Idea diffusion and influence

I was prompted to think more about Duncan Watts’s ideas by Sarah Fraser’s comment on my post on influencer communities, and by her post on professor Watt’s theories.

In fact, I’ve just finished reading Watts’s Six Degrees, which was excellent and more accessible (I found) than Barabasi’s Linked. So Professor Watts is top of mind right now.

Here’s where I think the key difference between what Watts says and my practical experience. Watts talks about the role of influencers in the diffusion of ideas. As Seth says, use ‘sneezers’ with influence if you want to crack a market. Watts disagrees that you can predict what ideas diffuse, or even whether you can identify influencers that might make diffusion easier or more likely. It’s pretty much random, according to Watts.

I agree. If you’re trying to use influencers to spread ideas and concepts, then good luck but don’t bet the firm.

My own view is that influencers can be identified, and can assist greatly, in the decision-making process. That is, not whether an idea is spread or not, but whether an idea is adopted in the end. Idea diffusion is part of the process, but it’s just the start. A decision-making process is often a long and time-consuming activity. In the B2B world especially, a decision may take years to emerge. Idea diffusion is necessary, but not sufficient.

I explored this relationship between influence and the decision-making process in the book, and also posted on it (in summary form) here. Idea planting (as I called it) is right at the start of the process, but is relatively low down in the awareness of senior decision makers. Thus idea diffusers (connectors, sneezers, etc) may not be that influential in affecting the ultimate decision. There are a whole bunch of other influencers that intervene after ideas are sown.

Idea diffusion is also important in the process of deciding whether to do something. Do I adopt SOA? Do I need a Web2.0 strategy? Do I need a new car? But it plays less of a role in the subsequent decision of what to buy. Different influencers are in play at this more practical stage, like product reviewers or case studies or implementers.

The only problem I have with Professor Watts’s arguments is that when he doubts the role of influencers in any aspect of life, it doesn’t fit with real world experience and intuition. My guess is that we can all think of people who are influential in certain areas of life. Fitting this experience and intuition into a practical marketing approach is what Influencer Marketing is all about.

Hugh and friends discuss influence

Hugh McLeod does a regular podcast with Rabbi Pinny Gniwisch, Johnnie Moore and Mark Earls. This week they talked about influence.

Well, they did for about 10 minutes – for the next 30 it was mainly about success in marketing and creativity. Interesting, nevertheless.

Listening to these intelligent chaps solidified my view that influence is grossly being misunderstood and/or misrepresented. As Johnnie Moore said, there are two views. One is to think of “cool people” that tell the rest of us what to do. Find those influencers and success will follow. The other view is that life is more complex (duh) and success is often just down to luck, or random acts of traction (as Hugh puts it). (Echoes of The Wisdom of Crowds and Fooled by Randomness here.)

This is being played out on the blogs as Malcolm Gladwell versus Duncan Watts.

I think neither of these views is right – this polarisation masks the real complexity of influence, which is that it’s damned hard to pin down in what it is and how it works.

I can’t criticise these guys for saying it how they see it. In fact, I think the biggest culprits are consumer-facing WOM agencies that claim to be able to identify influential consumers or, worse, to position celebrities as influencers.

The podcast does actually acknowledge that influencers do exist, though these may be the people that “show up.” In other words, anyone can be an influencer if they are committed and diligent enough. I think that this is true in large parts.

A couple of their comments jarred with me:

“The Influential model is most often touted by people who would like to be seen as Influentials, or at least, friends of Influentials.” Ouch. In fact, I “tout” Influencer50’s approach because I see it working with clients. Some influential people don’t even know that they influence the market, and are surprised on being told such.

The idea that once you find influencers it’s a simple task of pulling the levers and success follows. My experience is that although identifying accurately is complex, it’s actually the easy part in the process. Engaging with influencers is much harder.

It’s also cemented my view that influence in the B2B world is different from B2C, in that B2B lacks a strong sense of peer-to-peer communication. Business people don’t talk to others outside their organisation because of the lack of opportunity, or due to competitive sensitivities. Influencers act as proxies here, acting as go-betweens for firms. This role is critical, and underpins the entire consulting and industry analysis business models.

