Network holes and how to plug them

Posted: July 31, 2013 | Author: | Filed under: Uncategorized | Tags: , , , | 1 Comment

Notes on the practice of innovation and technology commercialization.

“There’s a hole in the bucket, dear Liza, dear Liza…” A children’s song with obscure origins in 18th century Germany.

My last blog discussed strong and weak links in networks. We should now talk about how to connect networks and what is meant by a network ‘hole.’ 

Networks, like the bucket in the song, may suffer from holes. In buckets holes need to be plugged; in networks, holes are bridged. In social networks they are called ‘structural holes’ formally defined by Ronald S. Burt, a leading researcher in networks and structural holes at the University of Chicago, as ‘when two separate clusters possess non-redundant [that is, complementary] information, there is said to be a structural hole between them.’ This suggests that a broker or boundary spanner – such as an entrepreneur – who bridges the hole could gain competitive advantage by engaging in information arbitrage. Connecting networks of entrepreneurs with their own strong and weak links by bridging structural holes may provide access to valuable new ideas, alternative opinion and practice, early access to new opinion, and an ability to move ideas between groups where there is an advantage in doing so.

Image

Previously we noted evidence from Innocentive http://www.innocentive.com that as the number of unique scientific interests in the overall submitter population increased, the higher the probability that a scientific or engineering problem was successfully solved through Innocentive’s crowdsourcing method. Bridging structural holes between disconnected networks brought a diversity of potential scientific approaches to a problem and was a significant predictor of problem solving success.  Innocentive and ideas coming from other industries than the seekers http://www.innocentive.com/blog/2008/07/25/5-questions-with-dr-karim-lakhani/

However, let’s control our enthusiasm because in a moment we will see that there may also be a downside.

Brokerage is about coordinating people between whom it will be valuable to trust, but where there is an inherent degree of risk. An earlier blog [http://innovationrainforest.com/2013/03/24/solving-the-right-problem-part-1/] discussed the importance of understanding context; the effectiveness, or otherwise, of brokerage is also highly dependent on context.

Much more can be found in Professor Burt’s paper Structural Holes and Good Ideas
http://www.econ.upf.edu/docs/seminars/burt.pdf

From what was discussed in my last blog we learned that within a closed network, with many strong links, no behavior goes unnoticed. Problems may be identified early and at low cost so that non-performing people can be removed from the network. In the early 1990’s, while working in Russia on technology commercialization issues I was surprised that the organization for which I was consulting employed several people who were related to the boss. Nepotism? No, said the boss; it’s an incentive for them not to screw up and have it immediately known by their relatives and be the cause of family shame.

Mentoring for early stage companies in venture accelerators relies of having robust knowledge of the business to apply the fail-fast fail-often principle, but this will be the topic of a future blog.

A series of new studies by Dr. Yuval Kalish of Tel Aviv University suggests that, in some cases, networking can do more harm than good. “If you’re at the intersection of two previously unconnected niches of a network, you’re occupying what I call a ‘structural hole,'” says Dr. Kalish. Filling that space can lead to prestige, opportunities and power ― or it may have quite the opposite effect.

“While it’s been reported that people who occupy these structural holes become more successful, some structural holes may be ‘social potholes’ that can harm you and your business,” he warns. For example, someone who cut across formal hierarchical organizational boundaries within a company may be resented.

Finally, networks are useless unless “transactions” occur among the networked parties. Communication is a transaction – an exchange of knowledge – and is also context dependent. The six degrees of separation theory that everyone is six or fewer steps away, by way of introduction, from any other person in the world is well known, but not much practical use unless at each stage a request gets transmitted. I figure I’m two links from the leader of one the world’s most powerful countries, but I very much doubt if any advice I send him would be received, and the transaction would fail.

Once when studying two linked group of researchers from a research center and a corporation in two different countries, my colleagues noticed an occasion when communications (transactions) between both groups were acting effectively, but then stopped. It was a mystery. Why? Answer: ‘culture’ prevented further transactions. The culture of the country in which the research center was located was that it was not good to transmit bad news. So, when the R&D ran into problems – completely normal and to be expected – they decided that giving no news to their corporate partners was better than giving what was perceived as bad news was the better action. Of course, from the company viewpoint it was essential for the company to know if there were problems so that the two partners could work together to correct them and proceed.

The remainder of the children’s song describes an increasingly frustrating tale of attempts to plug the bucket’s hole. Network brokers balancing cultures, transaction effectiveness and costs, trust, and so forth, may have a similar experience.

Next time: Notes on the practice of innovation and technology commercialization blog will be on vacation during August, and will return in September. Blog topics for the rest of this year will include how to create effective support ecosystems for technology transfer and commercialization in developing countries, agile commercialization systems, slow systems, and re-usable learning objects for technology transfer and commercialization.

The ties that bind us: networks, strong links, weak links, and expanding our knowledge

Posted: June 30, 2013 | Author: | Filed under: Uncategorized | Tags: , , , | 1 Comment

Notes on the practice of innovation and technology commercialization.

“The ties that bind us to life are tougher than you imagine, or than anyone can who has not felt how roughly they may be pulled without breaking.” Anne Brontë, Agnes Grey, 1847.

Anne Brontë gives us a useful introduction to the concept of weak and strong links and the stability – resilience to external perturbations – of our professional networks. Think of a network’s ability to roll with the punches like the shock absorbers and springs on a car as it rides potholes in the road.

