Archive for October, 2009

Measuring your SlideShare success

Wednesday, October 21st, 2009

I know that a lot of my leads for speaking engagements have come through SlideShare. People who invite me tell me that they actually found my ideas through SlideShare and were convinced I would be a good speaker or sparring partner for their case. Until now I haven’t really thought how to analyze what works and what doesn’t. I just know what presentations are viewed most often.

Anyway, my experience is that by sharing your presentations you will get more than you would get otherwise. More leads and valuable feedback. The downside is that you become conscious that giving the same presentation twice doesn’t help your online distribution at all. You have to keep on changing and that’s great for everybody.

I took the views, downloads and favorites stats of all of my presentations and put them on a spreadsheet. This was easily done by looking at the document stats at LeadShare (business extension on SlideShare to encourage leads).

Then I looked at the following things:

  • The relative percentage of downloads compared to views. The assumption is that people are more likely to download the presentation if they find it useful.
  • The relative percentage of views + downloads for a single presentation compared to all views + downloads for all presentations. This gives you a good overview what presentations are actually leading the way (or have got most exposure).
  • The relative percentage of favorites to views + downloads for a single presentation. The assumption is that people will favorite a presentation because they love it or want to store it for later reference.

Of course it’s hard to get an objective view here, because:

  • Certain good stuff is picked up by more popular bloggers and some perhaps even better stuff sometimes never gets picked up at all.
  • A great enhancer for traffic is also the moment when your presentation gets featured by SlideShare. This has happened to several of my presentations.
  • In the other hand, time is here an issue: my presentations are published in around two month intervals since October 2006, not all of the presentations have been available for the same time.

Therefore, the view, download and favorite counts are not good enough indicators of how you are doing, but rather the relative percentages I’ve been calculating. Below you can see my current situation on SlideShare:

The most popular presentation by far is my Web 2.0 Business Models presentation with 40.35% of all traffic. This doesn’t mean it’s the best presentation. If you look at some of the relative percentages, you can see what presentations likely generate most value to their viewers.

Most downloaded presentations compared to views:

18.39% – Web 2.0 Business Models
13.07% – Vision of the future: Organization 2.0
11.43% – Culture Matters – The cultural requirements for Web 2.0 powered innovation, networking, and collaboration
10.69% – Innovation and Microinformation
09.48% – Age of Real-Time: Future Trends in a Digital World

Most favorite presentations compared to views+downloads (I have highlighted the ones that are also in the most downloaded chart):

2.34% – Collaborative Edge: Real-Time Social Technologies in Organizations
1.69% – In the age of real-time: The complex, social, and serendipitous learning offered via the Web
1.40% – Age of Real-Time: Future Trends in a Digital World
0.92% – Vision of the future: Organization 2.0

0.91% – Using Social Technologies to Run Better Events

How would you improve these stats?

Fractal learning

Sunday, October 11th, 2009

One day I asked myself the question, what would learning look like if it could be visualized?

322px Mandel zoom 00 mandelbrot set Fractal learning

A fractal. Latin fractus, meaning fractured. It is recursive by definition.

What comes to my mind is the Mandelbrot set. In 1975, Benoît Mandelbrot first coined the term fractal. Mandelbrot emphasized the use of fractals as realistic and useful models of many “rough” phenomena in the real world. In The Fractal Geometry of Nature (1982) he writes:

Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.

If something is rough, that’s learning. As you approach a new topic, you start from a fuzzy idea of what it could be. As it comes into focus, new details expose themselves on the fringes, enabling you to discover even more interesting perspectives you were not aware beforehand.

Fractals are seen in many parts of nature. Even fractal cosmology exists as an area of study. In a New Scientist article (2007) Labini & Pietronero asked the question, “Is the universe a fractal?“. Their study of nearly a million galaxies suggests that the matter in the universe is arranged in a fractal pattern up to a scale of about 100 million light years.

The Second Law of Thermodynamics states that the total entropy in the universe increases over time, as change happens. In layman terms that would be analogous to a room getting messed up over time as people live in it. In thermodynamics, entropy is a measure of the amount of energy in the system that is no longer available. As entropy increases in the universe, at the same time incredibly intricate and detailed order emerges from the details. Think of the human brain on planet earth for example.

250px Fibonacci spiral 34.svg Fractal learning

Fibonacci spirals also depict the fractal pattern of beauty in nature. Golden ratio is a very well known principle in mathematics and art, first originating in the Liber Abaci (Book of Calculation) in the 13th century. Good examples of forms with Fibonacci spirals include the spirals of shells, various flowerings, the branching of trees and arrangement of leaves on a stem.

The internet looks like a fractal.

So what do fractals have to do with learning?

When considering learning, we are pattern recognizers. Just like fractals, our neural networks evolve over time and extend outside of us. As our environment changes, so do we.  As we process information, in addition to entropy, new patterns emerge. By increasing the ammount of information, you increase the possibility of new patterns to be recognized by people.

In the digital world, entropy is information overload and order is the pattern that emerges from the interconnection of such information.

Knowledge is like a hologram. In holograms, even smaller pieces of it include the picture of the whole object. Knowledge is like a hologram. The experience changes as your point of view towards the object changes. The knowledge is not in a single image, but distributed on a network.

This is pattern recognition. And it’s the culmination of fractal learning. It’s a Mandelbrot set that zooms into the details indefinitely. Universe is fractal by nature. So is learning fractal by nature. It’s rough, it’s self-similar, it’s recursive and increasing the likelihood for serendipity is key for building higher structures.

Here is a recent Finnish presentation recording of my talk on the subject at a conference (Verkkoja kokemassa):

Warning: video ID not specified!

Here are my slides from the Distance Education & Teaching conference in Madison, USA (still waiting for the presentation recording to be published):

Real-time web and management cybernetics

Sunday, October 4th, 2009

On 1st of October I gave a presentation at MindTrek entitled “Collaborative Edge: Real-time Social Technologies in the Enterprise” at the “Social Media: Now What?!” track and later on 7th of October I spoke briefly about it also at the 5th World Conference on Mass-Customization.

My presentation is built around the ideas of Stafford Beer, who was the founding father of management cybernetics. His ideas are now more timely than ever, because of the advent of the real-time web. Stafford along with his team built the first real-time computer controlled planned economy at the government of Chile in the beginning of 1970′s. I’m very interested in this because I was part of a team that created Real-Time Economy Community.

Stafford Beer’s project was called Cybersyn. It aimed to create an electronic nervous system for the Chilean economy. As progressive as they were, they included machine learning with a Bayesian filtering (cool in email spam prevention in the early 2000) and social features by letting every citizen and factory worker to influence the decision making. He also included some social innovations too, like having a diverse cross-disciplinary team (rather than a group of generalists) working in a futuristic Opsroom: the ultimate combination of man and the machine.

250px Cybersyn control room Real time web and management cybernetics

Cybersyn Opsroom inspired by Tulip Chair design by Eero Saarinen from Finland.

Recently I’ve been thinking about how to implement the real-time web in the enterprise. If there is one guy who really knows how to do it, he is definitely Stafford Beer with his Viable System Model (VSM). The aim of such a system was to remain viable to its users by involving the ability to adapt to changing conditions. This requires real-time data to be generated, reflected and interpreted by every employee (and customer) to lesser or greater extent.

This is very close to what I’ve said in the past that learning is not a separate process to be managed in organizations through training, but rather an inseparable part of all such activity that seeks to avoid stagnation and remain useful over time.

His ideas were more bottom-up than top-down: due to limitations of single or small groups of individuals to comprehend everything what is going on (=top management), one needs communication and conversation with employees, partners and customers – The very ideas that concepts like crowdsourcing or open innovation aim to address.

Stafford’s contribution was also to emphasize the importance of increasing the amount of variety in highly fluctuating systems, where you cannot predict the possible states of the system beforehand. This is exactly what companies like Apple do: by not knowing what applications to run on the phone, keep the number of features (apps) to the minimum and let users innovate and personalize through an App Store. In comparison, Nokia thinks they know their users and load the phones with apps that in general are underused by typical users. The same logic goes with most user-friendly web services (e.g. anything that comes from 37Signals): if you do not know what features your users need, release a limited version, open up the APIs and listen to your customers.

This is exactly how you achieve collaborative edge to provide best services to your customers: in case of doubt, tear down the firewalls and listen. Turn your organization into a complex adaptive system.

On September 11, 1973 (notice the date), Stafford’s dreams came to an end as Salvador Allende’s government was overthrown in a military coup with the support of the United States government. Along with Allende, the project went into grave. How unfortunate, how typical.

My question is, why haven’t we done it yet?

See my presentation here:

Browse the slides:

See also Stafford Beer’s lecture about Cybersyn (vintage, 1974). Memorable quote about the opsroom chairs:

No paper – there is an ashtray. There is room for a drink and there is a place for a creative session.