Saturday, 25 February 2012

Is virality really that easy?

Going viral: a critical review
In this article, I critique existing paradigms around viral marketing and pull three key messages from peer reviewed literature:
Key take away #1: If you want your product to go viral, it's gotta be a shit hot product.
Key takeaway #2: you've got to make your website visually appealing
Key takeaway #3: If you want to get Word Of Mouth recommendations, focus on building a tight knit community, create a reputation system and reward altruism

The tipping point
In the past month, I've heard several people say something to the effect of "Oh we're just going to use viral marketing. You know, tell our super-connected friends about it and before long we'll have 100k page views". They're smart people, so I figure they're not just spitting out nonsense here, they must have some reason for believing that virality of that magnitude is easy. When I asked them a few more questions, it turned out that many of them had Malcom Gladwell's book "The tipping point".

It's a fun book to read. Gladwell weaves a bunch of compelling stories of trends that have spread faster than the flu in a crowded train (eww). The Hush Puppy story is the classic example he bases his arguments on. Apparently Hush Puppies were the height of dagginess until some trendy New York hipsters started wearing them. Everyone wanted to be as avant garde as these hipsters, and started rushing to Hush Puppies stores in droves to join in the craze.

Meet the Influentials
Incredible right? If this story is correct, then all marketers need to do is find those hipsters - the super influential people their target customers respect - and convince them to wear their shoes/drink their water/sign their petitions and before long EVERYONE will be scrambling over each other to get their hands on your product/service/change making initiative.

Ed Keller and Jon Berry pick up on Gladwell's argument in their book, "The Influentials". They propose a breed of marketing superheroes: the Influentials - the people that everyone else admire and sycophantically imitate.

I've got this image in my head of marketing directors rubbing their hands in glee, thinking about how they can slash their marketing budgets by appealing directly to these Influentials instead of forking out mega bucks for a Superbowl commercial.

Are the influentials really that influential?
Duncan Watts, the guy who retested the '6 degrees of separation' theory in the modern era (it still applies), is pretty adamant that the Influential hypothesis is flawed. In his 6 degrees research, he showed that contrary to Milgram's original finding that most messages pass through 'hubs' (super connected people with disproportionate amount of influence in the community), most messages went through plain old ordinary folks with average amounts of influence.

Since then he has also done some nice empirical research into information diffusion through twitter. He found that the likelihood of a link going viral through twitter has very little to do with the influence of the person who started it. Instead, the best predictor of virality is how interesting (based on reviews from Mechanical Turk workers) the linked website is and how many people start off talking about it.

In other words, paying super influential hipsters to wear daggy shoes is very unlikely to create a trend. (A lot of people think Gladwell got the Hush Puppy story wrong - it probably had more to do with the parent company hiring a new marketing team). They've got to look and feel good first.

Key take away 1: If you want your product to go viral, it's gotta be a shit hot product.

Making ridesharing viral
The reason I'm interested in this topic is because I want to come up with a way to make ridesharing mainstream. Given that we have zero marketing budget and given that word of mouth is more effective than traditional advertising, virality is going to come from word of mouth promotion. So the question I'm asking myself and the team at CoCoRide is "how do we make our service interesting and valuable enough that people want to tell their friends about it?".

What makes people try a service the first time?
We'd better start with first things first. We can't get anyone to promote it if we don't have any users in the first place. So how do we get users? A fundamental rule for intangible services like ridesharing is that consumers regard them as high risk and they're unlikely to sign up if they don't have a referral from a friend. However, if
the user feels like immediately understand the service and that the organisation behind the service is credible, they may well lower their barriers and sign up on the spot without asking anyone else. For this reason, if you can make the website for your service appear credible, then you will get more signups.

What gives a website credibility?
B J Fogg has done a lot of research into website credibility. His team at Stanford interviewed 2600 people and asked them to assess the credibility of a bunch of health and finance websites. The findings of the study were intriguing because they fly in the face of how people should assess a website's credibility. Rather than looking at the factors that really matter, like whether the website had a sound privacy policy (you wouldn't want to sign up for a website only to find out later that the dodgey operator sold your contact details on to a telesales company) and who was sponsoring the site, people primarily judged the websites' credibility based on how good it looked.

Key takeaway #2: you've got to make your website visually appealing.
We've got a lot of work to do on this front at CoCoRide. Our website looks amateurish right now and that probably explains why we've got a pretty poor signup rate.

What makes people promote a service?
Let's assume that CoCoRide finds an amazing designer and we transform the website so that it wows people when they first see it and they snap their mouse button in half in their rush to sign up for our service. What next? How do we get them to tell their friends about it?

Cheung et al. attempt to answer this question in their paper: "What drives consumers to spread electronic word of mouth in online consumer-opinion platforms". They interviewed 203 people who were part of the restaurant review site,, and found that the top three reasons for recommending a restaurant were:

1. Sense of belonging
The members who answered "Extremely Agree" to questions like "I am very attached to community. (Extremely disagree/Extremely agree)"
were more likely to leave positive reviews about restaurants. This finding ties in nicely with Seth Godin's Tribe Marketing approach, where he argues that building a tribe of consumer advocates is the best way to market your service. The implications of this for CoCoRide is that we need to bring our users together in some way so they can feel part of the community as a whole. Perhaps some exclusive CoCoRide events would make sense. It would also be good to provide an easy way to recognise other CoCoRide users. Perhaps CoCoRide bumper stickers would be in order.

2. Reputation
This one's interesting. Apparently OpenRice members who wanted to boost their professional credibility (perhaps they were restaurant reviewers or chefs) were more likely to leave positive reviews because it would allow them to be perceived as experts.

CoCoRide could tap into this by publicly demonstrating each user's expertise. This would reaffirm our gamification strategy, which involves providing badges to users once they have shared sufficient rides.

Reputation could also tie into the phenomenon of 'conspicuous conservation', in which consumers buy environmentally friendly goods and services so they can beat the Joneses. Those bumper stickers would tie in nicely here. Perhaps something fun like "Traffic Jam Unblocker".

3. Altruism
Apparently it's true: helping others does make you feel good. OpenRice members who posted positive reviews often said they were doing it because they liked helping other people avoid bad restaurant experiences.

CoCoRide could tap into altruism by labelling the act of picking a stranger up as helping them out. Drivers could have a badge on their profile: "John has helped 67 people get to their destination".

Key takeaway #3: If you want to get Word Of Mouth recommendations, focus on building a tight knit community, create a reputation system and reward altruism

What are your thoughts on generating virality? What have I missed out?

No comments: