself away from the Texan sunshine to check out the “Marketing budgets have gone social – is it working?” #marketingsocial panel with General Mills, Pepsico and Samsung & the “Do agencies need to think like software companies?” talk with representatives from Google, Simple Geo, Barbarian Group and Tribal DDB.

Brands are taking social media very seriously and starting to spend big.

According to @pepsicojulie, Pepsico have increased investment by 30% across their 500 brands in 2011 while EMarketer report predicts the tipping point will happen in 2012 when 60% of all marketing budgets will be social.

Samsung’s Kris Narayanan explained there were 4 key areas in which this increased budget is being spent:

1. Listening – what’s being said about the brand?
2. People – communicators and social media strategists.
3. Product campaigns – full-on cross-platform social campaigns.
4. Brand platforms – proprietary streams and other platforms.

However it was clear that measuring ROI is still a nascent science for brands.

At present social media KPIs are in their infancy and are being set for consumer engagement and conversation sentiment alongside traditional brand trackers. This is because these measures can be benchmarked accurately and measured over time.

Clearly this is not enough and the next step for brands is to overlay other data sets – for example web analytics, sales and geo location – to understand ROI.

***

So we know clients are allocating serious budgets to social media and naturally their agencies are asking themselves how do they structure themselves to meet this new demand?

The answer is that they, just like their clients, are still working this out. Most agencies are led by a generation who don’t understand technology and where the command and control culture still dominates.

How can this culture be changed?

The general consensus is that there’s a need to hire more technologists. However, Rick Webb from Barbarian also emphasised the need to liberate technologists from a department by getting them to work across the agency influencing strategy, ideation and execution while Matt from Simple Geo talked about the benefits of staying close to emerging technologies and startups to inform the organisation of any implications and applications.

A key learning from Rob Rasmussen from Tribal DDB was to focus energy on first developing long term engagement platforms for and build things on top to help engage and understand consumers.

Ben Malbon from Google Labs encouraged a structure where smaller teams have the remit to do things fast, who think by making and build business cases by piloting rather than by Powerpoint.

But what does this mean for researchers? Well for me there are two clear lessons and areas  for #MRX to address.

1. Clients need help understanding social media data to develop ROI models, who are better placed than objective researchers to help them crack this challenge

2. However, research agencies need to adapt and change just like ad agencies. We need to hire technologists and a new breed of researcher that understands the social media ecosystem and data so that we can continue to play the role of strategic adviser