In fact it was not Anne, perhaps the least well know of the literary Brontë sisters, but Mark Granovetter, a Harvard graduate student in the 1960’s, who began much of the contemporary discussion of networks in his now classic study The Strength of Weak Ties http://sociology.stanford.edu/people/mgranovetter/documents/granstrengthweakties.pdf which investigated how job seekers found employment. To cut a long story short, he found that rather than through relying on their close friends most found their jobs through contacts they knew less well, such as friends of friends. The analogous situation is to consider how we find new ideas. If we only talk to the people we know well it’s likely that discussions will proceed along familiar lines and new thinking will not emerge. Granovetter noted in his follow up study in 1983, The Strength of Weak Ties: A Network Theory Revisited, http://www.soc.ucsb.edu/faculty/friedkin/Syllabi/Soc148/Granovetter%201983.pdf “that individuals with few weak ties [what we now call links] will be deprived of information from distant parts of the social system and will be confined to the provincial news and views of their close friends. This deprivation will … insulate them from the latest ideas.”

InnoCentive  http://www.innocentive.com/ is an online marketplace which connects organizations posing technical problems (seekers) with freelance problem solvers in a multitude of scientific and technical fields. Their regression results showed that the further the solvers rated the problem was from their own field of expertise the more likely they were to have solved the problem. For example an electrical engineer came up with a successful solution to a problem dealing with adhesives. Diversity of potential scientific approaches to a problem was a significant predictor of problem solving success in InnoCentive.

A national innovation fund in a rapidly developing nation was recently having problem making investments that would yield expected rates of return; the networks of the fund managers had many strong professional links but insufficient weak links to open them up to new ideas and different ways of thinking. But you might point out The Rainforest: The Secret to Building the Next Silicon Valley http://therainforestbook.com/ makes much of the need to reduce transaction costs by building trust. Surely, strong links imply a higher level of trust then weak ones? This discussion will have to wait for another time; a clue – it’s all about context. Context is the subject of my March blog Solving the Right Problem: Part 1. http://innovationrainforest.com/author/alistair2013/

The LinkedIn professional network is another example of the power of weak links. In addition to its main purpose of job seeking, it can be an effective network to expand our thinking and absorb new ideas from participation in its multitude of LinkedIn Groups. If you want a visual representation of your LinkedIn network, go to http://inmaps.linkedinlabs.com/network

We have been talking about weak links, so we had better define them. A weak link is defined as a link in a network if it’s addition or removal does not significantly change the functioning of the network. In our own professional networks weak links are formed and also disappear all the time. There are more technical definitions and, as might be expected by anyone who has read my previous blogs, these definitions are highly context dependent. Much more about weak links can be found in the fascinating 2006 book by Peter Csermely – Weak Links: Stabilizers of Complex Systems from Proteins to Social Networks.

Weak links as stabilize complex system networks. Strong links also contribute to network stability, but if a strong link is lost the networks will behave in a different ways. Both kinds of links are necessary for stable networks and stable societies. However, as we saw from the national innovation fund example, if the system becomes too stable it cannot change or develop and new ideas will not form.  

It is often recommended that to commercialize innovations in a complex global marketplace innovators must break free of regional constraints and ultimately become interconnected to today’s globally distributed markets, sources of investment, partnership prospects, and talent.  This certainly seems to make sense. So should we just go and create as many network links of all kinds as we can and just link to everyone? It may not be quite that simple.

Neil Johnson in his 2007 highly readable book Simply Complexity and a follow up, not so easy to read, research paper with colleagues, Dynamical Interplay Between Local Connectivity and Global Competition in a Networked Population http://arxiv.org/pdf/cond-mat/0401527.pdf, describes populations which might, for example, be competing for business. Their research found that for populations with modest resources adding small amounts of connectivity between members of the population increased the disparity between successful and unsuccessful people and reduced the mean success rate. In higher resource populations they found that the levels are interconnectivity increase the mean success rate and enable most (objects) to be successful. At higher levels of interconnectivity, the overall population became fairer (smaller disparity in success rates) but less efficient (small mean success rate) irrespective of the global resource level.

Or, as stated by Justin Marlowe at the University of Kansas in a related 2007 study of public debt management Network Stability and Organization Performance: Does Context Matter http://www2.ku.edu/~pmranet/conferences/AZU2007/ArizonaPapers/Marlowe_2007.pdf  “the consequences of ‘wiring up’ an additive population depends quite dramatically on the interplay between the local connectivity and the global resources.” This finding is consistent with a growing literature that suggests networks in general, and stable networks in particular, distribute resources in ways that may reinforce inequalities.

We are beginning to see that not only is it important to have the right mix of strong and weak links in a network and that also simply saying that we should link to everyone may not always be the right choice. Also, this mix of link types is certainly necessary, but what is really important is the ease with which information moves through networks – the subject of the next blog in this series.

I admit that all this is becoming confusing; although it’s clear that the networking story has still far to run and that we will continue to learn more about how networks support innovation. Let’s hope the lesson are worth the effort and not as Anne Brontë  said elsewhere in Agnes Grey –  “All true histories contain instruction; though, in some, the treasure may be hard to find, and when found, so trivial in quantity that the dry, shriveled kernel scarcely compensates for the trouble of cracking the nut.”

Next time: There’s a hole in the bucket, dear Liza, dear Liza – Network holes and how to plug them.