I wrote last week about the role of incentives in online communities (and specifically how they can lead to behaviours that you might not want). In the post I give an example from Dan Ariely, a behavioural economist at MIT, that shows why incentives can breed unintended behaviours. Today I came across his YouTube channel and thought I would post his video explaining just this.
I’m actually working on the issue of incentivisation at the moment. Many clients we work with, building and managing online research communities, want to explore the different ways in which they can motivate people to take part in the community or in activities. Our approach is not to incentivise in the way traditional in market research (by giving financial incentives) as this moves the experience out of the social and into the market environment. We become traders of information rather than members of an online community of people working towards a shared goal, or with a shared set of interests and aims. However, sometimes we want to reward behaviours we want to encourage but keep this in the social context.
I’m working on a white paper discussing just these issues at the moment, so watch this space…
Thirteen people have paid a serious price for what they have done on Facebook. Most of us have been tagged in photos when we’d rather not have been, or had the odd inappropriate comment from a friend on our wall. But these 13 people have lost their jobs. They all worked for Virgin Atlantic and had discussed safety standards and said some less than polite things about passengers on Facebook.
The conversations had been between employees of the airline in a Facebook group. Among the many issues they discussed on the group they expressed concerns about some of the aircraft and apparently called some of the airline’s customers both cockroaches and chavs. They were fired for bringing the company into disrepute.
We can’t know the exact details of who discussed what, most of us only hear this news second-hand (at best) and have never seen the actual group where the discussions were had. But what we can learn is this - we should be careful what we say on Facebook.
Social networks are built on networks of friends. I have 175 friends on Facebook and my experience of the site is centred on this group of individuals. I get updates on their events, photos and statuses. My Facebook experience is limited to a fairly small group of people. But we are actually part of something much bigger and much more public. Conversations that, to those taking part, may be between a small group of their friends are actually much more public than this. This is how social networks work - bringing together small but inter-meshed groups of friends into a significantly larger group.
Now we’ve probably all talked about work gossip, probably with a small group of friends, privately in a bar or over dinner. What these Virgin employees did may have felt just like that - they were in a group with their friends sharing work gossip. The problem is that unlike that secluded table in the bar or restaurant, they were talking in a very public place. Perhaps the most public of places. This was their mistake.
Because of the nature of social networks, it is easy for us to believe that we are sharing our thoughts, photos and opinions just with our friends. The truth is that we’re probably sharing them with a much larger and more public group. We’ve posted before about social media manners, and this episode is a great example of how we need to take active control over what we discuss and what we don’t discuss online, who we’re friends with and who we choose not to be in our various social networks.
Amex to launch online community for travel managers
One of the best examples I know of a brand using information they have to add a social layer to their site is the Members Know site from American Express. On this site Amex use the data from spending using their cards to highlight restaurants and hotels in certain cities that are popular with their members. Once you’ve signed up you can share your thoughts on these establishments and exchange travel tips with other business traveller. Today, they are launching a new online community, and this time it isn’t aimed at the business travellers, but at the people who organise their travel for them.
Business Travel Connexion is aimed at corporate travel managers and will combine editorial from Amex and other suppliers with user-generated content. Amex hope to create a real-time resource for the members and also build a fairly homogeneous community of a group of people who would valuable to marketers. They will be able to share information and ideas with each other and also with Amex and other suppliers. The site includes a “Product Lab” area for feedback and co-creation.
So what can we learn from this?
Amex are a great example of how brands are adding social layers to their existing sites and products, delivering real value to people and making the most of the product and information they already have. With Members Know they took data that previously wasn’t used externally (data on spending in hotels and restaurants) and repackaged this in a way that was both useful for members and encouraged them to interact and upload their own content. In the same way, Business Travel Connection, links a set of individual customers who are isolated (often working with no peers in their organisation) but who share a strong common bond (they all deal with the same problems). That they can be linked through the Amex brand is even more powerful.
When thinking about ‘going social’ - building online communities or using social media - too many firms build approaches that don’t always address their unique position in the market or capitalise upon what they may have to offer. Amex have done things the right way. They’ve thought about their strategy and about why people would engage in an online community that they manage; and about what they have to offer that’s different. These are important stages and ones that we at FreshNetworks spend a lot of time on with clients. Working out why people will engage and why they will engage on your site is a critical first step to any online community.
There’s a debate to be had in the world of online communities, especially in the world of online research communities about how we incentivise people to take part. At FreshNetworks, in the online research communities that we build and manage for clients we tend not to incentivise, at least not in the traditional way that we see in market research. So for a community we ran earlier this year with C-level executives from across Europe we didn’t incentivise and got high levels of participation. If this had been a focus group or telephone survey, traditional market research would have included incentives (either to the individual or to a charity on their behalf) of over £150 per person. Some other agencies who build and manage online research communities always pay incentives, but as a norm we don’t.
I’m currently reading Dan Ariely’s Predictably Irrational (a great read for anybody who is interested in human behaviour!), and he discusses exactly the reason why not paying traditional incentives can be an effective strategy. We know that humans operate in both the social context and the market context and Ariely shows how the mere mention of money, quantifying an effort, is enough to place an experience firmly in the market context. Take the example of a lawyer who is asked to do some work to help a group of disadvantaged people - offer to pay them a reduced day rate and they will probably say no; ask them to do it pro bono (for free) and they will probably say yes. In the former case they think they are being devalued (a market context), in the latter they know they are doing it for free as a favour (a social context). The mere mention of money shifted the engagement to a market one.
Ariely describes an experiment that throws more light on how to encourage participation. They recruited a group of students to take part in an experiment - they were asked to spend five minutes doing a mundane task (moving a circle on the left of a computer screen into a box on the right) as many times as possible. One group were given five dollars to take part, another 50 cents and another nothing. The outcome was measurable - the more times the circle was put in the square the harder the participants worked. Whilst it was true that those paid five dollars worked harder than those paid 50 cents, it was those who were paid nothing who worked hardest of all.
A second iteration of the experiment strengthened the theory that mentioning money creates a market context. Rather than offering money they offered gifts - to one group a chocolate worth about 50 cents, to another a box of chocolates worth about five dollars and to a third group nothing. Money wasn’t mentioned, just the gift that they would received. This time all three groups performed equally.
The outcome from these experiments is clear. The mention of money as an incentive for doing something shifts the context from a social one to a market one. People make a decision of how much they will contribute based on the value they think they are receiving. Without the money (or even with a gift not described in monetary terms), people operate firmly in the social context.
When we are building and managing online communities we want people to take part in the social context. The communities are not market-based transactions, but social environments. Monetary incentives (or equivalents) will only go to create an environment at odds with this. That’s why we tend not to incentivise participation in our online communities, we find that we usually don’t need to. Where we do incentivise we tend to do so with gifts, information or access - non-monetary offerings which leave people firmly where we want them in the social context.
It seems to be a week for research. A couple of days ago, we reported on research from Rubicon Consulting showing that online reviews are the second biggest influencer on purchases for US adults. Today I’ve been reading research from Razorfish which looks into this in more detail and in particular the role of advertising and the place of brands in social networks. Their Consumer Experience Report shows that 40% of online consumers have made a decision based on an advertisement they saw on social media sites and that three-quarters welcome brand advertising in these spaces.
So consumers are, it would appear from this research, comfortable with brands being present and advertising in social networks. It it worth comparing, however, the 40% of people that have made a purchase decision based on an advert in a social network with the 51% of people in the research who have made a purchase through a recommendation in a social network. This would indicate that whilst advertising and the presence of brands is accepted, personal recommendation have 25% more influence on purchase decisions.
