Still Fishing Where the Fish Are?

[#1]

Michael Wu, Ph.D. is Lithium's Principal Scientist of Analytics, digging into the complex dynamics of social interaction and group behavior in online communities and social networks.

Michael was voted a 2010 Influential Leader by CRM Magazine for his work on predictive social analytics and its application to Social CRM.He's a regular blogger on the Lithosphere's Building Community blog and previously wrote in the Analytic Science blog. You can follow him on Twitter at mich8elwu.

---

Last month I wrote a couple of posts exploring the strengths and weaknesses of Facebook (FB) from a relationship perspective. Although these posts were meant to highlight some of the proper and improper use of FB for businesses, the discussions are still rather scientific. They use many network concepts derived from social network analysis, and many sociology and social anthropology ideas. Today, I will try hard to paint a coherent picture with the scientific findings from these two posts for brands and businesses.

1. The Social Dynamics of Facebook Fan Pages
2. The True Marketing Power of Facebook: Sociology Perspective

Facebook Fan Pages are Wide but Shallow

Last time, we learned that the small world and navigability properties of our offline social network make it a very efficient medium for spreading relevant information. The structural properties of our real social network enabled a very nice balance between:

          a). Keeping the information content relevant

          b). Reaching a wide audience rapidly

Because FB tries to mimic our real-world social network, it inherits these properties from our social network. For example, the average path length for a random pair of persons on FB is 5.73. This is very close to the well known claim from six degrees of separation, which is a result initially derived from our offline social network. The balance between relevance and reach is why FB fan page is so great for driving awareness and creating interests for brands and businesses.

We can compare this to another popular social media that also enables social networking: Twitter. The average path length between a random pair of persons on Twitter is 4.12. This means Twitter’s follow network is a smaller world, and people on Twitter are more closely connected with only four degrees of separation. A positive side effect of this network structure is that information spreads faster. But the negative side effect is that the information is also very noisy and less relevant. Many marketers focus on total reach and the velocity of reach, because they are easier to measure, but relevance is crucial. Just remember this: regardless of how fast the information spreads, if it is irrelevant, people will treat it as spam and ignore it.

Clearly fan pages on FB are very effective at driving the unaware consumer into the shallow upper layers of the purchase funnel (Figure 1). The interesting question is can the fan pages push these consumers deeper into the funnel where it is more profitable for businesses. There are several factors suggesting that FB won’t be able to push most of their fan very deep down the purchase funnel.

As we learned from The Social Dynamics of Facebook Fan Pages, users on FB are usually not there to engage with brands. People get on FB to interact with their friends (i.e. the strong ties that already have pre-established relationships). Due to the attention economy and the conflicting social spheres on FB, fan pages are usually not able engage their fans sufficiently to drive action.

I must clarify that this doesn’t mean that fan pages do not lead to any purchase action. The purchase funnel is called a funnel because it has natural dynamics that channels consumer through their decision journey (i.e. AIDA – awareness, interest, desire, and action). After the initial draw into the early stages of the funnel, fan pages simply don’t contribute much more to the funnel’s natural conversion dynamics. However, by virtue of increasing the number of aware consumers, brands will benefit from a proportionally greater number of conversions at the other end of the funnel. In other words, simply by fishing where the fish are (with a fan page), brands will end up with more fish.

Communities are Narrower but Deeper

From our cyber anthropology discussion, we learned that there are really only two main categories of social media:

1. Social Networks
2. Online communities

Moreover, these two social structures serve complementary roles in the social history of mankind. (If you are curious about this topic, please see the mini-series on cyber anthropology.) Interestingly, online community also has a complementary role in consumer’s decision journey as we shall see below.

Like a fan page, a brand-sponsored community consists of interest focused discussions around a brand and its products/services. In fact, a fan page is really a community on the FB platform, because it has all of the characteristic traits of a community as we discussed in Community vs. Social Network. However, brand-sponsored communities are never on FB, so they do not suffer from the problems that arise from the attention economy and the conflicting social spheres. Therefore, these communities can engage and interact with their members at a much deeper level to build meaningful relationships with customers.