In B2C, sure, there are influential consumers. But I’ll bet that no agency can identify which fellow consumers are influencing me on my (ongoing) new PC decision. But they could identify which web sites, retailers and magazines I might consult. Fixating on consumers as B2C influencers is missing the primary sources of influence: the supply chain and value-adding influencers.

It’s clear to me that most firms looking for influencers amongst consumers are looking in the wrong place.

More on Duncan Watts’ theories

I suddenly remembered, after posting on Duncan Watts’s theories of idea spread and influence (or lack thereof), that Prof Watts currently works at Yahoo!.

Spookily, we included a case study of Yahoo!’s use of influencers in the book. I wonder how Prof Watts is getting on with Yahoo!’s influencer marketing team…

Duncan Watts – influence killer?

There’s a bit of a stir going on regarding Duncan J Watts and his theories on how ideas spread. In particular he’s becoming known as the influencer killer because he says that influencers have no catalyst effect on the spread of ideas. Indeed, they have no special role in trends at all.

I’ve been following his research for about a year, since this was published. But the recent article in Fast Company magazine brings the issue to a head (and I think Clive Thompson has done a great job in raising the relevant issues).

As I see it, Prof Watts highlights some of the basic misunderstandings of influence. There is, frankly, a lot of nonsense and assumption talked about influence. Much of it stems from the Keller & Berry book The Influentials, which states that 10% of people tell the rest of us how to vote, where to eat and what to buy. I loved this book. I’ve read it three times. Nowhere does it say how Keller & Berry came up with the 10% number – why not 9% or 11%, or 1%. It is based on assertion, and the data in the book simply reasserts the 10% assumption. The Influentials is really useful in understanding how people gain influence in society. But it’s not useful in understanding who or where these people might be, or how to get them to influence markets on your behalf.

Another key misunderstanding on influence is that the people with influence are marked out through the roles they play, or the jobs they hold. Or worse, through the degree of celebrity status they hold. The consensus of research shows that we are more influenced by people with expertise than people with celebrity. And we are more likely to be influenced by people we know than by any other group.

Equally influencers don’t influence evenly through a decision process, as I wrote in this post and the book. And influencers aren’t influential in every category – influence is context-sensitive. So an influencer in buying a house may not have influence in buying a digital camera. This should be intuitive, but marketers sometimes get carried away by the promise of influencers that hold the keys to increased sales.

However, it’s clear that influence occurs in markets, and it’s attractive (I say essential) that firms understand how and what this influence is, and who are the conduits of it. So the final misunderstanding of influence is the assumption that it exists only within the consumers themselves. Clearly this is nonsense when you think about it (except marketers don’t tend to think about it). As consumers we are influenced by a host of people who are not (in the roles they play) consumers.

We’re clearly influenced by a range of people in different roles, some local, some national. What we do as decision-makers is to process the variety of influencers and then make a judgement. Not all influencers are successful in influencing every decision.

Consider the process of buying a house, something that most of us will be familiar with. You’ll be influenced by the house vendor and a real estate agent. Both have a vested interest in the transaction, so you do more checks. You’ll be influenced by any people that you know of within the area. You’ll be able to check out whether a ‘neighbourhood watch’ system is in place. You will consult published crime figures, possibly even calling the local police station for advice. You’ll refer to school inspection reports and league tables. Local councils will be quizzed for pending planning applications. The local head teacher may be critically important if the main attraction for the property is its proximity to good schools. Builders and other tradesmen may be asked for quotations for structural work you want to do.

I see this even more pronounced in B2B markets, where we can identify the individual people that influence a specific market segment. These are people that are not particularly obvious but undeniably carry influence in their area of expertise. B2B influence works in a different way to B2C, primary because B2B markets don’t have the word of mouth communication that B2C markets do.

But even in B2C markets we find that influencers in the supply chain and its ecosystem carry the majority of influence, with consumers having limited impact on the market as a whole.

I don’t believe that Prof Watts signals the death of influencers. But perhaps he has hastened the death of much of the nonsense that surrounds the notion of influence.