This research does not come as a surprise to us here at FreshNetworks. We know that social networks are very much a me space (I go to Facebook to see my friends, upload my photos and plan my events). It’s a very personal space and it is difficult in this environment for a brand to have as much impact as those people who are in your network. Advertising does work and people do accept it (possibly because they are used to advertising across the web pages they visit) but it cannot have the same impact as something that capitalises upon this network and personal space. It cannot, therefore, have as much impact as a personal recommendation from somebody else in the network (even if you don’t know them).
As we are hearing more and more, people trust people like them more than they trust an anonymous brand in an advert. They make purchase decisions based on things they read in social media, and on social networks, but are more likely to be influenced by recommendations from other users than they are from advertising. This is why word of mouth is so important in a social media context and why brands should be making the most of and amplifying the organic natural discussions that are out there, rather than necessarily advertising in the way they always have done.
We’ve written a lot in the past about co-creation, from Mass Customisation to Community Product Design. In some of the more devloped examples of co-creating with customers crowdsourcing is common - getting feedback and ideas from customers, the people who know your product best.
The online communities that we build and manage at FreshNetworks make use of crowd-sourcing for innovation and insight. We use them to help clients create better experiences when on holiday, or to come up with ways of marketing a product better. Other brands are also making use of their crowd - from Oxfam looking for a new slogan to Starbucks wanting to improve its product - it’s a technique that is being used more and more, possibly because social media now gives us the tools and the audience to do it easily.
An article in today’s Guardian (The customer knows best) looks at how crowdsourcing is being used by brands (and has an interview with our CEO Charlie). As the article points out, getting ideas and feedback from users is not new (they cite the example of the Oxford English Dictionary in the 19th century) but social media tools and, more importantly, their growth and common acceptance, is making it easier and quicker to seek ideas and opinions from customers. It’s also making it easier for smaller businesses to capitalise upon the power the crowd can bring. As the Guardian notes:
The evidence from Starbucks and P&G shows that some of the world’s biggest companies can easily engage a crowd. Smaller businesses, naturally, find it much harder to source an army of volunteers, let alone get them to engage with their brand. Recently, however, a relaunched service from Amazon - a pioneer of customer generated reviews - is creating a market that might be able connect the two. Companies subscribed to its Mechanical Turk (mturk.com) service can post simple tasks, such as image tagging, data collection, basic market research and product comparison, and offer to pay potential click-workers a few pence to complete them.
This is a real sign that crowdsourcing is becoming both more available and more widely accepted - more business (big and small) can make more use of their customers in this way.
Of course opening up your business to ideas and comments can throw up challenges - what happens if you get negativity as well as those positive useful comments we all expect. This is of course true, but as Charlie is quoted as saying in the article:
…some companies are “definitely nervous” about this new, more open form of business, especially in terms of “opening themselves up to positive and negative criticism online and encouraging debate with their customers who aren’t happy.” He says that the evidence is to the contrary. In his experience, most people who participate online want to be positive
Crowdsourcing offers real benefits to businesses large and small and the evidence is that more and more people are experimenting with it. We expect that this kind of engagement will become even more mainstream in the coming year and even that pretty soon customers will expect to engage in this way.
We’ve written on this blog before about research and reports that show the influence that social media and online reviews can have on purchase decisions, including a post showing how 25m American adults make decisions based on social media. A report out last week supports this and shows the role of online reviews, in particular, to be greater than we might previously have thought.
The report, by Rubicon Consulting, shows that for American adults, online reviews are second only to word-of-mouth in terms of influence on purchase decisions. What is common about both of these, of course, is that they both rely on consumer-to-consumer recommendations. We’ve seen this before with research from Forrester, McKinsey and with our own experience at FreshNetworks of talking to clients. A consumer-to-consumer message is stronger than a brand-to-consumer one. And now it appears that getting this message through online reviews is second only to word-of-mouth.
The detail of the report shows some more exciting and interesting observations. Whilst the influence of online is large and growing, it is worth exploring the differences that emerge, especially those which show where we are influenced weakly by online information.
The areas where the web has greatest influence includes purchase of consumer electronics, vacation planning, choosing a movie to watch, buying a car, looking for a job or choosing a restaurant. In each of these, more than one in three consumers say they are heavily influenced by online information when making a purchase decision. The lowest areas are choosing a mechanic or solicitor and, interestingly given the strong use of social media by the Obama campaign, choosing who to vote for.
It may be that some decisions lend themselves more naturally to being influenced by online content, but I suspect it is just that the tools, sites and content do not yet exist to make influence meaningful in some areas. There are more places where you can share thoughts (and learn others’ thoughts) on consumer electronics and vacation ideas than there are for solicitors. Of course, things change rapidly - DeJuridica, a French site where you can review your solicitor, among other professions, is launching soon.
What we are seeing here is the democratisation of information online - more users reading more reviews from more people on more products. It is only when a tool becomes used by a broad base of people that it truly becomes useful and can have an impact on the way we live. That’s what is happening now with social media and online communities. And why they matter quite so much.
I was speaking on the Future of the Industry at the ESOMARPanel Research 08 conference in Dublin last week, an international market research conference. Over the two days of the conference, a lot was said about social media and online communities (see previous posts on lessons we can learn from the team at MomConnection). Speaking in the last session on the last day is always more difficult, and I wanted to leave the people at the conference with something to think about. So I spoke about two issues we have seen develop FreshNetworks.
Firstly I spoke about how online communities, and more importantly the ability for consumers and brands to talk directly with each other using social media, is changing the client-agency relationship in research and other marketing services. Whereas previous agencies played the role of standing between the consumer and the brand (or the client and respondent in market research terms), now their role is more to facilitate these two groups interacting directly with each other. This sounds like an easy change but really it isn’t. It shifts both the role of the agency (from intermediary and translator to facilitator and advisor) and it throws up it’s own problems. In the research industry, for example, the agency, standing between brand and consumer, has an important role to play ensuring that any research is conducted in an honest manner, designed and carried out to make sure that the results are meaningful and that business decisions can be made on them. With the role of the agency and client changing, there is a need to change processes and techniques. The first step is to recognise that the role has changed.
Secondly, and building on this I showed that whilst we’ve had online research communities for some time, to date they have typically been used as new ways of doing old things. Today, with the significant shift-change in the use of social media for customer engagement, online communities can now do completely new things. We are seeing more and more organisations building online communities as a way to engage with clients. Indeed just a few weeks ago a report from Gartner predicted such communities at more than 60% of large US firms by 2010. The challenge for the market research industry here is that, with so many communities being built and so many firms building them, they may lose the initiative. Communities are a brilliant source of insight (either planned and managed or insight through UGC). Organisations will begin to rely more and more on these tools as sources of insight and research, whether or not the community has been built or managed by a research firm.
So it was a bit of a “if we don’t do it somebody else will” speech. A sentiment that really does apply to the market research industry today.
I wrote a paper for ESOMAR on this and if you’re really keen it’s available here (or just ask me if you want to chat about it).
In Nara, in Japan, there is the largest wooden building in the world. And in that building is one of the largest Buddhas. Every year there is a festival to this Buddha and I was at the festival a couple of weeks ago. Central to the festivities is an afternoon long performance of a traditional Noh play. As I stood in the sun watching the play I realised I had no idea at all what was happening. Not because I don’t speak Japanese. I’ve watched plays in foreign languages before and at least manage to follow the gist of what is happening. With this Noh play I couldn’t even do that and so it was just a group of Japanese people in costumes singing and playing musical instruments. Great as this was to see, I did feel I didn’t get as much out of the performance as I might have done.
Reflecting on this, it’s clear what the problem was. Whilst any play I might see in Europe or indeed anywhere with a Western tradition of theatre would have pretty much the same broad plot. There would be goodies and baddies (the latter would seem to have the upper hand at the start but eventually lose). There would be a nouement and denouement. Some traditional roles would play out. In Noh plays there is a very different tradition and a very different set of habits and customs. Which makes it difficult for me to engage with.
This need for tradition, habits and customs is one that is common to all social interactions, and something that we see playing out online, especially in social media and online communities. If you follow these traditions and habits then you will make it easier for people to engage with your site. And if you introduce a whole new set, then people will find it difficult to engage.
We saw this recently with Facebook when they changed their user interface. A large number of users complained, setting up groups (like the almost 2.5 million people in this group) and posting their dislike of the new format. What they were complaining about wasn’t the look and feel, at least not from an aesthetic perspective. They were complaining that the habits and customs they had got used to had gone; they needed to learn new things and do things in different ways. Facebook responded to this by starting to reintroduce some old formats and supporting old habits. Not a complete reversal, but a change in design that reflected the need to respect the traditions and customs people were used to.
We see this in the online communities that we build for customers at FreshNetworks. If people are used to a way of doing things then we need to respect that. If they expect to be able to message their friends in the community then we let them. If they are used to being able to vote and comment on photos then we should let them do this. And if they use one set of language and terminology then we should mirror that.
It’s why we spend time when we set-up a community to understand the target audience, and both their relationship with the brand and their use of websites and social media. The more we know about their habits and customs online, the better we can design the site for them and the more likely they will engage with the community.
Traditions and habits matter and identifying and respecting the ones that matter most will help to ease engagement.
I’m at the ESOMARPanel Research 08 conference in Dublin for a couple of days. Tomorrow I present a paper on the future of the market research industry, talking about how online research communities are changing the agency-client relationship. Today I get to listen to the other speakers, talking on subjects from data quality to online communities. The latter sessions were obviously of most interest to me.
The site has been running for five years now and in part of their presentation they highlighted ten insights they have learnt from running the site:
Right size, don’t supersize - the team from MomConnection keep a community of 5,000 people and say that anything larger would be of less use from a research perspective
Protect the sponsor and the brand - when you’re building a branded online community it is important that you recognise and act as a brand ambassador
It never gets cheaper - the team claim that the need to constantly innovate and develop the site mean that there will always be a cost involved with running and improving the community
It never takes less time - the MomConnection team do a series of monthly, quarterly, semi-annually and annually activities to maintain the community and membership
Encourage shared ownership - the team talked about shared ownership between the client and agency
Respect your panel [sic] - they engage in direct contact with members and have named individuals dealing with them
A real community isn’t in it for the money - if you build a real community you won’t need to incentivise them to take part
Make the most on evolving needs to innovate - listen to the ideas of community members and the client on how to evolve the community
Stay focused on the client’s goals - make sure they get out what they want but also that they make the most of the community
Speed breeds success - the online research community format is built for quick turnaround research and so build tools and processes to support this
What are your thoughts on these as the ten main lessons for online research communities. I’m looking forward to debating them with the rest of the team at FreshNetworks when I’m back in the office, but reflecting on them today I think there are areas I would expand on. The ‘never gets cheaper’ and ‘never takes less time’ may be true but from our experience, although there may always be a cost and time invovled in managing the community this does not have to increase proportionately to the increased size of the community. If you get things right at the start you can scale the community with a relatively decreasing cost per member. I also agree with the need to encourage shared ownership, but we would rephrase this, saying the community owns the community and so the client needs to be part of that to make the most of it.
Anyway - more on some of these things in my presentation tomorrow. I’ll right about it afterwards.
There’s a debate to be had in the world of 
It seems to be a week for research. A couple of days ago, we reported on research from Rubicon Consulting showing that 
We’ve written a lot in the past about 
We’ve written on this blog before about research and reports that show the influence that 
I was speaking on the Future of the Industry at the 
In 
I’m at the 