One of the greatest differences between a community and a fan page is that a healthy and vibrant community cultivates superusers (a.k.a. advocates, evangelists, influencers, and many other names). Superusers are typically a tiny fraction (less than 1%) of the community population, but they can generate up to 50% of the community content (see The 90-9-1 Rule in Reality). So basically these superusers keep the community running by answering questions, responding to new comments, and keeping the conversation alive. And they are doing all this for free. Because these superusers are usually not affiliated with the sponsoring brand, communities are often viewed as voices of the peer, where as fan pages are still viewed as voices of the brand.

Because a well-run community can leverage strength from an unpaid army of dedicated superusers, it is able to scale with the social web. That means the community will be able to address individual needs at the scale of the social web. For example, people can ask very specific questions and get tailor answers that address their specific needs. Notice that everyone can get this level of customized response. This makes the community extremely relevant, down to the single user level. Moreover, conversations in the community will often evolve a brand’s messaging and explore edge use-cases for its products/services. This often makes the products/services even more relevant and adoptable.

By providing objective answers from other customers to critical issues whenever a user wants it 24/7, communities can go much deeper down the purchase funnel and drive actions. However, communities are doing this at the expense of having a narrower reach (Figure 2). From another perspective, the narrower reach of communities is what allowed them to reach deeper into the funnel via high level of relevance and specificity. Furthermore, communities can continue to empower their customers after the purchase action with proper incentives and rewards. This is what cultivates superusers who made everything we discussed possible. So, contrary to a fan page strategy, which is to fish where the fish are, a community strategy involves farming our own fishers, so they can catch fish for us.

Conclusion

So what have we learned today? I have applied the sociological principles we learned to show that fan pages on FB and communities actually have complementary roles for businesses. They act on different stages of the customer decision journey.

1. A FB fan page has a very wide reach, and it is great for driving awareness and creating interest. It mainly works on the shallow upper layers of the purchase funnel, which correspond to the early stages of a consumer’s decision journey.
2. A community has a narrower reach, but it can drive more actions and conversions. It can reach much deeper into the bottom layers of purchase funnel, so it can affect both the early and later stages of a consumer’s decision journey.
3. If you are only fishing where the fish are (e.g. by having a FB fan page), maybe you should start thinking about farming your own fishers (via a community). That is another way to get a steady stream of fish in the long run.

OK, this concludes our sociological discussions about relationships and its applications to business. Next time, I like to start something new. I haven't decided what I'm going to write about yet, so if you like to explore any topic scientifically, please let me know. In the mean time, please let me know if you have any questions or comments. As usual, discussions of any kind are always welcome. See you next time.

[#2]

Great article as always. Thanks for sharing deep understanding and help market to understand important differences between online communities and fan pages.

[#3]

Hello Hans,

You're welcome. Thank you for stopping by. There are actually a lot of interesting sociological complementarity between social networks and community. This post just highlights those that are important for businesses and enterprises.

Thank for coming back. See you next time.

[#4]

Michael,

While the whole article is very interesting I'd like to ask about one small aspect.

You touch upon a very interesting point about the relative degrees of separation between any two individuals on Facebook and Twitter.  With Facebook closely mimicking real life social networks and Twitter representing a smaller world.  You say, the smaller wold "less separation" on Twitter causes information to spread more rapidly, but creates more noise.  Could you talk more about this?

Are you saying that the noise on Facebook dissipates faster than it does on Twitter or that less noise in the first place gets created.  Also, since no one follows everyone on either site, isn't noise a factor of the users own network and how they filter.  After all noise is not an issue if you are selectively tuning for certain frequencies.

[#5]

Hello Andrew,

Thank you for stopping by and commenting.

Noise (or irrelevant content) on Facebook are better contained within the the relatively smaller and more tightly niched social network of each individuals. So indirectly, you can think that noise dissipates faster. But what really happen is that your friends simply won't re-share your content if they are irrelevant.

Twitter on the other hand has lot more irrelevant content because the follow/follower relationship do not correspond well to true relationships in the real world. So the individual's network is not as niched and clustered. So even though information still diffuse through the network, because the network structure is not as tightly clustered, information diffuse further beyond and more randomly across the network. This purely an effect of the precise structure of the network.

