More on networks vs hierarchies


Creating a hierarchy based on total scores and overall ‘influence’ risks lowering the value of a social network because total contributions do not mean every individual contribution is a good one

An example to follow on from a recent post: Social networks do not need a hierarchy.

A lot of social networking tools are focusing on the use of scores, badges and ‘gamification’ to encourage participation and highlight the key players, the ‘influencers’. Personally, I’m not a fan. I’m all for awards representing a significant achievement. But becoming mayor of the local train station simply by ‘checking in’ more times than anyone else is not on my list of priorities. However I have to accept the evidence. Simple games and badges work in certain scenarios – they increase participation compared to similar systems without them. But as Steve Jobs once said:

Incentive structures work so be careful what you incentivise people to do. Because it can create all sorts of unintended consequences.

The following picture shows the points and status for two people who have responded to a question on a Microsoft technology forum. At first blush, who is likely to be the authoritative source?


Yes, the one on the left is me. Or rather, my SharePoint clone. I haven’t participated regularly on a tech forum since the 90s. If I had taken the screenshot a few minutes sooner, I had a far more impressive null point (say it in a fake French accent and think of Eurovision). It seems I’ve been awarded 5 points for answering a question. On the right is another Microsoft partner who is also an MVP – that’s a Microsoft Most Valued Professional, an award given for contributions within the Microsoft community. And he has a tub-thumping 5,645 points.

So who’s answer would you trust the most based on this information? It’s OK, I’m not offended. I wouldn’t pick me either on this basis. Naturally, that’s the point of this post…

The question being answered was to do with enterprise search. I answered the question in its entirety. 7 steps that could be easily followed. The other partner added a general comment and a link to his blog post related to the topic but not answering the specific question. It provided no extra information but will have given him some more points and link love for his web site to boot. And why wouldn’t he, that’s what people are being incentivised to do. (Side note: I’ve greyed out the identity because the person involved is very knowledgeable about SharePoint and his MVP status is well deserved.)

I have a strong technical background in enterprise search and SharePoint. But that was quite some time ago. Even if I’d been given a big badge for it at the time, it wouldn’t matter now. Because the search functionality has changed dramatically over the various version releases and plenty of others have since caught up with my knowledge and surpassed it. The only reason I decided to answer a question in a technical forum for the first time in over a decade is because the person’s question had turned up in a search result. I was checking some information, saw the question was nearly identical to what I was checking and that it had gone unanswered for over a week. So I decided to respond. Likelihood of answering another one in the near future? Not high. And the world will (hopefully) keep orbiting the sun.

All a social network needs is a mechanism for connecting people with knowledge to people with questions. In short, it just needs a damn good search engine. This is why Flickr is such a brilliant example. You don’t search through photographs based on the points awarded to people based on their contributions. Not unless you are a fan following a celebrity photographer. Search and tags help you find the type of picture you are looking for. We are insanely good at judging whether or not what we are looking at is what we need. If not, move on.

Will that stop us from using badges and points in social networks and communities? Of course not. People are naturally competitive (whether we admit it or not) and crave recognition. Some are more easily satiated than others through simple rewards and manipulation. And when the network is owned by an organisation, there is value in discovering who are the most frequent contributors. Microsoft’s MVP programme is well run and those rewarded deserve the credit given for their ongoing commitment to Microsoft technologies. They donate a lot of time to help others. But displaying high scores on individual questions risks lowering the value of a network by focusing on the hierarchy of badges rather than the content. The better solutions are those that allow thumbs-up/down on an individual item basis. It doesn’t matter if you’ve got a grand total of 405,421 ‘thumbs-up’. What might matter is that you’ve got 10 thumbs-up for the question I need answering compared to no more than 2 thumbs-up for everyone else who answered it.

The power and beauty of a social network is that each individual connection between two nodes can be as valuable as the next one. Introducing a hierarchy risks losing that equality and weakens the system.

Related blog posts

Flickr image: Badge collection by Drew McLellan. As with the original post, kindly shared and no badges or hierarchy required to discover it

Creating a problem by solving a problem


An interesting article was recently published by the Washington Post – If this was a pill, you’d do anything to get it – that shows how solving a problem so often throws up a new one

Medical research has taken amazing steps towards curing or immunising people against infectious diseases – acute illnesses responsible for high mortality rates. The results are shown in the image below:

Graph: death rate for infectious diseases

The outcome is people living longer. When we live longer, we are more likely to suffer from a chronic condition – an illness that develops slowly over time and often cannot be cured but can be managed through early detection and ongoing medication or lifestyle changes.

And so healthcare is being disrupted by its own success.

In the US (as reported in the article), one healthcare program adopted a new approach. They targeted individuals based on their medical history (had a chronic condition and had been hospitalised in the past year) and instead of waiting for potential patients to phone when they get sick, they arranged for a nurse to visit the patient’s home on a regular schedule (weekly or monthly depending on the severity of the condition).

The results: Reduced hospitalisations by 33% and reduced Medicare costs by 22%.

You should read the article for what happens next. Not quite a happy ending and the clue is in the title of the article. Perfoming house calls is assumed to be expensive and inefficient, despite the success of the program. If the same results had been achieved by giving the patients a pill…

It’s a great demonstration of the need to think differently when solving one problem creates another.


The value in gossip at work


I watched a fascinating TEDx talk this weekend and have embedded it below to share. If you’re not enthralled at the start, I encourage you to stick with it. It leads to the conclusion:

Celebrity gossip is the conversation that exposes who we are… a reflection of modern human behaviour and culture. In observing the changing nature of standards and morality, gossip is the play-by-play of our social evolution

We often see nicknames for forums and comments sections of web sites such as ‘the water cooler’ or ‘chatter box’. But we kid ourselves if we think they are true replacements for the real thing. And gossip can be such a powerful mechanism for collaborative working. Many people have been criticising Marissa Meyer over the past week for the announcement that all Yahoo employees must now work at offices and not from home apart from exceptional circumstances. Whilst I think that is too strict an approach and there are big benefits working from home in terms of individual productivity, I do feel that organisations often under-value the benefits of informal collaboration through hall-way meetings and unplanned disruptions that tend to occur in close proximity rather than over long or digital distances.

Here’s the video. Make a brew and set aside 20 minutes for a very different TED talk

[ba-youtubeflex videoid=”oFDWOXV6iEM”]