It is correct that you can selectively tune for a certain frequency. In that case, you have to do the work, b/c you are doing the active filtering. Tuning in is the same as searching for what you want and filtering. You are choosing who to listen to and selectively ignoring all the rest. So FB vs Twitter is really like the difference between you going to google to look for what you want and have relevant information deliver to you without anywork from you. In other words, on FB, info delivered to you is pretty relevent, so you don't have to do too much work to get the info or to find extra info. On twitter, you have to do the filtering yourself, otherwise, the info become pretty useless due to too much noise (irrelevant content).

Alright. I hope this address your question. Thank you again for commenting on my blog. Hope to see you next time.

[#6]

Thanks for aswering my question, Michael; I find the topic fascination.

I would love to discuss it further because I find exactly the opposite to be true for me and my particular Facebook and Twitter use cases and practices.  Perhaps it is my perception or how "relevance" is defined vs. what is percieved as "valuable" content.

Because I use a Twitter client that I set up over time, and continually adjust, to filter content, and because of how I selected my network and topics to follow, I get a tremendous amount of relevant and valuable content out of Twitter In facebook since, where "friend" connections are reciprical, my network is constricted and less content is available to me, also I do not know of a way to easily filter the content coming from my Facebook network and even though those I follow are mostly real life friends, that does not neccessarily mean that the content they post is more relevant than that posted by online friends. I have little control and must rely on Facebooks algorithm to accurately show me the most valuable information which it has yet to have proven itself effective.  I miss a lot and must go to most recent news and wade through the noise.  I suspect the experience to be exactly the opposite for others who use these networks differently than i with different goals, objectives and use patterns.

Perhaps, as I said in the beginning this is just my personal perception, but I still don't think that network theory in itself without consideration of use cases, filtering, whether a user is a node or outlier (probably the wrong term), and multiple other factors can reveal the true noise level that any particular use will experience on either platform.

Thanks for the discussion,

Andrew

[#7]

Hello Andrew,

Thank you for coming back and furthering the conversation.

I think you have just proved my point. You have to set up filter and adjust them over time in order to make twitter content relevant. If you don’t do these filtering and adjustment over time, then twitter content tend to be much noisier. You find Twitter more relevant because you are comparing a hand-tuned filtered result of Twitter stream against a raw, un-filtered Facebook streams. This is not a fair comparison in terms of which network can deliver more relevant content.

I can also go to Google Alert and copy/paste all the search terms, key words, names, basically everything you follow on twitter and get the same relevant content as Twitter. Moreover, you can get even better/more complete coverage than Twitter, because Google has the WWW + Google RealTime, which includes the twitter feeds. But again that is not a fair comparison either.

Back to the fundamentals of the network structure. FB is much more relevant in terms of the content they delivered from a network structural perspective. But that does meant you can’t do some work to set up filters and select the relevant stuff out of the twitter stream. Most people ignore a large portion of the contents on their twitter stream simply by not being in front of twitter 24 hours a day. By setting up a filter, or monitoring a subset of their follower through twitter lists, they are filtering out even more content. If we can calculate the number of tweets that a user actually see, compare to number of raw tweets that reach a user’s public time line, that ratio is very tiny.

If all we do is use filter, the what people follow doesn’t matter anymore. You can follow a million user and have a million following you back. But if you only look at a list of 10 people, then effectively, you are only following 10. If you use search the bring up information, you can do that even when have 0 followers.

Anyway, I hope I am making sense to you, and hope I clarify more than confuse. But I’m happy to discuss more if it is still unclear to you.

Thanks again for asking the question. I think this discussion will benefit many others.

[#8]

Last week Andrew Mueller and I continued this discussion on twitter, which we thought would be nice to share with the rest of the readers. I captured the conversation by TwUniverse, edit it for readability, and posted below.

@andrewmueller: I understand your points about total noise in networks, I hope you understand mine about the amount of noise actually experienced

@mich8elwu: yes, I totally understand your point. Just want to compare apples to apples.

The noise experience is not a property of the network itself, it's a property of the tools build on top of the network.

So if we want to compare noise experience, it shouldn't talk about network, it should be about the tools on top :-)

@andrewmueller: The network itself dictates the tools that can be created and used the are not independent.

@mich8elwu: But network is not the only thing that determines tools though. Theoretically, we can build all the tools for twitter on FB

But FB don't open their data, making it difficult. But the network is still there. so does property/strucuture of the network

@andrewmueller: noise exists in use case only, it is not heard anywhere else therefore a non factor other than how it affects network perfomance

@mich8elwu: ah... but it will affect network performance though. even though we build tools to hide problem in network underneath

but agree that to an end user who don't care about the underlying layers, then it doesn't matter much

but network noise+problem tend to be poorly scaled, so eventually tools may still break b/c network breaks.

@andrewmueller: So I think that we agree, Twitter contains more noise than Facebook but doesn't necessarily mean the use experiences more noise

@mich8elwu: yes, that is true. but i like to make the distinction between the network itself vs the tool that ppl use...

it's like cell phone network vs cell phone. if network is noisy, we can make better phone 2 filter it, but network still...

if the network get's so noisy that we can't make phones to filter out the noise, then everything on top of network will fail

network is therefore somewhat a more fundamental object of interest

@andrewmueller: but we can and do make tools that filter the noise, the problem is that the noise sometimes makes the network itself fail

@mich8elwu: we can do better than make tools filter noise. We can impose rules on how ppl form these network, so they don't break easily

@andrewmueller: but that is a completely different issue that whether or not a user is exposed to more noise on one network than another.

When we look from a user experience perspective we cannot separate the network form the tools. They combine 2 deliver experience

@mich8elwu: yes. to the user, it is just another bad ux. they may not care where the problem arise.

@andrewmueller: interestingly, it is not only noise that is an issue, but sheer volume of the relevant and duplication.

@mich8elwu: precisely. if the network don't fix fundamental problem, then tools can only do so much.

@andrewmueller: ah, so the crux of the question is whether noise is more of a problem to each perspective network not which has more noise

@andrewmueller: It would certainly be interesting to post this conversation into your comments on the Fish post?

@mich8elwu: good idea. let me see if i can extract this conversation and post it.

Hope you find this conversation intereting.

Feel free to continue the conversation too.

[#9]

Hi Mike!

Thanks for this awesome finishing article! I weel a lack of good discussions about

Your picture of coherent power of using fan page and community through reference to funnel is pointed!

- I realize as you the situation with perception of fp as brand voice. It does block a lot of incentievs to use it in some another way. 

- Actually fp usage now does only be the cuurent style of usage. Realy we have enough strong ties within communities too, but our partners there aren't friends in offline.  FB is in the stage of accumulation relationship history, social history, that can make a base for extended relationships managment as it seems for me (support, sales and R&D). 

- FB will have chance to attract some volume of conversations and helping actions on the fp, I think. Chance... But it must be a long road - to close new social circles within. 

I am interesting in your opinions about some related topics: value attribution models, incremental WOM, networked CLV and another about measurement of social networks effects. But I know to have fun time with reading all posted on your blog! 

[#10]

Hello Andrei,

Thanks for the nice comment and glad you like the picture.

To your 2nd point: Communities are the natural place for relationships to develop. In fact, it is very often the case that people develop a social network within a community. There will be some strong ties and some weak ones. But the strong ones tend to remain and become part of our egocentric network even after we leave the community. If you are interested in how this dynamics work, please check out my post on cyber anthropology.

Thank you for letting me know of your interest. I will definitely put them in my to-do reasearch list. Thanks again for commenting. See you next time.

[#11]

Wonderful article it's always nice to see some true depth in a post. So many are opinions and such rather than true studies and while those kinds of articles are nice it's good see the science behind yours. Keep up the good work and I've be sure be mentioning and promoting your post.

@kstaxman (on twitter & amplify)

[#12]

Hello Frank,

Thank you for the very nice comment. And I'm very glad to see that you actually care about the science behind how social media works. That to a scientist is the best reward, because it is one of my goal to explain the phenomena on social media with social science principles.

Thank you in advance for being so generous as to mention and promote my blog. Very much appreciated. I will definitely keep up the good work.

Hope to see you around Lithosphere sometimes.